Discussion:
[Freetel-codec2] RFC: 1300C
Jeroen Vreeken
2017-03-10 11:12:16 UTC
Permalink
Hi all,

I have been experimenting for a while to extend the 700C codec2 mode
into a 1300C mode. Mostly by just trying out different things like
adding more bits for pitch or energy and trying to create better codebooks.
It is a fun way to get to know the different parts of the source and get
an understanding of the codec.
I finally settled on a frame format that adds a 3th, 4th and 5th stage
to the vector quantization of 8bits each.
The first entry of the three codebooks contain all zeros, this is to
ensure that the additional stages will always have a result that is at
least as good as 700c. Also transcoding from 700c to 1300c is simply a
matter of zero padding.
For the training I used over an hour of material that could be heard
over a typical repeater. (It started with some recorded QSOs, but ended
up with mostly bulletin samples).

A patch for this new mode is attached to this mail. Personally I think
it might be a good candidate to use for the 2400A/B/T FreeDV mode.

At http://video.vreeken.net/~dml/patches/all.1300c.wav you can find an
example of the output.

One of the things that I noticed while testing was that the current
pitch estimation sometimes makes a mistake and ends up with twice the
real pitch. I also got the feeling (but no proof yet) it was worse with
samples with heavy filtering under 300Hz (e.g. to get rid of CTCSS tones).
So there might be some room for improvement there...

Regards,
JEroen
Jeroen Vreeken
2017-03-10 12:51:56 UTC
Permalink
To make comparing the sample a bit more objective, here are the error
values of the all.1300c.wav file per 40ms frame.
First two collums are the 2 codebook entries selected for mode 700C and
the third column is the calculated squared error.
The next five columns are the 5 codebook entries selected for mode 1300C
and the last column is again the calculated squared error. As you can
see it is not unusual for it to be four times lower.


233 422 147.647919 503 505 11 179 219 44.553898
325 510 181.368240 451 21 21 78 100 45.464817
451 258 148.718140 451 258 16 21 29 45.768272
228 80 90.499649 34 237 37 123 121 43.202187
228 37 157.971237 228 37 2 37 39 49.921417
508 151 107.038239 508 151 39 201 79 31.536287
140 167 102.829575 436 164 23 46 95 40.998692
163 495 99.522934 163 78 181 102 194 43.852077
386 386 210.730103 386 140 39 61 55 46.412121
386 295 246.201874 386 295 160 49 9 41.924938
8 253 150.807434 8 396 145 142 15 31.501804
395 290 281.666107 395 57 24 68 128 52.487968
156 378 298.061066 156 378 160 199 163 46.268307
252 140 143.043137 252 140 32 58 31 49.562840
137 281 300.300781 137 493 127 77 152 42.387012
87 41 328.088165 49 149 85 96 191 82.831955
91 313 92.516937 91 342 222 67 22 36.041050
128 459 180.377136 278 100 108 182 209 60.553558
49 252 139.408264 49 252 15 26 19 40.837978
289 75 131.074295 248 322 90 59 103 17.612642
355 146 265.222168 355 407 222 71 99 73.439407
111 56 92.347755 111 56 0 64 3 41.532867
265 239 157.375458 411 393 0 93 224 53.163589
241 66 153.000595 241 399 253 64 63 47.846157
326 10 138.359451 326 240 210 63 123 41.266605
44 109 127.177170 44 109 9 62 65 45.500179
224 378 201.815964 69 475 32 136 138 49.213757
288 346 72.265251 288 359 16 156 0 42.942123
178 510 148.850525 178 146 1 61 113 41.565170
333 180 312.276978 333 180 145 133 57 69.454086
451 193 69.917313 404 135 16 27 56 32.561127
451 285 133.718277 451 30 151 173 187 29.812428
140 167 88.817291 436 164 23 39 74 37.682289
140 322 169.870010 140 158 169 39 91 56.077801
391 362 429.265015 352 345 116 115 48 71.408165
116 307 155.442047 355 19 142 63 67 60.193554
87 339 119.485100 87 339 151 102 66 45.062138
91 206 117.031303 91 383 105 58 169 54.464661
219 186 126.147766 219 186 16 131 170 28.837233
505 161 286.248840 505 161 85 109 249 52.463730
3 229 199.624817 3 229 7 239 8 55.336121
411 479 218.442596 411 479 86 49 19 56.919838
434 32 183.146500 434 140 116 49 39 50.288322
371 405 109.610397 248 436 22 45 184 43.478477
74 362 127.279114 74 362 0 3 54 48.129131
94 101 229.217987 94 101 70 37 41 73.690918
264 251 147.646072 264 189 73 108 136 37.705185
63 479 218.821442 63 502 110 180 150 49.792801
443 22 166.788361 443 22 37 40 33 65.506363
190 124 372.739441 190 124 36 171 20 70.339455
401 378 140.835541 198 378 41 197 127 38.892132
309 11 152.522980 209 115 212 61 177 57.784550
441 24 155.904083 441 310 7 180 5 33.559052
246 140 199.056381 445 239 88 70 130 45.895939
280 341 178.415756 280 341 32 21 60 44.076591
280 55 240.220337 280 55 251 174 182 52.880894
280 55 246.933685 280 55 173 59 250 48.752373
215 183 294.185394 215 183 6 120 200 68.541962
240 374 589.771912 467 479 87 248 178 153.107224
190 164 521.321350 190 164 107 113 189 59.873711
374 295 324.502533 374 295 33 15 54 71.200981
374 376 187.600235 374 376 235 59 196 42.682747
144 471 232.218857 336 424 235 157 58 69.440002
343 484 107.534515 343 408 1 71 207 45.475224
451 510 210.914673 451 510 4 156 209 42.572525
451 483 94.179573 228 253 29 12 0 23.819105
228 253 86.819855 228 253 29 55 63 35.485752
451 270 164.034653 451 195 70 122 120 46.696651
451 195 102.187294 161 317 4 74 25 40.133240
451 452 80.114838 451 452 15 39 59 41.008793
501 349 179.203720 501 349 62 48 32 38.993614
228 106 98.077240 490 463 25 52 120 35.614235
269 301 194.679169 477 463 222 73 92 44.784603
343 222 166.625000 510 6 21 119 0 47.770222
451 510 126.503082 451 452 21 21 59 39.129257
451 324 109.060669 451 65 25 3 240 38.090450
295 87 234.316742 295 87 181 55 239 42.107018
17 209 164.704346 17 65 11 6 216 42.009460
295 80 83.156197 295 80 37 193 200 32.770020
510 186 162.467087 510 186 4 13 77 39.908054
228 510 97.679512 228 510 25 60 232 37.310242
85 464 182.283112 85 464 2 26 220 49.726727
228 119 71.275261 228 493 127 43 176 40.666931
278 277 155.832123 92 162 58 128 83 46.144493
165 24 230.953735 165 24 34 14 43 46.388229
504 257 161.134201 504 180 142 82 184 40.884586
504 133 192.721268 504 133 66 84 159 32.248768
278 392 169.781036 504 339 66 249 41 44.624557
411 239 242.630920 411 239 210 63 96 57.502659
441 271 338.805573 441 314 27 119 96 82.635696
439 316 267.039337 216 250 134 82 159 51.129704
382 370 158.099899 382 394 160 46 204 46.497486
511 475 126.148911 511 475 134 82 90 46.956760
316 201 193.514938 316 201 221 144 135 66.750290
278 180 346.641968 278 180 34 174 191 101.094009
278 235 278.418610 278 57 168 150 186 70.582520
278 94 377.501556 278 94 48 32 93 86.529900
278 351 82.235809 278 351 27 80 99 39.739120
411 339 137.460785 411 76 207 82 74 53.458317
123 170 97.023842 426 285 37 15 0 31.428448
123 87 259.639282 123 87 74 51 20 59.508896
123 378 120.335922 123 378 11 98 0 34.948620
382 375 120.038483 382 375 143 54 111 41.571789
241 386 115.464653 241 144 23 31 21 41.859516
374 135 219.017807 374 135 14 11 229 42.774288
170 378 393.866211 374 379 186 64 243 114.211861
278 427 137.561661 278 196 186 71 24 36.868145
472 295 356.442963 93 226 99 86 194 85.974640
144 307 207.329681 144 307 127 139 18 60.010933
504 464 304.849640 278 277 126 76 114 57.215942
458 421 202.581253 458 421 58 33 172 51.385429
85 508 402.402924 203 201 58 83 50 84.216980
203 345 270.876648 203 345 76 68 106 48.698490
385 392 170.233795 385 392 235 49 111 43.972664
203 119 86.822021 203 119 218 48 232 40.068478
203 422 202.442596 203 394 168 74 234 43.222450
104 110 275.648926 104 373 13 41 249 49.043472
104 348 266.199371 104 271 28 197 86 52.214844
104 252 191.544662 104 195 19 5 211 43.213074
358 195 187.263733 358 75 32 33 115 52.057529
123 167 164.245819 123 167 7 80 50 32.579319
504 376 171.398941 504 112 178 187 114 46.229137
504 277 138.936096 504 376 196 31 104 28.474892
504 78 94.993759 504 78 4 61 92 45.096107
278 277 130.231995 278 273 58 0 250 30.309448
458 24 112.469917 85 10 145 69 0 32.034763
239 510 128.195053 239 464 20 66 8 43.673615
411 251 360.230011 411 251 37 61 216 118.545074
18 475 183.174393 278 337 66 204 177 54.846336
18 475 191.226074 504 290 176 116 180 53.586987
504 386 248.484558 504 471 218 100 112 60.656765
336 339 473.547241 161 66 134 139 96 119.135857
374 369 230.323227 120 118 222 78 112 53.549206
120 119 123.004074 120 247 29 158 228 26.175037
374 363 251.198105 374 74 186 57 59 71.274734
120 283 226.418976 120 205 110 40 46 70.797424
146 479 256.512573 146 479 74 61 88 63.681484
433 479 138.461685 433 479 234 103 130 38.755703
459 446 66.830429 459 446 0 14 21 36.754955
241 130 111.982574 320 10 152 152 2 29.169571
62 328 200.562531 62 328 7 139 176 56.256886
343 187 132.245346 381 483 7 162 233 21.865145
435 474 202.272583 435 277 8 12 233 51.620163
239 146 117.966721 239 146 12 54 62 39.960644
510 349 134.886719 228 510 17 84 166 40.897224
239 341 196.233215 228 341 39 31 103 50.234772
510 349 112.233604 510 349 16 52 32 48.926579
249 106 115.316940 249 106 8 31 0 35.920151
343 504 127.106728 17 151 16 42 215 40.367420
182 229 200.424561 182 178 53 33 204 29.265814
277 178 266.337952 477 318 6 106 218 57.603645
355 292 162.089249 355 253 8 3 96 31.156256
343 292 88.509003 343 292 0 48 0 38.736374
192 344 130.198730 192 344 4 99 236 37.388569
343 333 138.942383 343 333 11 79 15 37.133476
116 229 270.158508 116 229 142 23 110 45.011440
52 510 87.352119 52 510 0 18 93 45.209808
325 510 143.634537 325 510 9 91 43 39.347355
194 510 205.536011 257 510 53 129 82 58.083015
506 68 85.273178 506 68 0 49 106 42.576874
147 301 181.074051 147 301 16 32 118 50.069469
153 474 82.073601 153 459 5 57 93 33.566837
153 220 173.064240 153 220 26 13 137 38.424530
153 407 110.639366 153 301 9 5 42 39.026779
347 65 142.518524 347 138 7 42 94 46.708382
110 295 247.107162 110 241 53 105 17 54.594131
116 317 119.566788 116 317 12 0 176 29.083282
352 229 94.294914 352 498 33 5 49 43.035069
335 424 200.365997 179 317 62 30 50 55.245422
174 186 170.652405 174 219 2 70 202 32.857292
482 505 173.127411 482 505 74 11 236 53.686886
287 395 116.955307 287 395 14 79 47 39.602467
480 178 293.381958 480 151 136 200 210 66.143051
458 194 366.798889 458 173 70 136 49 80.977913
460 5 272.445068 460 401 55 152 22 46.699188
17 398 152.638901 17 398 22 51 95 37.531593
189 258 223.314636 52 162 12 25 219 61.376823
477 474 378.934601 325 510 89 180 120 75.793541
503 161 228.985626 325 510 74 51 254 38.478050
3 384 169.252609 3 140 14 166 62 47.930191
3 449 156.154449 355 209 234 7 45 42.810516
339 27 201.400146 339 27 16 187 10 48.779484
295 200 173.968536 295 200 142 5 5 54.538311
263 110 179.021729 263 375 8 10 37 40.011387
487 317 318.863495 487 392 76 181 69 57.584518
487 236 268.247101 487 114 82 26 233 50.507023
263 138 163.310547 263 138 134 168 63 55.016895
135 123 88.396645 135 390 6 57 175 30.236193
486 504 125.991142 486 504 5 13 0 41.193729
486 504 113.581985 486 317 33 160 230 40.373703
364 438 245.013748 364 254 10 50 47 54.621643
486 344 228.499893 486 344 26 189 26 28.818779
390 76 227.955063 486 358 13 6 100 43.732052
34 317 198.521210 34 317 31 44 139 38.810276
221 139 327.343689 221 139 21 164 166 57.542618
335 510 318.242432 335 236 2 42 12 43.897850
54 510 189.270309 54 10 12 5 215 43.278805
52 460 201.484268 341 401 145 137 41 31.246778
269 62 343.178986 269 62 198 160 94 82.395134
332 261 344.598450 371 344 223 44 254 57.759369
249 423 267.760376 477 339 1 3 10 48.280682
258 510 97.456078 258 510 29 23 150 36.042400
159 423 322.590057 159 178 21 105 248 63.374748
6 510 235.354126 429 349 40 28 152 29.500975
95 248 88.906479 322 108 22 62 53 39.844501
100 435 322.427612 15 80 1 31 78 38.104881
257 510 222.856918 257 510 151 4 65 50.386269
349 358 89.922020 350 96 13 151 0 24.581480
48 17 168.921677 48 17 10 176 255 45.346428
272 483 71.282883 272 483 23 181 195 39.135277
142 510 168.095856 310 223 2 50 17 44.568478
358 492 85.421181 358 296 0 145 152 44.913574
17 511 246.651535 182 332 220 135 174 38.448586
329 28 238.350922 430 435 40 131 147 51.102749
433 510 246.757217 433 106 2 123 44 50.862461
145 58 150.312836 145 58 10 65 44 47.601433
417 417 285.235016 417 417 24 39 131 42.146423
61 80 306.936005 61 381 1 123 193 70.669983
183 80 240.914185 183 80 2 104 181 47.101334
123 510 384.395538 123 510 44 8 10 52.057308
329 165 281.826385 329 240 14 112 2 77.577774
181 237 186.874191 181 237 21 82 136 50.378815
417 150 59.091526 417 150 33 38 150 31.866770
89 129 190.633209 237 401 8 39 53 49.474411
201 254 276.309479 201 254 4 18 243 70.407616
48 375 253.680634 358 298 1 34 50 51.268711
436 459 389.771851 311 61 24 5 6 40.219494
335 463 301.925323 54 459 18 158 208 51.866928
54 349 151.886841 54 349 24 29 112 42.789013
335 6 331.694641 335 391 21 22 96 55.959045
329 326 230.071625 329 391 17 130 82 46.681980
315 473 198.854950 315 473 2 47 144 43.836811
421 380 184.206879 421 380 181 108 106 50.503910
44 80 116.859413 44 80 13 37 200 27.257744
472 510 197.322464 255 252 18 0 53 41.251579
278 333 97.667343 504 463 70 84 0 35.577976
18 186 131.909485 405 337 134 104 0 21.519114
18 344 192.890839 18 483 145 189 15 34.800640
24 349 188.128174 24 435 2 22 133 29.597569
182 510 180.082443 239 435 252 89 64 45.530792
148 510 104.977905 133 435 3 33 184 36.159462
335 341 172.186478 335 341 1 109 34 51.494316
148 510 182.695419 148 510 1 168 101 44.043751
133 349 247.681870 133 6 17 6 83 41.856827
413 199 140.644608 349 178 1 77 65 32.262177
474 313 98.795944 474 313 22 0 164 39.546806
349 445 180.875137 349 299 41 57 192 51.825001
474 178 67.767838 474 229 23 0 229 30.218361
474 178 192.194153 474 510 19 67 166 48.587341
81 463 212.329697 107 27 21 56 207 49.662514
413 380 206.716309 413 380 22 85 41 41.259850
107 415 107.704369 107 415 17 0 134 26.594116
1 415 144.769577 474 64 12 73 253 35.272659
189 460 167.326950 189 42 1 2 245 33.280163
349 349 90.204231 349 204 62 45 11 42.548073
133 178 167.056198 81 169 1 51 249 42.984161
335 349 399.844238 335 100 50 67 60 58.604259
329 151 338.842468 329 17 1 22 164 49.336372
364 122 419.889587 477 122 28 13 24 48.162663
364 178 511.318726 477 73 180 199 19 77.418495
181 146 272.835205 133 214 68 14 83 63.274105
492 122 390.854889 492 373 161 192 240 61.385036
249 362 191.785950 249 362 131 42 123 51.444504
390 156 239.121368 390 156 17 200 249 43.727821
390 47 257.870667 390 324 29 188 147 59.880806
314 469 168.264359 314 220 253 69 47 55.649929
81 435 109.216400 81 489 23 2 208 30.149994
133 384 227.298126 133 384 24 98 100 53.991497
31 366 122.603653 31 366 6 71 243 43.139828
31 178 187.797897 31 178 11 133 250 39.217407
132 178 473.191803 132 56 139 177 210 66.341393
37 178 181.717911 37 178 4 185 27 49.471935
60 131 97.474792 60 131 5 85 228 37.330070
17 324 252.029572 17 113 236 126 222 56.622406
335 225 263.700195 335 225 53 159 134 44.868179
133 373 236.462860 133 384 2 132 213 62.575714
413 384 164.151428 413 164 23 129 222 45.890434
108 144 341.660645 108 144 249 71 178 85.662056
135 295 253.072800 315 223 119 72 59 43.613934
135 226 318.433899 135 154 72 33 23 71.809639
135 504 259.748901 135 427 21 34 152 47.430466
345 349 320.098877 345 459 28 67 152 50.958382
390 80 186.760101 390 80 6 178 58 40.168888
249 122 413.818420 249 122 213 100 251 88.958443
486 384 290.774170 486 384 18 43 79 47.076160
345 56 287.781982 486 473 46 139 15 37.855244
135 473 396.501282 345 318 123 153 87 42.879585
181 155 352.224365 181 155 2 91 70 72.382980
133 165 217.943085 133 165 9 4 116 46.869568
329 349 288.300842 329 349 141 142 180 52.834007
148 186 285.914490 148 510 151 46 210 47.524330
413 252 238.904343 413 252 28 176 64 54.312027
486 201 406.626465 486 201 28 66 145 68.914185
486 435 582.723633 345 303 56 199 132 75.046631
345 56 542.413574 345 56 226 160 6 62.693348
364 435 547.637573 345 505 238 23 2 60.445408
364 373 212.645477 364 373 23 82 93 50.355743
135 381 122.305786 135 281 5 52 80 37.376141
340 358 254.119400 340 358 25 2 67 42.265915
415 165 235.309875 415 165 123 180 117 54.738956
340 459 160.101242 340 84 25 95 44 40.860294
17 384 158.082748 359 61 25 28 139 38.349548
17 110 386.239716 17 186 208 109 109 42.562607
135 178 429.225800 406 178 80 63 59 47.152443
477 415 313.900635 477 415 70 73 212 52.837177
148 510 294.066254 148 510 24 21 211 27.031700
329 510 306.401276 335 6 24 2 134 51.507977
477 362 447.535889 477 322 122 115 175 77.936913
413 435 146.567947 413 435 7 88 162 52.427139
450 110 249.523941 450 110 148 129 16 42.935234
122 358 182.356308 122 358 43 10 71 40.571388
122 5 253.141861 122 122 254 41 181 51.130211
114 422 256.876801 338 435 34 30 129 61.652111
500 474 343.076111 450 161 21 38 9 78.343872
500 510 226.485519 500 474 27 4 239 34.825642
350 225 208.132370 350 225 88 9 138 43.513119
113 122 241.361328 113 181 2 13 19 50.337078
413 96 163.465851 413 396 253 34 129 36.212509
349 33 253.573761 349 33 25 126 10 65.854645
236 165 252.177063 236 165 12 6 172 50.155025
494 503 309.414673 494 503 218 18 41 59.059231
45 199 496.044769 478 373 72 164 85 67.242134
415 122 221.690384 415 122 62 157 206 54.130291
332 269 326.054199 332 269 98 24 88 53.817867
105 229 156.786285 105 229 6 71 76 36.762486
214 324 243.874084 214 324 36 34 66 46.924377
507 83 317.828400 507 415 116 155 6 55.506199
53 61 302.631195 53 173 130 126 39 43.440712
60 476 265.870178 19 220 129 70 46 47.467262
250 396 276.598969 250 108 24 191 170 57.361050
58 474 124.527641 58 474 13 9 34 36.006855
182 423 325.231018 182 423 185 163 154 88.259148
450 423 220.631653 450 510 29 6 72 62.953739
31 56 239.018646 432 423 127 58 161 65.508507
31 492 168.153702 31 492 16 34 109 53.099895
404 468 166.901901 404 219 12 26 130 40.258400
135 450 226.356049 360 204 32 190 126 51.697121
182 324 283.983948 182 288 21 6 209 63.582184
182 288 313.769714 182 37 236 104 191 73.353485
349 435 111.060486 349 6 1 52 93 36.004604
99 159 276.033142 461 496 2 184 83 59.719421
480 318 339.910889 480 318 224 176 163 65.457108
347 100 259.421692 176 301 99 201 41 38.962120
306 80 196.801590 306 80 216 115 232 35.598560
132 349 140.777054 31 351 47 38 119 38.229294
192 1 288.850037 276 299 31 94 4 38.712978
414 459 262.482086 414 101 70 4 98 32.127850
236 396 214.706909 236 113 31 15 241 31.056892
461 459 255.851318 461 390 2 142 124 65.254333
413 151 151.870178 413 145 24 88 185 39.640770
313 56 338.673767 313 496 24 185 105 80.790581
332 504 210.930420 332 194 173 77 93 39.738956
31 317 325.959656 31 390 35 58 125 64.372330
306 200 251.686996 239 275 99 50 109 72.417549
182 268 227.625137 182 268 39 17 99 53.669231
235 435 416.023987 235 155 216 62 248 49.758518
35 463 167.047119 35 463 181 174 219 38.410957
31 351 191.548828 31 2 5 86 86 39.786945
435 510 269.248535 435 510 127 192 3 64.688934
210 510 364.836426 210 510 252 20 24 62.890697
413 299 163.754379 189 373 29 6 127 42.010044
54 510 139.472504 54 505 2 19 56 41.948162
243 182 562.665405 243 61 198 55 235 114.999527
31 151 92.924286 31 121 5 15 83 40.084400
31 454 109.741699 31 454 1 63 130 49.920334
122 200 176.176498 122 510 108 50 49 47.615810
189 56 263.616425 413 53 6 122 154 57.474846
325 396 113.304916 325 396 41 188 0 23.334967
31 510 77.070221 31 474 16 0 65 41.204967
370 450 137.019363 370 102 39 126 106 41.948460
306 510 124.880241 306 200 6 89 3 41.368443
243 178 117.972206 202 396 1 30 38 39.477562
107 489 85.446098 107 489 0 86 155 27.965599
202 122 179.576416 202 122 90 6 58 45.596432
413 122 156.350006 413 122 14 67 236 41.873314
494 294 252.199249 368 459 100 56 40 61.773361
75 435 242.651398 75 394 67 136 236 42.864998
81 384 139.276108 81 384 2 185 15 36.259575
133 96 169.992737 53 96 20 35 117 37.242477
181 407 104.697845 181 407 49 33 0 42.826145
335 96 211.893387 148 304 12 190 7 57.624748
133 384 269.619019 148 341 17 123 190 48.851448
148 341 198.406342 148 156 179 28 3 48.897285
133 373 160.840408 133 6 47 98 26 47.255619
133 96 194.663422 133 473 24 103 18 35.181946
148 304 269.410034 148 84 31 49 44 44.293282
133 140 178.602127 133 386 2 18 166 46.458385
148 341 267.881287 148 341 24 68 167 37.918552
133 96 298.917877 466 225 21 175 217 67.251053
466 373 190.647491 133 156 3 159 55 37.129463
413 269 183.363907 413 2 70 157 179 52.438866
53 407 265.491516 53 407 12 51 41 52.137581
53 107 329.426941 53 107 119 44 30 65.572563
53 107 187.529831 53 107 62 156 90 59.590279
53 44 105.151169 53 182 2 80 0 26.774677
53 83 385.143768 53 83 150 36 138 67.169411
109 32 188.113617 109 32 6 91 138 37.061695
327 209 232.550903 327 422 26 20 9 52.717365
474 415 177.038544 341 384 2 128 97 41.000015
413 370 133.549576 413 225 7 68 59 36.056927
202 147 230.723907 202 393 216 113 239 41.640762
432 378 136.219666 334 14 6 197 60 42.292229
17 112 194.121033 98 396 2 108 33 44.365376
261 235 202.410339 261 186 218 2 140 42.266262
359 187 201.225433 113 214 210 146 162 43.930191
111 288 226.858047 111 489 22 39 102 50.840778
314 423 290.093048 314 178 72 83 130 55.842354
329 229 315.397980 329 229 139 104 150 64.574310
427 423 603.357300 413 449 147 114 214 96.532448
413 68 256.145386 413 272 13 146 120 50.147644
58 229 304.310547 58 229 124 167 75 50.621483
201 57 93.034401 201 57 17 9 60 35.875923
340 57 334.789246 340 386 45 85 163 39.579079
314 386 267.537720 33 100 57 122 26 42.718647
335 415 327.822754 335 199 188 26 28 42.299301
329 229 534.291138 329 229 106 121 160 119.410362
68 423 393.283447 68 324 80 77 238 63.600616
329 37 547.594727 329 37 203 103 6 47.240463
148 423 373.683105 329 229 62 22 102 77.997551
335 110 714.789185 335 61 122 36 13 93.534973
413 343 548.034058 335 274 139 161 187 105.143097
202 510 171.725037 202 510 10 65 183 32.677959
335 510 308.390198 335 510 62 97 207 60.289001
62 463 394.989777 62 463 206 170 112 56.615082
59 55 265.130768 59 277 18 135 46 60.204945
141 260 163.428619 141 121 16 56 218 37.774696
114 505 112.692474 114 103 16 73 106 42.091656
31 55 135.118286 31 55 1 101 3 41.758308
202 151 151.023605 202 165 29 120 170 35.658493
122 476 275.936249 122 350 73 83 138 36.724121
122 200 330.693207 71 435 17 39 75 73.819077
250 301 376.121979 250 301 36 52 148 59.050793
250 83 476.728516 250 83 123 186 27 101.003113
480 324 404.993347 480 324 18 181 109 42.535606
282 151 198.290070 282 151 1 97 0 42.922276
367 199 253.536377 367 56 196 186 2 52.276615
189 168 238.248138 500 168 47 111 202 37.411728
185 459 210.781433 185 459 131 105 67 39.929676
383 510 340.450348 383 510 44 105 195 58.977131
196 349 310.271240 196 349 24 184 219 71.041565
239 423 283.667786 239 423 41 97 33 70.512558
442 373 262.208832 313 5 165 65 15 58.742386
436 432 260.081421 436 432 70 67 174 66.340218
31 268 302.207733 31 268 131 184 78 69.895996
503 200 421.269043 503 200 245 87 195 104.590813
249 288 750.672180 24 288 99 154 24 198.890167
261 85 273.623413 261 83 131 75 85 60.567589
390 308 289.656982 112 220 192 87 143 39.923637
343 139 338.874939 343 2 189 17 162 69.983864
343 387 173.756546 343 387 10 87 23 60.641930
137 417 235.709534 137 417 7 137 26 39.453682
486 324 228.436386 486 373 39 137 59 42.511158
364 503 483.388306 364 503 255 32 26 87.994408
486 331 251.426422 486 331 26 132 165 74.734505
470 209 210.151947 470 209 3 85 126 43.130127
137 390 501.485687 111 407 180 186 225 91.520477
135 10 482.010468 360 510 101 86 216 68.391960
151 463 316.710999 381 101 174 110 204 75.374908
189 505 168.797043 189 151 43 31 43 35.438747
413 122 308.976288 413 373 42 157 102 65.141724
133 178 172.311584 53 178 21 82 0 36.084698
79 510 235.941269 503 453 152 8 27 50.019543
511 510 505.017334 511 510 18 50 200 74.298225
62 510 372.392426 62 474 106 45 185 63.241489
296 504 464.148315 410 450 18 40 252 55.003971
339 106 415.946381 339 237 24 126 21 44.497242
50 80 610.353210 50 234 57 178 27 127.399391
236 156 160.630936 236 156 177 31 181 30.946772
146 96 331.339233 146 96 18 51 227 24.774479
236 407 248.969131 105 100 32 3 103 59.423054
199 316 205.806229 199 316 218 25 230 44.451485
332 424 150.772385 495 56 93 200 216 37.515072
343 396 178.337234 343 122 6 62 0 36.822884
269 6 146.214478 269 453 218 3 50 37.309151
135 446 197.571564 135 446 2 71 9 46.210270
345 459 216.274155 345 349 12 85 107 43.580421
433 358 264.287842 112 220 228 128 60 42.257889
276 35 184.867493 276 35 31 200 75 44.975544
343 100 275.453949 343 100 17 55 252 48.965454
413 89 298.001862 466 309 6 104 170 60.732758
327 25 300.168945 327 25 113 163 104 67.439255
450 510 244.578751 103 122 57 170 92 39.042446
31 317 277.025970 31 317 73 4 36 54.164722
210 299 390.856812 250 473 24 13 137 74.936188
77 510 501.462830 435 510 237 136 166 68.553558
150 1 347.124817 31 222 94 165 235 45.456646
338 484 206.814606 151 304 143 97 242 50.247417
249 288 384.050049 306 349 159 96 105 82.609657
249 80 400.590424 249 80 69 65 245 101.867226
477 299 160.724945 477 299 6 86 41 42.253899
413 317 182.709564 54 435 207 25 2 47.398659
398 268 436.065063 398 268 44 143 74 50.868774
215 375 378.378845 215 375 64 40 150 71.351257
471 344 587.607178 471 114 229 151 17 123.034714
18 354 925.058167 18 135 136 246 102 184.672882
26 375 564.334534 26 375 19 243 137 129.548569
295 403 299.478180 295 403 16 131 97 87.033020
18 343 373.814789 18 343 29 16 140 107.099388
228 343 393.327179 228 343 181 216 157 98.066360
116 5 310.864868 34 236 44 197 63 42.164474
364 496 1045.624023 364 496 128 70 79 147.172897
364 5 520.951050 364 317 254 138 42 79.833176
390 104 490.325989 390 104 32 190 200 113.157097
83 85 728.741699 83 503 115 131 122 166.361252
128 222 981.082642 128 222 225 31 218 107.910973
345 496 232.749176 345 401 57 32 87 44.952789
364 369 856.649109 364 369 92 140 66 102.418694
38 375 434.218536 38 375 255 107 169 69.152023
38 294 622.221375 38 294 19 214 217 102.580177
38 1 478.447601 400 60 101 60 112 86.179062
116 128 130.190231 116 128 0 50 72 36.825951
116 60 283.032379 116 445 222 153 112 63.671860
458 261 252.404053 116 403 231 57 141 69.103790
116 285 279.061981 116 475 3 7 240 60.259464
116 236 283.256042 116 236 21 16 57 50.462460
473 161 703.780273 364 344 226 3 242 59.680908
364 510 504.145355 364 510 226 67 242 74.799683
390 392 478.734741 390 104 127 246 190 90.076401
38 85 451.492432 83 222 176 40 200 100.651031
83 85 547.667847 83 222 214 171 9 67.177956
390 423 535.456604 390 435 82 148 173 82.256317
240 423 412.075378 240 246 57 104 222 65.963257
128 503 448.236694 128 503 147 91 140 49.208725
38 85 379.851776 38 85 13 9 187 92.096931
400 60 496.329163 400 60 249 57 200 109.371330
219 451 453.679260 219 451 190 77 103 86.714378
116 222 443.455902 116 507 5 223 120 45.624802
116 285 205.107254 116 483 7 7 209 37.726471
295 317 380.332611 295 222 226 136 120 40.903961
295 511 556.687744 295 233 183 121 225 74.696594
116 60 239.233322 116 60 68 65 251 52.036221
128 222 434.070831 128 155 225 103 96 59.934429
128 222 408.646973 128 222 79 69 94 74.214073
249 222 514.147888 390 116 246 47 121 100.528419
364 435 566.628906 364 161 225 86 41 52.680771
364 201 984.655151 364 201 147 140 199 157.813828
128 503 469.800018 83 155 214 171 14 58.912674
400 503 528.521790 400 104 170 56 225 90.587128
355 503 305.612854 355 503 28 113 36 50.532936
116 75 198.673676 116 285 186 122 211 51.370380
116 75 336.261475 391 124 186 68 47 63.548328
116 350 251.800903 187 60 186 42 139 51.244854
34 476 284.528076 116 496 166 191 127 41.535210
361 344 251.330917 34 161 5 29 40 58.676086
135 5 310.915405 182 129 188 72 12 49.578995
391 454 280.253754 391 454 166 117 211 61.142204
239 387 661.064941 239 387 124 70 138 104.957237
480 268 249.534683 480 450 108 111 60 61.198166
173 392 349.157043 36 469 85 88 42 65.730858
458 105 342.878418 458 105 58 196 195 76.079346
247 511 592.083984 247 511 219 164 106 94.285522
187 268 530.190613 250 222 207 29 66 129.861557
116 129 408.187805 116 506 226 174 90 70.677467
128 100 559.670959 91 511 71 172 92 86.480682
128 489 594.752075 128 463 115 48 199 79.450333
34 129 406.807312 34 129 158 129 119 62.317123
364 483 732.162598 364 131 42 184 236 82.488785
364 5 587.449219 364 209 184 114 61 70.891861
91 131 414.245850 91 131 226 64 66 59.001953
83 222 539.111877 83 222 176 103 122 81.421532
38 85 618.354919 83 222 176 40 9 164.024963
128 85 651.091187 128 489 92 31 96 60.519157
345 229 222.414185 345 366 45 71 163 48.268444
364 369 249.660583 364 369 26 168 104 40.030807
486 505 462.969025 486 209 30 125 21 59.519337
128 294 563.027466 83 339 79 18 61 89.223091
91 235 462.422516 91 507 135 32 246 80.131927
278 298 371.121033 278 85 116 215 18 83.654922
278 281 370.781250 278 281 187 246 155 70.593811
116 403 315.525299 116 403 19 232 70 61.499958
116 483 205.531601 116 483 8 4 109 45.922512
34 510 178.570938 50 390 16 79 107 35.194290
116 24 326.791901 116 24 19 139 139 52.689445
85 252 375.788818 355 483 76 75 178 68.708130
116 483 278.610626 352 104 76 50 131 43.826508
295 268 467.228516 295 268 123 34 87 67.530441
295 317 447.181793 295 332 249 246 129 55.473324
116 165 200.541504 116 165 215 158 32 47.717434
116 252 316.814423 116 94 28 148 64 65.200554
116 415 348.540100 355 104 101 9 135 45.311558
116 375 279.775757 116 449 28 109 150 38.264709
116 354 656.132324 116 354 42 109 222 74.701042
364 268 732.560608 364 258 226 126 27 89.469742
473 510 356.483490 473 276 223 34 187 60.352104
413 245 396.628357 413 245 32 235 196 56.636497
345 268 226.403244 345 5 22 67 205 53.860245
116 375 359.820496 116 355 42 87 23 58.542732
91 235 254.792130 91 503 9 64 217 50.731503
83 503 286.127045 83 375 17 64 83 74.005974
83 104 330.282684 83 104 32 255 84 89.973801
83 155 309.258667 83 155 32 120 182 71.618187
128 155 538.701904 91 104 176 104 230 61.500771
128 358 485.161194 91 233 68 173 211 84.157997
83 155 301.341064 83 326 181 131 209 93.238487
83 155 344.422577 83 68 131 29 14 74.513542
83 155 482.332153 83 222 176 40 99 89.638512
83 451 673.709961 83 451 75 190 21 117.091827
116 85 509.289520 116 104 115 17 218 64.535057
128 303 339.579102 128 303 144 88 43 68.829849
24 267 202.841324 24 267 43 101 95 52.403034
102 505 285.753448 210 438 43 88 60 46.785706
210 303 408.944672 239 268 43 44 50 108.107887
400 268 354.112030 113 141 55 134 113 59.037476
128 222 436.973267 38 398 89 133 187 60.792362
128 466 655.655029 116 511 125 139 112 79.009445
116 60 418.365082 116 449 28 190 215 47.472301
486 422 219.897278 486 209 40 40 171 37.053413
364 511 716.318665 364 369 183 192 222 71.990395
34 201 405.397705 34 201 21 6 63 73.640686
38 375 323.127502 38 375 49 56 137 66.352936
400 104 271.501556 400 503 26 139 49 53.748955
355 503 329.027740 278 451 178 128 40 47.991707
458 261 359.151245 458 261 73 177 183 80.560181
458 104 280.858490 458 252 26 27 233 48.882004
477 505 416.685486 477 505 75 34 28 88.906448
249 56 264.906647 249 178 21 31 213 50.772308
314 14 321.333557 17 489 36 85 74 50.417957
343 390 220.627991 343 390 3 118 161 41.148731
17 489 205.965881 492 459 1 1 9 38.753826
249 474 179.514359 249 474 7 46 164 37.590096
477 122 200.133545 477 435 47 10 69 43.064911
413 452 315.295380 413 223 214 33 249 64.968628
477 435 271.528351 477 96 80 65 155 66.082161
413 229 328.927795 413 229 4 204 125 62.380939
364 6 317.269287 364 6 75 80 64 48.375294
116 292 183.840973 116 292 5 80 105 32.648464
215 375 257.279938 215 375 22 112 29 58.849804
458 201 430.417084 476 85 19 103 200 92.477615
173 104 411.893402 173 104 98 131 66 93.893936
476 85 407.256561 280 442 147 66 150 86.526276
458 161 331.009399 458 161 12 119 225 65.182938
280 452 355.830597 280 285 214 86 57 36.428162
476 355 331.457092 76 511 32 189 83 84.757095
446 369 245.416245 446 85 16 29 111 56.818523
398 450 357.452423 210 375 121 157 151 59.358055
116 210 275.652466 352 343 5 29 23 57.922241
477 1 219.277985 477 1 1 9 222 60.149990
413 292 197.068878 413 292 61 89 13 46.052128
355 171 178.425049 355 40 39 70 213 44.921139
317 450 459.338287 210 222 197 57 180 92.850662
290 104 400.471954 290 369 32 29 200 75.196152
328 511 354.466248 173 511 31 121 159 55.348019
34 161 149.128189 34 161 19 48 44 39.397743
295 355 111.497581 295 355 22 55 120 38.269020
187 124 305.418335 187 124 240 50 57 59.326138
480 268 495.334198 250 268 76 113 136 113.259544
446 375 410.639526 446 375 115 42 185 61.303070
317 268 344.763367 102 129 197 42 71 46.896461
352 129 275.469238 391 129 31 48 95 69.472107
116 80 220.261642 116 80 33 31 6 44.242306
34 122 367.009399 34 165 55 123 199 67.321533
477 438 416.637970 249 435 46 43 10 94.141190
116 375 548.759277 116 104 65 190 151 80.799522
510 317 204.476974 510 387 158 37 142 60.478264
210 65 341.098175 210 268 48 94 78 61.179115
102 450 298.793304 102 450 43 94 17 64.975433
400 5 334.169922 113 505 85 189 236 53.691730
116 222 738.372803 116 129 188 55 94 104.107033
116 222 419.514404 116 155 238 104 220 55.762062
26 202 406.585938 26 202 237 35 129 53.904408
458 204 128.753998 458 317 88 104 218 35.607815
458 300 397.834015 458 300 73 69 188 99.004509
116 483 165.944916 116 483 73 4 147 48.943577
66 503 285.880066 66 104 55 201 247 55.760136
66 503 239.668625 66 503 16 190 190 54.750561
253 375 377.878418 355 31 213 58 64 65.672676
249 267 347.430267 38 41 24 109 33 67.730865
413 474 187.772156 413 474 7 30 120 45.245041
81 474 105.833611 81 474 6 12 46 30.746305
62 438 634.410339 24 93 203 138 45 105.738556
480 80 221.727081 480 65 254 133 232 52.860352
280 70 406.042755 280 480 236 40 114 103.066734
476 303 378.378021 476 222 85 56 76 103.459900
476 222 589.816772 480 390 44 34 102 142.859451
116 401 437.596039 38 265 43 153 103 74.698364
128 100 547.387207 116 222 125 112 224 80.198204
128 100 540.374268 91 483 115 95 138 60.008087
473 317 433.817993 473 201 82 87 27 79.575188
364 317 327.790283 364 317 148 89 172 40.594410
116 237 141.128784 116 234 5 48 64 44.918022
128 222 624.669312 91 85 183 131 59 84.358032
38 438 912.182129 38 489 174 19 132 93.646988
427 299 276.575836 427 322 82 108 165 53.868980
240 136 586.982239 345 469 172 188 52 107.731949
128 504 490.951965 128 503 44 42 110 54.462223
128 503 371.442444 128 503 17 214 232 100.515320
128 172 259.476593 128 172 59 109 140 56.330376
116 476 238.287537 116 115 222 147 120 29.696314
458 293 268.202637 458 293 165 19 36 45.545952
116 403 188.471359 116 403 41 69 42 48.928902
355 449 360.743958 355 449 121 40 91 98.128426
116 222 257.029205 116 449 236 23 39 50.675468
128 489 513.630005 128 303 138 241 193 80.809555
364 5 377.508270 364 401 42 161 45 53.022915
364 317 914.086548 364 317 226 235 19 108.947678
295 317 447.203766 295 317 189 144 180 63.640289
116 129 148.460159 116 129 93 133 249 34.998146
400 129 186.134216 400 3 7 173 80 44.547924
128 390 321.262543 128 390 12 254 33 82.184456
116 129 443.368622 219 354 44 198 87 85.253242
91 131 605.108398 91 131 226 197 157 58.245140
116 511 632.763062 116 511 249 65 8 112.260399
116 104 418.054169 128 65 101 197 94 76.388664
116 454 105.895699 116 176 11 55 37 38.470875
115 408 247.649033 115 408 108 10 8 70.400421
351 162 317.694824 351 142 16 190 143 69.294853
66 505 268.419250 351 210 31 143 1 55.719810
225 366 280.638733 225 366 12 42 189 54.346577
400 469 230.752625 91 407 16 13 82 57.546310
249 510 317.570129 34 457 30 63 68 36.628098
34 220 293.463379 24 435 56 37 64 63.375610
249 423 531.495361 249 432 18 31 222 106.147583
477 473 365.092407 477 504 55 123 100 49.430531
413 452 429.595123 413 452 131 4 73 67.208954
306 510 282.475159 451 510 44 147 153 44.945728
24 510 302.502167 239 510 192 86 39 60.350037
24 510 285.086609 197 510 42 52 27 58.218102
295 510 302.810181 34 106 64 99 43 44.278290
34 237 225.219269 24 237 10 10 67 42.896881
24 510 173.464340 24 459 95 50 184 43.284321
239 229 195.669968 34 106 39 187 230 29.308136
239 229 318.830353 510 220 44 124 179 48.792130
197 510 270.003174 197 80 114 6 86 56.185501
341 344 95.280449 341 344 27 0 217 32.353088
306 229 67.713898 306 229 4 113 22 29.126867
17 178 246.350952 17 151 3 106 8 60.363178
249 510 105.178230 249 474 62 136 187 35.794575
306 366 286.847565 197 496 86 184 208 70.007103
34 122 166.299530 306 510 7 18 159 43.630909
424 510 252.526154 306 459 93 156 122 59.838161
477 130 163.441574 17 488 10 88 103 42.286533
197 510 139.753571 34 106 39 12 133 42.964310
295 510 256.495667 239 349 132 142 47 57.678341
239 474 224.555679 239 474 5 1 14 43.685364
239 229 271.410797 239 229 30 102 64 46.301579
295 459 284.224609 295 459 24 50 221 65.413582
34 401 309.700073 510 317 18 79 121 46.298603
34 401 239.376556 34 401 52 126 3 39.069851
295 510 271.212341 295 107 246 20 210 55.094440
24 510 321.377014 239 229 44 62 103 53.543266
40 424 232.749374 10 20 218 118 6 52.439484
274 49 210.841644 3 121 16 25 70 41.589062
452 49 257.392487 452 373 31 72 79 49.291958
450 422 253.387482 343 121 6 72 235 55.112873
60 473 385.101349 134 473 184 87 221 59.670078
85 294 350.781433 85 89 79 13 186 74.191551
116 426 134.706512 116 426 70 104 218 37.934162
116 47 225.221252 116 294 24 92 100 49.969227
116 110 293.065399 116 110 162 52 89 69.905464
34 31 301.651917 34 31 119 81 227 56.248604
34 237 131.595184 34 237 24 113 248 35.329155
34 229 316.002289 197 510 136 77 83 73.360931
228 459 125.162735 228 358 4 31 67 23.938215
116 110 470.347565 355 110 42 31 70 39.100376
364 459 490.243195 364 459 18 31 66 53.453598
364 178 373.848145 364 452 174 25 72 61.121555
269 435 447.776489 50 504 28 10 99 39.817158
427 366 541.725281 427 344 89 61 174 56.193695
473 122 859.577576 473 122 184 166 252 135.805420
2 422 278.150208 2 422 12 9 36 54.376938
136 348 144.916458 136 348 23 86 174 33.707176
355 463 185.485046 355 101 127 54 10 44.212822
364 511 281.615509 364 375 58 4 130 59.164719
473 161 886.317627 364 268 18 4 98 105.417183
477 138 345.765045 277 258 136 156 24 72.801147
398 236 260.502777 398 236 26 19 144 39.900196
416 317 126.244720 416 317 43 103 182 42.013260
416 392 147.808685 416 392 3 46 134 41.930595
116 220 143.061783 116 220 19 188 132 26.328228
34 375 82.338959 34 375 70 53 0 37.295238
34 504 187.393402 34 285 21 151 186 36.402977
116 460 239.071289 116 215 30 157 16 51.290287
116 233 170.324203 116 233 31 143 203 38.585381
34 209 171.279373 24 209 11 5 226 40.213821
136 369 218.329224 187 42 207 189 103 46.617859
122 103 254.479111 122 103 152 6 79 56.476467
212 469 282.335571 212 469 33 165 216 57.663628
212 392 189.415421 452 501 1 52 187 38.326210
364 504 754.477661 364 5 147 30 110 74.410385
335 483 912.996216 335 483 89 19 69 118.785553
473 510 772.357788 295 185 54 1 221 71.170738
343 56 292.583710 343 432 21 106 164 65.976112
510 370 137.222534 3 291 12 29 95 40.654976
228 260 70.312370 228 121 222 43 91 29.603203
187 170 254.307434 187 170 73 96 69 56.627682
115 369 134.436691 115 504 207 58 69 48.418140
458 132 190.871674 458 390 88 136 170 53.535526
34 469 154.514206 34 469 29 10 192 45.241055
136 95 168.091675 136 345 9 58 175 44.379402
263 171 141.811539 263 506 85 106 52 39.782639
384 443 385.232544 187 100 236 96 245 71.740036
187 507 354.195312 384 439 130 181 200 65.093300
187 95 196.733307 187 95 32 141 25 50.618050
311 10 217.283142 311 10 49 50 249 55.878498
80 130 322.411957 458 130 58 156 147 85.639458
458 345 274.738708 458 267 13 236 232 62.690529
458 124 330.557678 458 124 214 119 61 41.802395
476 452 302.850952 476 452 13 23 76 46.785561
452 482 210.635544 452 426 82 163 39 28.843355
428 396 131.180511 386 215 108 1 241 38.518494
428 292 221.376999 428 401 186 60 140 45.386646
452 80 170.324005 452 347 22 5 184 47.623623
298 402 167.933960 298 375 49 76 159 29.401974
80 293 257.356689 80 179 29 136 186 56.555515
306 505 211.987793 24 110 29 23 135 42.545086
306 154 86.185387 306 154 39 53 218 39.854336
116 293 261.635345 116 293 24 212 242 70.371880
452 403 362.684906 215 222 197 45 220 76.711929
505 486 187.762665 505 333 168 152 128 52.741661
358 103 173.928406 358 103 1 149 76 39.309746
457 185 183.094101 457 268 222 78 252 50.097252
458 474 55.906811 458 474 22 195 58 34.541042
505 261 223.067413 505 369 87 29 193 53.251656
355 139 105.161552 355 139 85 17 126 38.005016
355 382 210.919617 355 418 210 10 211 48.080307
34 373 195.464081 34 373 12 162 87 35.522427
249 122 205.013519 249 122 33 10 122 45.178127
34 505 139.932373 34 505 5 19 216 46.164749
347 511 171.206314 347 395 108 45 102 51.481941
169 499 296.998322 169 499 73 103 207 43.311363
382 80 455.795166 110 181 72 124 131 83.867050
432 476 279.402710 31 483 72 149 40 47.320160
181 210 268.182800 181 210 26 171 164 51.310219
335 161 264.455261 335 161 3 143 170 58.115463
427 505 465.039520 427 236 184 75 112 51.969944
335 483 444.713074 337 293 42 113 103 50.888092
477 114 150.507645 291 56 22 102 100 39.708279
24 510 160.296661 34 229 29 25 250 40.276810
239 178 219.766052 239 258 24 69 76 51.027069
34 220 111.885643 197 65 21 0 151 36.101089
477 318 197.902298 249 113 8 3 137 42.324684
306 344 209.188263 306 344 227 184 106 47.846699
451 477 130.010559 73 1 70 67 168 45.417728
451 83 141.837692 40 350 70 82 14 38.785831
40 344 89.611969 306 352 23 3 248 39.598961
40 276 104.839653 306 292 4 66 43 39.470329
451 477 137.994370 451 477 116 52 139 45.908642
343 56 158.434158 343 56 3 56 102 39.923409
325 452 156.294342 325 310 7 64 114 51.663906
503 80 261.723755 503 80 180 115 16 47.586914
341 318 175.937851 341 464 33 59 24 50.009232
500 510 138.394257 500 510 25 0 207 39.129143
341 510 160.701630 341 318 13 120 6 28.127811
477 55 187.424866 500 452 3 176 25 45.115891
503 510 105.773430 151 435 3 26 161 37.024471
497 1 311.389099 497 1 28 136 204 52.565876
306 167 111.102318 182 396 11 25 229 40.744659
451 463 179.820358 151 252 10 120 1 42.954628
404 318 216.085632 404 167 139 196 244 60.945152
451 178 183.788300 451 80 1 30 117 43.522217
451 510 229.898468 451 464 1 164 204 66.116646
306 510 234.074356 451 96 35 121 53 48.995724
451 229 91.198883 451 80 15 166 113 38.236790
122 83 222.257812 451 112 163 187 62 46.841564
451 10 100.323326 451 151 93 104 149 33.818146
451 258 79.894531 503 276 26 0 48 37.180080
151 178 143.098099 151 178 2 133 118 36.757164
451 510 154.645416 451 341 12 114 217 46.163242
277 510 170.168243 277 318 24 119 157 37.946152
451 510 126.928009 451 510 2 87 160 42.037525
306 229 262.354645 306 229 2 83 107 59.231461
277 178 174.115936 306 56 2 12 202 38.154636
116 362 684.151733 116 322 51 5 3 52.473171
390 423 495.792236 135 401 72 172 246 44.981323
181 349 373.801361 81 349 28 51 162 76.599167
181 349 492.105774 181 349 79 251 129 98.094254
505 294 551.545105 25 70 67 197 48 58.027592
25 510 729.834473 25 317 247 64 183 82.165108
116 358 692.161499 225 186 135 29 51 92.427605
174 510 501.832153 174 510 30 142 170 57.965012
477 459 312.487549 477 6 44 9 177 45.194134
17 178 492.408386 17 80 147 102 129 54.493042
477 484 417.589783 477 355 207 3 157 68.688004
405 510 225.242844 116 358 17 2 48 38.168083
25 104 245.306503 25 392 208 131 121 54.256142
25 489 207.170990 25 261 1 97 89 47.347599
451 318 271.035034 343 504 131 151 196 44.185440
15 351 237.048065 15 351 4 137 233 58.562599
494 463 303.815491 494 415 79 136 153 47.073639
187 362 200.678619 136 124 61 5 99 49.861576
487 403 321.330139 487 403 5 50 152 71.223488
36 142 283.900116 36 385 46 56 52 52.451710
386 390 141.796219 386 390 12 36 178 50.806728
23 510 201.633392 23 229 2 123 94 44.191471
181 148 185.846146 181 148 23 125 34 35.283165
181 14 217.460892 181 14 16 176 192 47.812855
181 29 221.265076 181 29 4 99 39 47.485039
116 408 183.039963 116 47 37 73 218 54.402695
116 155 478.923615 116 217 194 56 61 85.159378
505 223 402.448853 505 223 76 85 198 75.540733
187 33 150.296875 187 33 27 168 226 46.270065
36 258 296.921844 36 142 16 4 51 62.955082
322 366 200.617355 16 121 219 191 25 44.646561
322 258 206.660065 322 109 31 7 235 46.627468
305 375 508.786896 485 447 76 19 98 97.308678
358 474 373.700836 66 410 35 14 33 38.067059
44 489 120.595146 44 344 33 17 151 39.746838
413 360 244.931854 413 360 39 34 185 47.994488
413 463 134.748505 413 463 13 174 0 38.802818
413 370 171.074371 413 370 207 10 72 39.566296
497 510 150.259842 497 510 93 7 149 41.277458
279 186 143.904144 298 191 13 22 25 42.903351
95 369 237.739105 95 369 48 136 163 41.535526
322 355 300.654053 322 380 108 104 145 66.276154
322 355 344.531067 131 444 108 59 121 82.152382
322 254 278.428223 322 254 1 22 66 53.237400
369 93 286.598297 322 254 83 191 112 67.074135
142 114 247.586655 369 428 43 129 100 53.549530
369 40 310.235413 47 276 219 185 90 49.998631
311 31 336.237762 311 31 212 71 175 69.226746
355 229 149.012100 355 106 7 1 222 25.209621
236 366 236.978027 236 452 31 95 156 53.773544
500 510 275.107849 500 341 10 23 37 59.301289
451 510 133.322861 451 505 5 8 126 53.244144
210 500 240.938721 210 500 5 133 250 51.069290
243 84 342.993988 383 5 21 68 28 57.677525
487 489 85.820091 487 500 26 65 0 35.646255
305 483 262.851135 305 483 108 134 250 50.869026
136 142 164.857422 263 142 25 66 112 34.716354
24 510 95.574493 24 453 30 197 178 48.471451
225 474 205.829208 225 345 156 49 123 50.765423
203 346 186.907852 203 346 4 179 145 47.190914
44 186 139.183731 44 366 16 62 161 43.186489
135 106 351.763580 111 463 213 92 56 40.005596
17 151 138.080414 17 380 27 3 74 49.361965
477 122 147.092392 107 104 6 3 200 42.038780
181 146 132.370422 181 199 16 95 108 33.904343
314 505 234.597183 343 222 28 144 154 41.575668
239 435 349.823334 239 435 29 3 69 52.346870
122 339 199.022171 122 339 19 59 126 59.360821
451 463 164.469086 451 463 49 58 81 51.072025
424 510 55.491795 424 510 151 190 202 35.771576
18 423 243.451141 18 423 6 110 9 50.290825
38 167 295.033997 38 463 78 136 214 58.085033
335 423 520.922363 335 510 159 58 107 68.802879
181 358 504.754089 181 6 159 114 144 54.131260
50 130 175.704498 355 151 61 79 15 49.049416
104 158 203.846451 104 93 116 18 9 49.249969
398 463 205.239166 398 223 178 59 9 38.087627
355 362 98.916428 355 474 11 24 0 37.671741
135 332 66.370979 135 367 12 7 52 37.833889
135 387 118.517174 135 356 22 7 0 39.853096
135 387 125.771347 135 356 22 7 153 42.619301
25 222 165.166321 25 222 22 160 247 27.387846
355 1 184.726517 355 510 25 39 220 38.161770
34 459 197.411316 34 107 21 71 170 43.011993
225 463 120.349060 225 186 2 67 31 31.705887
135 1 204.771713 470 341 2 3 91 41.290913
470 510 283.379242 470 459 6 31 249 55.194794
122 96 156.105835 122 96 73 100 252 59.575558
233 55 180.602356 233 55 13 21 113 39.658134
306 510 173.074722 182 229 39 146 53 47.962845
355 6 159.205490 355 6 25 10 135 33.034687
447 260 179.605347 447 260 108 171 97 41.499683
36 185 268.006561 305 483 108 31 242 53.338848
102 349 208.686218 102 234 17 68 51 54.455643
503 276 118.092468 451 35 12 146 134 18.822672
427 510 410.514526 427 453 31 7 6 65.030418
181 435 176.021423 181 235 10 4 159 43.988319
235 178 392.981140 235 178 72 56 63 94.454323
25 320 294.844940 384 223 96 191 57 54.147762
211 25 178.583527 398 474 178 153 161 28.918533
26 504 329.554016 26 222 30 199 233 56.353901
16 401 148.615356 16 268 4 159 70 37.890408
256 510 284.327362 322 122 157 90 155 64.727722
142 1 171.578308 142 1 178 197 1 31.505001
108 366 145.912064 108 186 142 34 182 23.914379
413 463 242.441360 413 463 196 68 214 57.059772
413 415 324.165222 413 86 4 67 99 74.651840
500 264 210.186859 500 264 25 103 129 34.066059
500 341 136.169144 500 453 6 126 7 55.545486
279 20 140.067017 279 217 66 142 89 36.664814
298 380 150.722473 298 291 207 29 216 38.430363
298 380 210.105881 298 20 16 191 221 55.923077
298 42 200.304108 298 102 6 22 96 56.449223
322 49 222.051620 322 49 1 30 207 54.632835
310 245 172.774719 310 302 43 7 221 41.503979
118 311 105.027122 118 311 25 61 129 34.877869
238 466 152.753479 238 466 20 26 107 39.862431
477 396 112.676788 500 358 1 4 177 34.223362
451 318 130.093658 451 318 4 92 31 34.734512
497 236 160.453674 497 236 16 163 219 47.608360
84 244 171.600922 84 244 12 120 127 32.085804
432 423 351.589508 432 423 49 54 28 78.693588
432 349 184.292358 432 349 1 176 101 25.820454
487 470 182.129898 442 403 85 56 129 48.600792
263 403 237.490799 136 403 7 153 48 58.775555
24 459 258.359344 24 55 12 48 202 55.145733
17 435 196.601913 17 324 19 2 93 41.999153
224 55 184.149841 224 276 70 34 250 40.877472
396 56 345.444427 44 503 21 3 229 59.439495
135 61 333.883362 360 459 146 120 219 71.480972
454 56 204.318787 4 186 48 105 13 50.314613
477 438 543.381897 249 510 184 80 117 77.649925
228 510 119.923836 228 108 5 31 16 35.449486
135 343 266.267212 91 504 18 161 97 48.281639
135 344 303.355774 135 276 101 58 233 62.393452
477 100 241.136383 477 6 47 31 8 57.821014
413 0 459.209534 413 0 65 145 250 86.222260
269 96 165.586365 269 301 31 59 125 40.524532
136 248 255.085312 136 271 48 30 74 48.235668
398 415 135.555695 398 27 10 34 34 30.310791
25 510 199.515228 25 199 13 11 47 50.718784
25 317 237.850128 25 104 30 56 37 40.138199
91 459 309.746063 91 55 56 53 187 53.936016
135 317 347.146667 135 317 30 151 8 45.190548
306 510 234.711731 306 510 93 48 120 69.607864
451 510 206.551971 451 55 61 31 164 50.411823
181 223 512.515381 181 366 213 95 32 57.166813
358 484 152.552887 358 484 39 23 76 47.839981
358 214 132.143478 358 214 9 46 207 47.494217
181 510 465.491150 181 318 75 137 202 57.750237
181 423 233.731827 181 423 17 43 49 69.987366
341 510 622.068848 503 178 106 19 99 73.849640
335 510 339.415100 335 510 55 62 154 49.768135
405 463 192.888367 405 459 61 23 58 45.858185
240 459 695.191101 235 438 56 79 185 99.803963
384 214 311.657623 384 214 178 160 145 64.044899
384 174 227.347610 384 174 70 143 251 45.612633
384 174 238.960083 505 206 148 181 252 75.123138
384 82 510.924438 187 499 161 79 159 64.020187
503 304 197.021408 503 1 25 53 135 55.250713
413 217 233.938995 413 217 2 29 216 51.877621
413 213 96.114021 413 213 17 18 159 31.344210
413 462 165.319916 413 462 17 65 6 40.884861
466 505 167.431519 466 505 4 84 9 45.651382
236 229 104.926125 236 80 3 74 40 39.537395
500 229 138.357025 189 452 59 1 126 42.697960
432 510 252.686172 62 463 31 15 125 44.145508
451 341 355.465027 451 341 30 100 93 52.565460
122 460 221.905365 122 460 31 87 53 44.911129
458 453 153.881531 458 453 116 25 59 43.062309
130 254 206.559570 130 408 43 28 137 44.056210
95 254 308.818054 95 254 48 168 60 59.374184
136 95 161.201218 136 95 16 52 211 53.970467
187 459 123.026527 384 265 32 84 9 37.506149
384 186 223.879272 384 54 37 31 198 56.591087
187 490 229.478928 187 101 218 4 126 37.456051
181 401 349.448975 181 401 216 29 135 64.668152
181 358 273.123352 181 220 167 65 150 66.426048
361 344 253.341125 361 344 12 142 16 51.958237
113 422 93.130569 174 332 4 6 48 31.416088
225 35 132.998840 225 35 13 22 21 36.676895
225 276 171.903030 225 276 1 29 169 43.313324
225 510 294.372284 225 474 89 41 83 62.536755
400 80 231.496368 400 80 18 40 235 48.586185
474 349 188.918945 75 349 1 118 153 39.042450
181 318 131.906769 413 474 73 28 15 44.869389
236 438 473.186249 45 122 123 160 44 50.004753
510 349 236.842239 510 122 2 68 107 43.975262
205 344 114.739929 205 344 43 150 70 39.913204
48 155 345.190918 478 438 205 67 113 62.517773
384 223 248.259460 384 245 28 32 250 50.882580
116 155 387.284637 116 204 75 145 200 64.541107
116 294 427.677368 116 395 67 9 199 52.504475
135 138 132.947845 135 209 9 88 46 37.630402
236 459 267.048279 413 155 4 52 41 47.133183
500 349 211.019806 507 349 28 135 77 40.901058
413 278 273.288147 413 278 86 80 223 65.949135
277 106 159.415527 277 106 66 123 132 41.011967
36 459 232.579834 36 459 242 196 41 45.363228
476 463 298.025055 476 463 43 197 183 60.380768
487 345 406.403229 487 345 76 92 16 78.971748
487 345 338.547241 487 345 108 189 181 64.347786
487 21 235.014557 328 410 178 31 16 67.529991
487 382 217.757553 169 172 27 18 208 50.420834
424 510 192.908173 424 510 43 8 170 38.027184
413 76 119.607025 413 324 13 32 109 28.878138
413 474 113.382187 202 122 3 142 0 36.987282
189 5 157.557480 189 208 43 48 184 29.247208
131 447 173.097031 131 69 32 136 37 53.014217
95 246 229.084824 95 246 207 51 104 54.155979
95 369 196.692261 95 187 16 88 224 47.337799
322 392 202.355911 95 369 16 96 45 59.482502
322 392 215.786926 375 392 6 59 10 59.614872
287 505 285.001251 322 184 142 23 134 63.068497
310 379 241.319885 310 379 12 56 10 60.890381
299 254 272.357483 299 254 158 14 168 47.230347
299 438 254.219360 311 31 31 37 188 51.851116
292 80 301.078003 292 80 231 182 160 68.440056
505 24 236.948059 220 80 76 72 192 38.629623
249 510 220.003036 306 510 35 139 152 40.035740
413 195 386.259918 413 195 181 120 55 88.406738
413 167 143.531479 413 167 32 52 0 37.601643
451 510 148.995895 451 229 152 49 169 38.813805
497 258 209.573517 497 258 32 149 6 63.438549
500 318 204.087952 202 10 2 103 149 60.753212
503 100 193.100647 356 510 25 16 232 48.029102
500 510 279.027222 500 459 185 120 206 54.449142
189 438 238.857178 500 130 25 66 6 56.599152
45 505 232.182709 45 505 94 58 52 57.171989
306 510 122.983932 306 510 12 74 238 33.741184
451 318 134.223984 451 318 47 5 127 42.680882
497 178 336.373352 497 422 31 59 1 66.620239
451 510 218.300308 451 510 236 188 83 60.228302
451 510 183.746735 343 356 75 40 141 41.932858
477 510 93.808960 477 341 16 184 222 37.858929
197 510 176.887405 3 418 39 3 0 30.586830
298 20 429.536285 298 20 209 151 79 97.417480
95 116 260.177582 95 279 108 3 179 49.229603
95 380 192.234482 95 501 207 111 109 48.738350
95 246 338.340454 95 246 66 111 132 54.238396
180 254 473.409851 131 103 43 120 123 91.212433
310 452 343.732971 369 288 43 4 10 57.578812
134 505 314.160400 56 178 31 86 2 70.662766
17 438 435.916412 17 304 75 80 82 51.480347
25 407 426.503296 25 407 106 116 109 55.511860
413 55 291.924103 413 55 181 9 172 62.307789
477 438 296.813782 202 438 181 114 115 49.763252
477 459 205.000900 325 459 1 120 26 33.460091
341 435 149.332626 341 435 25 1 64 21.124840
341 510 141.037979 341 459 1 104 146 32.401020
487 397 182.558411 487 267 100 172 252 37.249813
476 463 285.786987 476 148 35 151 235 72.888748
476 452 292.453430 476 276 79 74 54 65.315056
476 510 248.763397 476 318 13 55 52 48.904854
130 503 165.584030 130 503 210 19 76 60.219379
347 460 157.648956 347 460 100 191 121 44.394005
189 229 232.423721 257 229 251 196 122 25.512299
474 88 213.300018 474 88 1 75 206 48.629063
477 206 134.187378 341 85 181 3 178 43.629158
413 435 280.512634 413 278 32 170 157 58.391613
325 101 276.483673 325 101 12 208 1 53.380310
447 178 123.963531 447 178 43 28 131 43.638767
431 453 142.840378 205 489 23 4 127 41.398560
478 463 274.218781 478 463 181 82 132 56.962532
470 55 172.249359 470 446 231 45 22 52.106026
486 474 176.240479 486 510 32 12 211 38.662292
486 317 335.418945 83 504 225 21 124 54.727039
486 375 505.447815 390 201 150 223 46 83.500984
486 435 540.229492 486 317 92 31 12 54.894924
341 510 208.127151 0 318 2 142 36 49.495583
451 510 263.561981 451 258 100 66 15 52.250118
181 300 421.680695 181 300 74 89 159 113.477165
181 165 216.551910 181 332 42 8 16 54.481659
181 300 192.617065 181 300 19 43 250 45.615143
348 133 373.258209 348 133 48 46 14 69.103027
116 27 206.592117 116 81 2 20 40 39.493992
225 452 120.006111 225 452 3 29 53 27.970388
400 463 276.967957 400 464 79 141 136 48.593369
400 341 218.379883 400 35 127 39 4 52.016773
225 438 404.237122 225 459 115 163 4 56.253792
135 1 405.015503 113 473 99 190 237 58.592857
113 435 434.657928 113 278 55 21 134 65.722160
113 438 347.156799 113 392 79 152 236 78.168472
225 463 278.732666 225 463 25 6 0 30.724430
340 438 193.707962 340 76 197 36 58 48.178085
306 93 340.899048 306 93 153 5 235 67.238724
110 279 251.889984 187 172 62 100 189 46.657055
85 510 59.866001 85 510 0 84 241 44.340820
458 351 97.817238 458 351 253 142 142 30.566395
110 184 181.069199 110 435 160 124 21 48.666439
110 176 320.572998 110 176 240 200 71 57.292118
95 107 310.958130 95 184 177 88 144 69.339966
131 185 308.873779 96 369 94 177 145 41.298729
369 185 348.011505 322 499 115 168 191 68.668213
369 178 281.969421 369 489 195 87 50 57.549625
136 92 135.775528 136 92 6 90 171 42.122524
136 298 206.534973 136 499 160 137 129 58.850628
352 501 312.677948 352 501 160 13 71 60.542664
503 446 294.182739 503 446 55 160 182 60.675301
500 55 168.473145 500 55 29 49 139 41.489048
500 276 226.239090 500 276 196 22 43 42.164204
239 459 241.595734 239 459 47 70 47 41.075836
10 463 391.432587 286 55 28 73 117 39.149994
497 233 200.425446 451 178 52 154 64 42.824078
451 80 109.657944 451 186 61 20 9 35.731106
306 510 156.518982 0 459 26 23 253 47.164101
306 510 418.096954 239 174 21 20 63 74.733208
451 510 191.149780 451 510 151 87 163 57.180038
306 362 145.582199 306 362 70 166 91 30.241018
138 464 250.629059 381 116 23 63 138 45.659611
306 510 165.386734 306 106 13 3 106 47.897518
306 510 233.928391 306 510 151 80 2 60.039459
282 463 127.992874 282 452 2 37 70 41.276814
275 186 174.812119 341 353 10 168 212 41.132725
343 281 287.816620 381 122 36 87 61 40.479179
451 510 90.996841 151 186 10 51 63 42.360779
138 258 137.673279 138 103 1 76 196 38.969421
197 6 77.801720 197 123 23 58 161 41.608437
239 167 131.017654 197 200 37 199 180 38.170650
27 510 170.045105 27 130 1 49 77 52.684536
197 322 176.754425 197 322 27 46 49 38.204720
34 424 108.541817 182 241 13 102 9 55.562069
197 162 276.953003 197 510 218 168 128 50.414467
510 349 144.878082 102 499 8 39 196 35.346596
228 204 137.251190 228 458 29 0 118 33.821251
34 501 58.092793 24 278 8 0 0 37.743721
45 392 167.948837 45 380 134 168 55 52.239109
306 341 192.619186 306 206 2 56 90 48.639194
45 132 212.300842 45 491 37 73 89 66.148674
45 374 286.757660 309 148 49 68 162 63.475201
111 424 385.448730 404 200 6 110 185 103.816856
111 80 168.892654 111 318 12 8 33 42.197750
197 359 114.908813 197 359 6 43 113 50.153664
197 510 191.589127 197 463 29 22 243 45.861729
116 130 82.907135 116 130 0 54 222 48.066967
239 155 168.603973 239 155 19 105 17 56.473190
0 459 134.296143 0 459 16 132 159 48.501213
228 459 254.032776 228 459 1 101 232 69.065559
17 317 470.062164 17 491 44 112 50 63.628769
116 269 193.496841 116 269 41 169 33 68.307816
249 459 323.033997 249 318 80 64 52 66.921638
182 276 124.811249 34 301 37 152 25 42.906609
18 85 128.372757 18 493 32 57 240 24.213387
18 338 272.320740 128 338 16 169 91 40.278496
295 438 246.576004 182 477 158 195 212 50.565571
181 369 289.561279 181 261 152 64 106 63.647217
426 58 220.127777 426 58 254 113 192 50.006142
258 380 211.529160 426 318 52 142 242 38.749760
426 464 268.999023 426 464 31 36 121 38.095428
25 366 264.339600 92 473 31 105 5 53.771381
400 178 216.815643 400 149 62 136 208 51.705242
306 57 143.193420 306 57 178 139 166 45.728470
503 332 109.663467 341 278 12 76 17 40.420135
268 347 65.436928 268 347 0 31 116 42.411285
322 206 243.935471 322 422 25 169 151 69.022629
287 56 181.169296 84 109 19 61 47 43.955280
61 228 200.973312 61 228 48 66 154 49.013412
37 178 135.289520 37 178 1 76 93 47.624981
24 459 190.603653 24 407 13 67 48 49.529583
181 339 232.515091 181 339 48 171 175 61.050842
451 505 113.727638 150 452 32 28 15 28.589842
240 199 201.091232 240 199 1 93 143 32.935192
278 215 225.797989 278 215 147 24 29 51.590019
18 264 417.951691 128 215 55 226 22 73.513702
345 54 345.697845 57 134 71 151 95 69.247643
427 174 343.089050 427 504 75 36 152 63.380527
467 504 663.563232 467 55 147 174 137 98.829140
486 422 485.633118 345 14 71 117 68 74.527794
57 215 432.423035 57 57 207 57 139 99.154137
278 122 180.017685 57 277 41 3 244 32.433537
13 178 186.029037 13 186 31 176 43 42.389854
343 107 224.844421 419 463 18 76 159 45.241089
432 422 210.681549 237 422 21 9 20 49.107685
371 505 127.358376 237 260 23 22 196 42.351772
416 32 193.340607 212 271 31 57 202 30.089798
50 194 195.658310 50 194 52 132 155 52.137253
451 55 61.162624 451 55 0 0 28 36.559788
17 254 138.496078 25 57 52 117 110 38.892185
18 54 204.907608 18 54 6 140 209 32.768429
116 230 343.177368 400 422 43 136 41 78.352654
111 366 138.282242 182 433 193 87 84 40.252659
325 36 386.157623 325 186 173 188 5 83.894073
92 384 215.427002 92 384 85 77 209 27.557423
278 242 96.732170 278 242 85 15 58 56.681141
116 149 227.459457 116 384 108 148 185 38.404800
494 209 176.284225 494 209 19 19 25 40.007633
325 336 198.261475 325 336 181 82 194 46.850403
295 426 123.708786 228 482 8 79 29 41.416824
18 452 72.094452 18 186 22 130 174 41.974163
203 473 283.823730 164 55 163 94 20 67.153984
104 84 180.600555 263 347 20 90 92 55.309265
303 318 101.529114 303 187 4 22 0 29.706388
256 500 109.255722 256 500 25 0 0 32.701653
171 355 60.175259 171 355 43 49 171 33.003490
358 384 174.096176 358 49 25 174 4 41.563152
275 16 154.771240 275 16 152 192 207 41.805458
117 146 85.038330 117 199 20 0 111 40.276718
497 235 56.639965 503 376 32 0 56 33.656239
313 182 126.500473 313 107 62 129 42 23.483536
84 186 201.487610 84 186 212 18 200 47.008873
84 504 251.355057 84 438 48 79 148 65.833679
371 127 169.284286 371 505 61 169 89 43.663277
61 186 219.512299 61 167 32 78 47 28.601624
358 25 271.921844 358 133 233 36 45 58.052410
249 14 189.200439 249 14 29 46 12 57.251431
102 497 748.579956 102 497 230 84 71 157.093109
451 55 122.345032 117 55 93 56 104 24.089685
181 410 323.094269 126 438 113 69 83 68.950279
171 384 177.526978 171 384 1 176 91 44.747227
416 301 132.773682 416 397 19 184 189 41.216106
447 206 226.598663 447 191 55 84 77 50.033745
171 58 192.427643 171 209 49 38 112 49.807461
116 64 231.541153 352 222 90 151 73 52.946362
197 37 145.707336 306 464 2 45 234 35.560463
241 142 142.616745 60 14 49 113 231 33.656925
310 444 218.072739 131 466 168 22 54 32.401009
298 188 132.229477 298 188 235 67 132 42.375595
487 122 133.518387 487 340 70 50 125 43.362068
282 178 149.851944 282 178 160 116 209 49.836018
507 446 122.455620 507 446 10 87 62 51.679726
275 153 70.521645 275 153 5 14 111 35.587902
205 380 121.435379 205 380 108 45 98 43.098103
186 102 247.733307 186 102 6 58 235 57.435329
289 114 199.574860 238 49 48 82 135 45.529427
116 351 244.416183 116 415 73 5 171 44.186909
116 149 200.373993 116 370 48 178 80 43.893288
18 422 404.213440 18 337 101 162 178 78.445145
278 56 292.338074 278 56 31 105 224 61.673462
224 505 379.108765 224 42 31 145 255 76.239861
224 116 301.149841 224 116 31 93 121 57.118301
505 192 324.867706 92 20 31 18 208 74.667389
182 438 318.226532 182 254 68 113 163 65.993126
422 155 311.373535 422 66 67 14 54 66.712952
111 459 176.038239 111 459 1 36 136 35.622562
277 459 117.243019 277 108 2 23 102 41.063038
503 301 200.779922 503 116 17 46 20 39.234734
236 186 168.078186 236 219 78 150 42 35.927475
500 378 214.506165 236 285 212 33 248 48.346756
402 55 159.889709 282 462 37 103 221 50.695171
503 463 125.403084 503 459 19 62 11 30.147730
503 459 255.979889 503 243 4 178 233 47.671467
111 55 219.488968 111 55 166 52 254 38.310135
451 318 56.216061 306 464 29 0 221 31.216866
500 474 122.795021 500 464 29 34 125 32.180756
239 95 166.878876 239 95 6 128 108 40.929031
123 85 202.973480 123 85 16 31 139 47.921673
240 349 221.158569 240 349 47 88 2 38.225994
181 332 175.720337 337 184 1 114 50 30.639971
291 301 241.393250 291 301 4 87 53 67.852348
464 324 155.781799 464 324 14 101 109 33.618599
278 450 166.357620 18 1 6 40 143 38.314289
18 369 441.574646 18 337 147 130 54 93.712959
116 510 193.934479 116 510 73 165 31 53.194412
116 477 208.808731 116 446 47 195 1 47.161297
458 223 140.663223 458 192 32 21 29 44.658001
458 460 263.536652 458 449 28 45 5 52.350639
458 130 290.923920 458 130 232 47 44 52.959328
322 490 261.107117 322 490 1 200 200 63.879559
100 56 366.432251 100 350 70 220 148 91.004845
100 270 226.194244 100 310 32 55 17 44.171200
352 25 249.044586 116 25 7 127 51 61.840015
116 415 229.699036 116 415 2 36 142 60.494816
116 130 176.266129 458 244 59 2 224 53.537498
116 483 158.972778 116 483 1 74 202 30.529396
349 236 91.286049 349 236 0 125 26 38.161312
325 505 341.683411 325 130 28 70 123 62.777061
17 505 239.483887 98 505 146 101 16 39.162643
25 369 122.512657 25 369 1 96 21 38.796104
25 369 163.764191 25 369 19 99 134 48.285969
25 438 308.792206 278 500 79 77 42 64.428627
25 58 285.026184 25 76 55 28 65 46.025360
278 358 335.492981 25 369 245 78 106 56.869629
182 473 195.612961 182 322 85 19 207 55.136578
503 130 129.460617 451 214 177 135 244 42.905884
325 463 185.090027 451 186 47 2 156 47.238289
182 277 113.563980 197 144 1 46 181 39.888599
371 301 107.264122 61 443 5 144 189 44.492435
136 457 214.645233 136 457 37 59 45 53.083302
292 85 186.613556 220 355 32 18 52 45.783859
278 305 94.226990 278 305 3 168 42 36.033970
128 186 365.303589 128 178 136 26 121 37.275131
128 264 307.056030 128 264 33 57 223 58.586548
128 54 445.872864 128 379 136 66 123 101.850296
500 55 92.193504 500 55 33 9 193 36.907444
197 6 114.841484 197 6 6 37 9 36.826595
197 241 286.162231 510 55 70 176 60 71.258270
311 82 123.804825 274 229 17 76 33 39.941692
157 438 246.842331 130 201 25 77 136 66.799889
173 438 267.040802 476 318 100 6 20 89.542931
157 438 235.978394 157 446 219 80 198 45.686020
130 201 326.612335 298 438 58 95 233 63.395443
130 201 332.098602 298 438 58 73 106 61.323986
98 368 190.545578 422 487 5 9 164 53.443497
1 276 71.496788 1 276 16 199 102 40.384224
202 224 72.120613 202 224 22 80 176 33.031097
500 243 97.925011 500 396 37 31 45 32.175804
279 298 252.703003 279 386 47 103 93 62.046902
84 318 192.677170 84 318 108 86 110 43.820744
84 178 111.476463 84 178 23 77 29 37.100552
237 167 197.336853 237 167 2 52 20 48.392750
171 491 119.698792 171 491 49 44 175 39.629692
197 149 135.481888 197 149 10 34 189 49.771797
17 510 270.835815 17 229 29 237 225 72.700523
325 452 104.685921 325 452 1 58 0 37.918144
181 452 150.123230 181 195 16 59 242 42.757824
467 510 270.913177 467 510 137 58 89 73.649704
278 84 125.491364 278 238 17 117 159 34.355080
278 194 179.108627 278 194 59 28 180 44.114697
98 403 177.328369 98 403 23 15 73 41.540424
24 167 114.877258 24 167 17 48 0 37.128677
58 432 235.945450 58 432 24 101 76 47.573830
413 285 235.548737 413 449 2 64 174 51.834591
337 301 267.106445 337 301 213 53 106 60.219315
181 275 210.027527 181 275 3 137 24 49.595627
455 304 227.313553 35 318 21 191 35 62.395283
258 339 168.837967 258 339 61 96 121 47.146626
417 235 155.043076 417 235 59 45 200 54.183502
182 186 185.989441 34 426 49 8 51 31.916677
380 490 176.778687 380 380 15 102 221 61.978725
116 466 284.536926 116 415 73 18 106 62.117664
116 252 527.047791 116 149 176 148 192 96.411972
352 104 423.580261 352 104 125 221 65 81.523163
17 246 201.766296 17 209 103 170 80 40.259724
239 220 156.237747 239 426 21 36 33 36.706348
183 54 482.269928 159 294 255 28 52 57.253925
426 510 200.894104 426 510 207 52 128 51.382942
278 482 106.624710 278 482 29 130 146 53.147854
278 186 138.486176 278 186 16 93 43 35.612274
182 146 205.778946 426 178 26 200 57 47.033573
98 341 47.664017 98 44 13 24 0 33.692657
116 483 292.170380 458 460 125 64 235 61.629974
116 483 433.825714 116 52 176 19 77 70.785355
116 483 434.883820 116 52 73 2 169 84.030334
116 464 237.826965 116 130 151 155 249 48.053772
100 318 70.929260 100 318 0 151 41 36.822994
6 85 260.408081 6 85 165 84 156 60.282345
100 187 203.968079 100 187 78 77 183 66.436752
237 507 201.486893 237 85 88 185 48 56.385132
351 235 279.043793 351 235 207 128 134 53.591526
458 252 305.923615 116 483 217 69 142 54.177254
116 483 269.275024 116 483 73 4 48 44.697224
34 341 180.054092 34 341 47 17 0 37.743149
451 112 287.964478 282 112 32 143 113 74.106689
239 301 103.108185 239 301 20 14 105 33.475262
426 206 363.343201 342 144 218 68 8 77.364647
480 116 205.718613 351 261 28 136 191 43.859833
316 504 266.708710 92 480 143 55 39 61.714291
92 379 240.662445 92 379 29 145 200 56.163765
116 52 338.679260 116 466 135 158 142 64.805786
116 294 376.951813 116 294 212 156 25 75.268204
311 346 184.268555 311 346 1 103 173 42.393711
302 39 130.105835 302 39 31 57 67 44.036514
295 55 161.984085 295 341 17 82 157 36.166924
306 167 238.954300 306 167 3 130 172 57.749004
197 463 226.442093 197 244 6 39 222 41.964920
197 170 190.853546 197 170 196 17 28 57.578754
111 380 182.465881 111 462 6 66 224 51.792667
404 10 75.002991 404 10 32 37 106 40.174133
426 82 196.377106 151 446 48 31 154 54.598576
451 167 132.702347 451 35 6 136 134 39.412369
45 447 67.404915 45 447 0 125 88 37.017025
241 184 166.026657 241 184 48 121 250 45.082191
306 35 94.795967 306 35 27 0 6 43.515770
111 56 251.812485 111 56 212 186 192 61.418972
111 107 101.017334 111 481 15 13 31 45.052895
David Rowe
2017-03-10 20:51:54 UTC
Permalink
Hi Jeroen,

Wow that's great work - thank you so much. So wonderful for me to see
someone diving into the code and making things happen at this level.
Also the results are pretty good - that guy vk5dgr sounded the worst :-)

Here is the same source processed by Codec 2 1300:

http://rowetel.com/downloads/codec2/all.1300.wav

What does everyone else on the list think of the two modes compared to
each other? There is an obvious level difference - an artifact of the
post filtering. When listening, compare one sentence at a time, e.g.
the first 3 seconds of both, then the 2nd 3 seconds etc. Think about
which sample you prefer.

Couple of questions/comments:

1/ How did you train the VQ - using the Octave code?

2/ The squared error can be directly related to spectral distortion, if
the difference (in dB) between the original and quantised spectrum, so
the extra VQ stages have reduced the RMS distortion by sqrt(147/20) =
2.7 dB to sqrt(44/20) = 1.5dB.

3/ Given more bits are available, might be interesting to expand the
bandwidth the VQ works over. The 1300 samples do sound a bit more
"wideband" to me, compared to 1300C.

4/ The post filter may not be so important now, and could be "relaxed".

5/ Yes I'm sure the pitch estimator loses it occasionally. To test
objectively pitch contours could be compared with different filtering.

6/ I have some other training material you could try, to compare
results. But sounds like you are on the right track.

Well done!

Cheers,

David
Post by Jeroen Vreeken
Hi all,
I have been experimenting for a while to extend the 700C codec2 mode
into a 1300C mode. Mostly by just trying out different things like
adding more bits for pitch or energy and trying to create better codebooks.
It is a fun way to get to know the different parts of the source and get
an understanding of the codec.
I finally settled on a frame format that adds a 3th, 4th and 5th stage
to the vector quantization of 8bits each.
The first entry of the three codebooks contain all zeros, this is to
ensure that the additional stages will always have a result that is at
least as good as 700c. Also transcoding from 700c to 1300c is simply a
matter of zero padding.
For the training I used over an hour of material that could be heard
over a typical repeater. (It started with some recorded QSOs, but ended
up with mostly bulletin samples).
A patch for this new mode is attached to this mail. Personally I think
it might be a good candidate to use for the 2400A/B/T FreeDV mode.
At http://video.vreeken.net/~dml/patches/all.1300c.wav you can find an
example of the output.
One of the things that I noticed while testing was that the current
pitch estimation sometimes makes a mistake and ends up with twice the
real pitch. I also got the feeling (but no proof yet) it was worse with
samples with heavy filtering under 300Hz (e.g. to get rid of CTCSS tones).
So there might be some room for improvement there...
Regards,
JEroen
------------------------------------------------------------------------------
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
Jean-Marc Valin
2017-03-11 03:52:21 UTC
Permalink
Post by Jeroen Vreeken
One of the things that I noticed while testing was that the current
pitch estimation sometimes makes a mistake and ends up with twice the
real pitch. I also got the feeling (but no proof yet) it was worse with
samples with heavy filtering under 300Hz (e.g. to get rid of CTCSS tones).
So there might be some room for improvement there...
Pitch doubling/tripling is a common problem in pitch estimation. I spent
some time trying to avoid it as much as possible in the Opus (CELT)
pitch detector. You might want to have a look just in case (haven't
tested) it works better than the existing codec2 pitch detector. Some of
the things I have found that can help (and that the CELT code does):

1) Upsampling the pitch correlation (2x) to get fractional time
resolution. Even if that resolution isn't needed, it helps reduce pitch
doubling).

2) Looking at fractions of the pitch period to find peaks. e.g. if the
correlation at T/3 is almost as good as the one at T, then we pick T/3
instead.

The code is part of Opus, in celt/pitch.c (remove_doubling() function).

Cheers,

Jean-Marc
glen english
2017-03-12 01:15:15 UTC
Permalink
I agree strongly with Jean-Marc here.

there are many signal processing examples where 2 x nyquist ("2x
oversampling") does a much better job . I do all my two way radio speech
processing at "2x", especially peak control, and speaker pitch
estimation (I try and tell the difference between a male and a female
speaker for multiband band-splitters for multiband AGC) .
*****
The 300Hz point is highly valid. Most CTCSS subtone filters would be
3dB down at 400Hz, 20dB at 300 , 30+ dB down at 250.
And of course those filters are cascaded in TX and RX.....



glen
Post by Jean-Marc Valin
Post by Jeroen Vreeken
One of the things that I noticed while testing was that the current
pitch estimation sometimes makes a mistake and ends up with twice the
real pitch. I also got the feeling (but no proof yet) it was worse with
samples with heavy filtering under 300Hz (e.g. to get rid of CTCSS tones).
So there might be some room for improvement there...
Pitch doubling/tripling is a common problem in pitch estimation. I spent
some time trying to avoid it as much as possible in the Opus (CELT)
pitch detector. You might want to have a look just in case (haven't
tested) it works better than the existing codec2 pitch detector. Some of
1) Upsampling the pitch correlation (2x) to get fractional time
resolution. Even if that resolution isn't needed, it helps reduce pitch
doubling).
2) Looking at fractions of the pitch period to find peaks. e.g. if the
correlation at T/3 is almost as good as the one at T, then we pick T/3
Jeroen Vreeken
2017-03-12 15:46:38 UTC
Permalink
Post by David Rowe
Hi Jeroen,
Wow that's great work - thank you so much. So wonderful for me to see
someone diving into the code and making things happen at this level.
Also the results are pretty good - that guy vk5dgr sounded the worst :-)
http://rowetel.com/downloads/codec2/all.1300.wav
What does everyone else on the list think of the two modes compared to
each other? There is an obvious level difference - an artifact of the
post filtering. When listening, compare one sentence at a time, e.g.
the first 3 seconds of both, then the 2nd 3 seconds etc. Think about
which sample you prefer.
1/ How did you train the VQ - using the Octave code?
No, it seems I am very good at completly messing up octave code...
I used a small C program load a set of vectors (e.g. error output of 2nd
stage) and it searches for vector that has the best results when applied
to all other vectors (measured in number of vectors that have a
significant improvement).
The best one is added to the codebook and all vectors that where
improved enough are removed from the set. Then the next vector is
searched by comparing it with the remaining set.
There are a few thresholds being taken into account to make sure the
improvement is big enough. (A very small vector would improve almost the
whole set, but would not add anything usefull)

It is not very quick, but seems to find a nice set of vectors in the end.
Post by David Rowe
2/ The squared error can be directly related to spectral distortion, if
the difference (in dB) between the original and quantised spectrum, so
the extra VQ stages have reduced the RMS distortion by sqrt(147/20) =
2.7 dB to sqrt(44/20) = 1.5dB.
3/ Given more bits are available, might be interesting to expand the
bandwidth the VQ works over. The 1300 samples do sound a bit more
"wideband" to me, compared to 1300C.
Do you mean the upper limit of 3700Hz? Or increasing the rate K from 20
to e.g. 40?
I tried increasing it to 39 (actually 40, but it was not really used)
such that the even factors would align with the 20 of the 700c mode (in
order to reuse the first two stages). But the results did not add as
much as I hoped...
Post by David Rowe
4/ The post filter may not be so important now, and could be "relaxed".
I'll try some different settings, I experimented a bith with it switched
of completly, but it didn't sound as good as switched on. But some
different settings (like 10db/dec or a lesser gain) might be worth trying.
Post by David Rowe
5/ Yes I'm sure the pitch estimator loses it occasionally. To test
objectively pitch contours could be compared with different filtering.
6/ I have some other training material you could try, to compare
results. But sounds like you are on the right track.
Sure, more material the better.. Do you have a download link or another
way to transfer it?

Regards,
Jeroen
David Rowe
2017-03-13 08:18:16 UTC
Permalink
Re (1), OK cool - yes there are many ways to train VQs. Given some
training data (a 12000 x 20 matrix train_120), the steps I used in
Octave are:

octave:12> size(train_120)
ans =

12000 20

octave:13> melvq; vq = trainvq(train_120, 512, 2);
mse stage 1: 3.105040
mse stage 2: 2.149767

This cld be extended to say 4 stages of 9 bit/s stage (36 bits):

Re (3) yes perhaps try increasing the upper and lower frequency range.

Re (6): Sure will send to you off-list.

Cheers,

David
Post by Jeroen Vreeken
Post by David Rowe
Hi Jeroen,
Wow that's great work - thank you so much. So wonderful for me to see
someone diving into the code and making things happen at this level.
Also the results are pretty good - that guy vk5dgr sounded the worst :-)
http://rowetel.com/downloads/codec2/all.1300.wav
What does everyone else on the list think of the two modes compared to
each other? There is an obvious level difference - an artifact of the
post filtering. When listening, compare one sentence at a time, e.g.
the first 3 seconds of both, then the 2nd 3 seconds etc. Think about
which sample you prefer.
1/ How did you train the VQ - using the Octave code?
No, it seems I am very good at completly messing up octave code...
I used a small C program load a set of vectors (e.g. error output of 2nd
stage) and it searches for vector that has the best results when applied
to all other vectors (measured in number of vectors that have a
significant improvement).
The best one is added to the codebook and all vectors that where
improved enough are removed from the set. Then the next vector is
searched by comparing it with the remaining set.
There are a few thresholds being taken into account to make sure the
improvement is big enough. (A very small vector would improve almost the
whole set, but would not add anything usefull)
It is not very quick, but seems to find a nice set of vectors in the end.
Post by David Rowe
2/ The squared error can be directly related to spectral distortion, if
the difference (in dB) between the original and quantised spectrum, so
the extra VQ stages have reduced the RMS distortion by sqrt(147/20) =
2.7 dB to sqrt(44/20) = 1.5dB.
3/ Given more bits are available, might be interesting to expand the
bandwidth the VQ works over. The 1300 samples do sound a bit more
"wideband" to me, compared to 1300C.
Do you mean the upper limit of 3700Hz? Or increasing the rate K from 20
to e.g. 40?
I tried increasing it to 39 (actually 40, but it was not really used)
such that the even factors would align with the 20 of the 700c mode (in
order to reuse the first two stages). But the results did not add as
much as I hoped...
Post by David Rowe
4/ The post filter may not be so important now, and could be "relaxed".
I'll try some different settings, I experimented a bith with it switched
of completly, but it didn't sound as good as switched on. But some
different settings (like 10db/dec or a lesser gain) might be worth trying.
Post by David Rowe
5/ Yes I'm sure the pitch estimator loses it occasionally. To test
objectively pitch contours could be compared with different filtering.
6/ I have some other training material you could try, to compare
results. But sounds like you are on the right track.
Sure, more material the better.. Do you have a download link or another
way to transfer it?
Regards,
Jeroen
------------------------------------------------------------------------------
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
Jeroen Vreeken
2017-05-04 00:10:39 UTC
Permalink
Hi all,

It has been a while, but I did make some steady progress.
(Trying different training strategies can take days at a time)
I made a new version of the patch with some changes:

- I cut out a lot of silence and noises from the training material (like
the breathing). There is no real point in reproducing such noises
perfectly. I also added David's samples and some other nice ones I found
that where different from all I already had. (Also some children voices)
- Instead of three times 256 values (resulting in 24 bits) I used two
tables of 1024 values. (20bits). 24 bits of course give a slightly
better result, but I think the bits are better used now. On average the
error is about 1.7dB lower than 700C.
- Wo and Energy each get an extra bit.
- The last two bits are used for voicing. One of the things I noticed
when listening to audible artifacts was that the combined Wo and voicing
was resulting in Wo being interpolated from the hardcoded non-voiced
default back to the voiced value. This sometimes results in an audible
pitch shift. Having a voice bits seperatly makes this a bit less likely.

Sample:
http://video.vreeken.net/~dml/patches/all.1300C_10b10b5e7wo2v.wav

73,
Jeroen PE1RXQ
David Rowe
2017-05-04 21:23:00 UTC
Permalink
Hi Jeroen,

Hey well done that's sounding pretty good except for that last guy :-)
It would be good to break the samples out and compare with the existing
700C and 1300 modems.

Brady O'Brien has also been doing some work on improving 700C, one
innovations was extending the low freq response.

Cheers,

David
Post by Jeroen Vreeken
Hi all,
It has been a while, but I did make some steady progress.
(Trying different training strategies can take days at a time)
- I cut out a lot of silence and noises from the training material (like
the breathing). There is no real point in reproducing such noises
perfectly. I also added David's samples and some other nice ones I found
that where different from all I already had. (Also some children voices)
- Instead of three times 256 values (resulting in 24 bits) I used two
tables of 1024 values. (20bits). 24 bits of course give a slightly
better result, but I think the bits are better used now. On average the
error is about 1.7dB lower than 700C.
- Wo and Energy each get an extra bit.
- The last two bits are used for voicing. One of the things I noticed
when listening to audible artifacts was that the combined Wo and voicing
was resulting in Wo being interpolated from the hardcoded non-voiced
default back to the voiced value. This sometimes results in an audible
pitch shift. Having a voice bits seperatly makes this a bit less likely.
http://video.vreeken.net/~dml/patches/all.1300C_10b10b5e7wo2v.wav
73,
Jeroen PE1RXQ
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
David Rowe
2017-05-06 00:54:59 UTC
Permalink
Hi,

Has anyone tried the FreeDV GUI 1.2 software on Windows 10? I have one
report of it "not starting". Not sure if it's the 32 or 64 bit version.
I understand there is also some sort of new Windows security feature
that prevents application program files being installed in the "Program
Files" folder.

Richard and I run mainly Linux so it would be great to get some feedback
from Windows users.

Thanks,

David
Post by Jeroen Vreeken
Hi all,
It has been a while, but I did make some steady progress.
(Trying different training strategies can take days at a time)
- I cut out a lot of silence and noises from the training material (like
the breathing). There is no real point in reproducing such noises
perfectly. I also added David's samples and some other nice ones I found
that where different from all I already had. (Also some children voices)
- Instead of three times 256 values (resulting in 24 bits) I used two
tables of 1024 values. (20bits). 24 bits of course give a slightly
better result, but I think the bits are better used now. On average the
error is about 1.7dB lower than 700C.
- Wo and Energy each get an extra bit.
- The last two bits are used for voicing. One of the things I noticed
when listening to audible artifacts was that the combined Wo and voicing
was resulting in Wo being interpolated from the hardcoded non-voiced
default back to the voiced value. This sometimes results in an audible
pitch shift. Having a voice bits seperatly makes this a bit less likely.
http://video.vreeken.net/~dml/patches/all.1300C_10b10b5e7wo2v.wav
73,
Jeroen PE1RXQ
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
Walter Holmes
2017-05-06 01:50:34 UTC
Permalink
I have heard reports that SOME of the win10 users have had to install the
32bit version, and it worked fine. Even though they were using a 64bit
version of Windows.

It has been inconsistent though, as several others were working just fine
with the 64bit version.

I had meant to chase down the root of that, but got sidetracked and haven't
heard of anyone recently having the issue.

Walter/K5WH

-----Original Message-----
From: David Rowe [mailto:***@rowetel.com]
Sent: Friday, May 05, 2017 7:55 PM
To: freetel-***@lists.sourceforge.net; ***@googlegroups.com
Subject: [Freetel-codec2] FreeDV GUI 1.2 on Windows 10

Hi,

Has anyone tried the FreeDV GUI 1.2 software on Windows 10? I have one
report of it "not starting". Not sure if it's the 32 or 64 bit version.
I understand there is also some sort of new Windows security feature that
prevents application program files being installed in the "Program Files"
folder.

Richard and I run mainly Linux so it would be great to get some feedback
from Windows users.

Thanks,

David
Post by Jeroen Vreeken
Hi all,
It has been a while, but I did make some steady progress.
(Trying different training strategies can take days at a time) I made
- I cut out a lot of silence and noises from the training material
(like the breathing). There is no real point in reproducing such
noises perfectly. I also added David's samples and some other nice
ones I found that where different from all I already had. (Also some
children voices)
- Instead of three times 256 values (resulting in 24 bits) I used two
tables of 1024 values. (20bits). 24 bits of course give a slightly
better result, but I think the bits are better used now. On average
the error is about 1.7dB lower than 700C.
- Wo and Energy each get an extra bit.
- The last two bits are used for voicing. One of the things I noticed
when listening to audible artifacts was that the combined Wo and
voicing was resulting in Wo being interpolated from the hardcoded
non-voiced default back to the voiced value. This sometimes results in
an audible pitch shift. Having a voice bits seperatly makes this a bit
less likely.
Post by Jeroen Vreeken
http://video.vreeken.net/~dml/patches/all.1300C_10b10b5e7wo2v.wav
73,
Jeroen PE1RXQ
----------------------------------------------------------------------
-------- Check out the vibrant tech community on one of the world's
most engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
----------------------------------------------------------------------------
--
Check out the vibrant tech community on one of the world's most engaging
tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
Freetel-***@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
Brian Richardson
2017-05-06 02:59:09 UTC
Permalink
I had the same problem. The issue is that the default installation folder is in the \Program Files folder and I think the GUI must attempt to write into this folder and Win 10 will not allow that.

Reinstall it into the C:\ folder and it will work. Otherwise run it as Administrator which also works but you have to do that every time you run it. I do not recall the problem with the 32 bit v1.2 but it is there in the 64 bit version 1.2.

Regards, Brian Richardson
Post by David Rowe
Hi,
Has anyone tried the FreeDV GUI 1.2 software on Windows 10? I have one
report of it "not starting". Not sure if it's the 32 or 64 bit version.
I understand there is also some sort of new Windows security feature
that prevents application program files being installed in the "Program
Files" folder.
Richard and I run mainly Linux so it would be great to get some feedback
from Windows users.
Thanks,
David
Post by Jeroen Vreeken
Hi all,
It has been a while, but I did make some steady progress.
(Trying different training strategies can take days at a time)
- I cut out a lot of silence and noises from the training material (like
the breathing). There is no real point in reproducing such noises
perfectly. I also added David's samples and some other nice ones I found
that where different from all I already had. (Also some children voices)
- Instead of three times 256 values (resulting in 24 bits) I used two
tables of 1024 values. (20bits). 24 bits of course give a slightly
better result, but I think the bits are better used now. On average the
error is about 1.7dB lower than 700C.
- Wo and Energy each get an extra bit.
- The last two bits are used for voicing. One of the things I noticed
when listening to audible artifacts was that the combined Wo and voicing
was resulting in Wo being interpolated from the hardcoded non-voiced
default back to the voiced value. This sometimes results in an audible
pitch shift. Having a voice bits seperatly makes this a bit less likely.
http://video.vreeken.net/~dml/patches/all.1300C_10b10b5e7wo2v.wav
73,
Jeroen PE1RXQ
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
Peter Reichelt
2017-05-06 04:48:05 UTC
Permalink
David

On Windows 8.2 X64

I have set up a shortcut on the desktop to the 64 bit version.

Right click the Icon select "Options"

Click on "Advanced ....."

Tick "Run as administrator"

Press "OK"

Press "Apply"

Now double clicking the icon will bring up a "warning dialogue" about
allowing FreeDV to change items on the computer, and when this is
acknowledged FreeDV now starts. This saves having to remember to right
click and run as administrator each time you wish to start the program.

Don't know if windows 10 is the same.

This info should be on the FreeDV page to alert users that the 64 bit
must be run with administrator privileges when installed in the "Program
Files" directory.

Peter VK5APR
Brian Richardson
2017-05-06 04:53:19 UTC
Permalink
The other option is to install in the root directory. C:\. Win 10 does not allow programs to write data into the Program Files directory.

Brian vk3ccr
Post by Peter Reichelt
David
On Windows 8.2 X64
I have set up a shortcut on the desktop to the 64 bit version.
Right click the Icon select "Options"
Click on "Advanced ....."
Tick "Run as administrator"
Press "OK"
Press "Apply"
Now double clicking the icon will bring up a "warning dialogue" about
allowing FreeDV to change items on the computer, and when this is
acknowledged FreeDV now starts. This saves having to remember to right
click and run as administrator each time you wish to start the program.
Don't know if windows 10 is the same.
This info should be on the FreeDV page to alert users that the 64 bit
must be run with administrator privileges when installed in the "Program
Files" directory.
Peter VK5APR
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
David Rowe
2017-05-06 05:20:14 UTC
Permalink
OK so maybe I need to make sure FreeDV isn't writing any data files, or
if it does write them, they should go to c:\ or some other temp
directory. I think I am just writing a log file as a debug feature
anyway, so it can be disabled.

I'll take a look at this over the next few days, then see if we can get
a new exe to try.

Thanks guys!

David
Post by Brian Richardson
I had the same problem. The issue is that the default installation folder is in the \Program Files folder and I think the GUI must attempt to write into this folder and Win 10 will not allow that.
Reinstall it into the C:\ folder and it will work. Otherwise run it as Administrator which also works but you have to do that every time you run it. I do not recall the problem with the 32 bit v1.2 but it is there in the 64 bit version 1.2.
Regards, Brian Richardson
Post by David Rowe
Hi,
Has anyone tried the FreeDV GUI 1.2 software on Windows 10? I have one
report of it "not starting". Not sure if it's the 32 or 64 bit version.
I understand there is also some sort of new Windows security feature
that prevents application program files being installed in the "Program
Files" folder.
Richard and I run mainly Linux so it would be great to get some feedback
from Windows users.
Thanks,
David
Post by Jeroen Vreeken
Hi all,
It has been a while, but I did make some steady progress.
(Trying different training strategies can take days at a time)
- I cut out a lot of silence and noises from the training material (like
the breathing). There is no real point in reproducing such noises
perfectly. I also added David's samples and some other nice ones I found
that where different from all I already had. (Also some children voices)
- Instead of three times 256 values (resulting in 24 bits) I used two
tables of 1024 values. (20bits). 24 bits of course give a slightly
better result, but I think the bits are better used now. On average the
error is about 1.7dB lower than 700C.
- Wo and Energy each get an extra bit.
- The last two bits are used for voicing. One of the things I noticed
when listening to audible artifacts was that the combined Wo and voicing
was resulting in Wo being interpolated from the hardcoded non-voiced
default back to the voiced value. This sometimes results in an audible
pitch shift. Having a voice bits seperatly makes this a bit less likely.
http://video.vreeken.net/~dml/patches/all.1300C_10b10b5e7wo2v.wav
73,
Jeroen PE1RXQ
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
Richard Shaw
2017-05-06 11:35:45 UTC
Permalink
Post by David Rowe
OK so maybe I need to make sure FreeDV isn't writing any data files, or
if it does write them, they should go to c:\ or some other temp
directory. I think I am just writing a log file as a debug feature
anyway, so it can be disabled.
It also seem like maybe we should install to a different program files
directory on 64bit systems? If we can determine where that is, then I
should be able to adjust the settings for the NSIS installer.

Thanks,
Richard
KF5OIM
Matthew Pitts
2017-05-06 17:28:22 UTC
Permalink
Richard and David,


Windows versions, if installed in C:\Program Files\ (or C:\Program
Files(x86) for 32 bit versions on 64 Bit OS)) should write any needed
data to the C:\Users\<uesrname>\AppData\Roaming\FreeDV\ folder. This has
actually been the preferred method since Windows XP, but has only been
enforced since Windows 8, from what I understand. Doing it this way
avoids the privilege escalation that occurs when running the software as
Administrator, which is just as dangerous (if not more so) than running
applications as root user on Linux.


Matthew Pitts

N8OHU
Post by David Rowe
OK so maybe I need to make sure FreeDV isn't writing any data files, or
if it does write them, they should go to c:\ or some other temp
directory. I think I am just writing a log file as a debug feature
anyway, so it can be disabled.
It also seem like maybe we should install to a different program files
directory on 64bit systems? If we can determine where that is, then I
should be able to adjust the settings for the NSIS installer.
Thanks,
Richard
KF5OIM
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
David Rowe
2017-05-06 20:40:14 UTC
Permalink
Thanks Mathew - I wonder if there is a Win32 system call to return that
path. As well as debug/log information there are also times when the
program records wave files, but the user gets a dialog box to choose
record file locations.

- David
Post by Matthew Pitts
Richard and David,
Windows versions, if installed in C:\Program Files\ (or C:\Program
Files(x86) for 32 bit versions on 64 Bit OS)) should write any needed
data to the C:\Users\<uesrname>\AppData\Roaming\FreeDV\ folder. This has
actually been the preferred method since Windows XP, but has only been
enforced since Windows 8, from what I understand. Doing it this way
avoids the privilege escalation that occurs when running the software as
Administrator, which is just as dangerous (if not more so) than running
applications as root user on Linux.
Matthew Pitts
N8OHU
Post by David Rowe
OK so maybe I need to make sure FreeDV isn't writing any data files, or
if it does write them, they should go to c:\ or some other temp
directory. I think I am just writing a log file as a debug feature
anyway, so it can be disabled.
It also seem like maybe we should install to a different program files
directory on 64bit systems? If we can determine where that is, then I
should be able to adjust the settings for the NSIS installer.
Thanks,
Richard
KF5OIM
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
Sebastien F4GRX
2017-05-08 11:21:54 UTC
Permalink
Hello,

that should be GetUserProfileDirectory:

https://msdn.microsoft.com/en-us/library/windows/desktop/bb762280(v=vs.85).aspx

the hToken parameter probably can be provided by GetCurrentProcessToken:

https://msdn.microsoft.com/en-us/library/windows/desktop/mt643211(v=vs.85).aspx

Another solution based on SDD (stackoverflow driven development) is
SHGetFolderPath:

http://stackoverflow.com/questions/9542611/how-to-get-the-current-users-home-directory-in-windows

Sebastien
Post by David Rowe
Thanks Mathew - I wonder if there is a Win32 system call to return that
path. As well as debug/log information there are also times when the
program records wave files, but the user gets a dialog box to choose
record file locations.
- David
Post by Matthew Pitts
Richard and David,
Windows versions, if installed in C:\Program Files\ (or C:\Program
Files(x86) for 32 bit versions on 64 Bit OS)) should write any needed
data to the C:\Users\<uesrname>\AppData\Roaming\FreeDV\ folder. This has
actually been the preferred method since Windows XP, but has only been
enforced since Windows 8, from what I understand. Doing it this way
avoids the privilege escalation that occurs when running the software as
Administrator, which is just as dangerous (if not more so) than running
applications as root user on Linux.
Matthew Pitts
N8OHU
Post by David Rowe
OK so maybe I need to make sure FreeDV isn't writing any data files, or
if it does write them, they should go to c:\ or some other temp
directory. I think I am just writing a log file as a debug feature
anyway, so it can be disabled.
It also seem like maybe we should install to a different program files
directory on 64bit systems? If we can determine where that is, then I
should be able to adjust the settings for the NSIS installer.
Thanks,
Richard
KF5OIM
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
David Rowe
2017-05-08 21:26:37 UTC
Permalink
Thanks Sebastien and everyone else for their comments and suggestions.
At this stage I've simply disabled writing to the log file. Richard is
building a new package for you all to try on Win 10/64 bit systems.

- David
Post by Sebastien F4GRX
Hello,
https://msdn.microsoft.com/en-us/library/windows/desktop/bb762280(v=vs.85).aspx
https://msdn.microsoft.com/en-us/library/windows/desktop/mt643211(v=vs.85).aspx
Another solution based on SDD (stackoverflow driven development) is
http://stackoverflow.com/questions/9542611/how-to-get-the-current-users-home-directory-in-windows
Sebastien
Post by David Rowe
Thanks Mathew - I wonder if there is a Win32 system call to return that
path. As well as debug/log information there are also times when the
program records wave files, but the user gets a dialog box to choose
record file locations.
- David
Post by Matthew Pitts
Richard and David,
Windows versions, if installed in C:\Program Files\ (or C:\Program
Files(x86) for 32 bit versions on 64 Bit OS)) should write any needed
data to the C:\Users\<uesrname>\AppData\Roaming\FreeDV\ folder. This has
actually been the preferred method since Windows XP, but has only been
enforced since Windows 8, from what I understand. Doing it this way
avoids the privilege escalation that occurs when running the software as
Administrator, which is just as dangerous (if not more so) than running
applications as root user on Linux.
Matthew Pitts
N8OHU
Post by David Rowe
OK so maybe I need to make sure FreeDV isn't writing any data files, or
if it does write them, they should go to c:\ or some other temp
directory. I think I am just writing a log file as a debug feature
anyway, so it can be disabled.
It also seem like maybe we should install to a different program files
directory on 64bit systems? If we can determine where that is, then I
should be able to adjust the settings for the NSIS installer.
Thanks,
Richard
KF5OIM
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
Alan Beard
2017-05-11 21:38:17 UTC
Permalink
Hi guys,

I've just downloaded http://files.freedv.org/freedv/freedv-1.2.tar.xz

on my Fedora 23 Linux box where FreeDV used to work until ~6 months ago.
Hil VK2IUW and I had some QSOs over 10Km on 6m.

This is the only thing I did to get it compiling and thus working:

grep 3.0 CMakeLists.txt
    set(WXCONFIG "/usr/bin/wx-config-3.0" CACHE FILEPATH "Location of wx-config binary.")
    set(WXRC "/usr/bin/wxrc-3.0" CACHE FILEPATH "Location of wxrc binary.")
set(WX_VERSION_MIN 3.0.0)

My problem:
Transmit mic in. Nothing on the "Level" bar-graph or the oscilloscope screen.

"Pulseaudio" from desktop "Settings" shows my mic is working.

I can make an RDP session available for your testing though RDP on Linux
doesn't give mic and speaker.

80 'cause a 73 won't do nothing useful here.

Alan VK2ZIW
Post by David Rowe
Thanks Sebastien and everyone else for their comments and
suggestions. At this stage I've simply disabled writing to the log
file.  Richard is building a new package for you  all to try on Win
10/64 bit systems.
- David
Post by Sebastien F4GRX
Hello,
https://msdn.microsoft.com/en-us/library/windows/desktop/bb762280(v=vs.85).aspx
https://msdn.microsoft.com/en-us/library/windows/desktop/mt643211(v=vs.85).aspx
Post by David Rowe
Post by Sebastien F4GRX
Another solution based on SDD (stackoverflow driven development) is
http://stackoverflow.com/questions/9542611/how-to-get-the-current-users-home-directory-in-windows
Post by David Rowe
Post by Sebastien F4GRX
Sebastien
Post by David Rowe
Thanks Mathew - I wonder if there is a Win32 system call to return that
path.  As well as debug/log information there are also times when the
program records wave files, but the user gets a dialog box to choose
record file locations.
- David
Post by Matthew Pitts
Richard and David,
Windows versions, if installed in C:\Program Files\ (or C:\Program
Files(x86) for 32 bit versions on 64 Bit OS)) should write any needed
data to the C:\Users\<uesrname>\AppData\Roaming\FreeDV\ folder. This has
actually been the preferred method since Windows XP, but has only been
enforced since Windows 8, from what I understand. Doing it this way
avoids the privilege escalation that occurs when running the software as
Administrator, which is just as dangerous (if not more so) than running
applications as root user on Linux.
Matthew Pitts
N8OHU
    OK so maybe I need to make sure FreeDV isn't writing any data
    files, or
    if it does write them, they should go to c:\ or some other temp
    directory. I think I am just writing a log file as a debug feature
    anyway, so it can be disabled.
It also seem like maybe we should install to a different program files
directory on 64bit systems? If we can determine where that is, then I
should be able to adjust the settings for the NSIS installer.
Thanks,
Richard
KF5OIM
------------------------------------------------------------------------------
Post by David Rowe
Post by Sebastien F4GRX
Post by David Rowe
Post by Matthew Pitts
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
------------------------------------------------------------------------------
Post by David Rowe
Post by Sebastien F4GRX
Post by David Rowe
Post by Matthew Pitts
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
------------------------------------------------------------------------------
Post by David Rowe
Post by Sebastien F4GRX
Post by David Rowe
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
Alan

Evil flourishes when good men do nothing.
Consider the Christmas child.
---------------------------------------------------------------------------
Alan Beard               Unix Support Technician from 1984 to today
70 Wedmore Rd.           Sun Solaris, AIX, HP/UX, Linux, SCO, MIPS
Emu Heights N.S.W. 2750  Routers, terminal servers, printers, terminals etc..
+61 2 47353013 (h)       Support Programming, shell scripting, "C", assembler
0414 353013 (mobile)     After uni, electronics tech
David Rowe
2017-05-11 22:12:06 UTC
Permalink
Hi Alan,

Sounds like a portaudio/sound card issue.

Couple of ideas:

1/ Try using the readme.txt option to compile various libraries (in
particular PortAudio) from source.

2/ Do you get tx audio when you use an 8kHz 16 bit int wavefile for the
audio source using the Tools - Start/Stop Play file from Mic In

3/ Can you use the Fedora FreeDV packages (see freedv.org)?

Cheers,

David
Post by Alan Beard
Hi guys,
I've just downloaded *http://files.freedv.org/freedv/freedv-1.2.tar.xz*
on my Fedora 23 Linux box where FreeDV used to work until ~6 months ago.
Hil VK2IUW and I had some QSOs over 10Km on 6m.
grep 3.0 CMakeLists.txt
set(WXCONFIG "/usr/bin/wx-config-3.0" CACHE FILEPATH "Location of
wx-config binary.")
set(WXRC "/usr/bin/wxrc-3.0" CACHE FILEPATH "Location of wxrc binary.")
set(WX_VERSION_MIN 3.0.0)
Transmit mic in.* Nothing on the "Level" bar-graph or the oscilloscope
screen.
"Pulseaudio" from desktop "Settings" shows my mic is working.
I can make an RDP session available for your testing though RDP on Linux
doesn't give mic and speaker.
80 'cause a 73 won't do nothing useful here.
Alan VK2ZIW
Post by David Rowe
Thanks Sebastien and everyone else for their comments and
suggestions. At this stage I've simply disabled writing to the log
file. Richard is building a new package for you all to try on Win
10/64 bit systems.
- David
Post by Sebastien F4GRX
Hello,
https://msdn.microsoft.com/en-us/library/windows/desktop/bb762280(v=vs.85).aspx
Post by David Rowe
Post by Sebastien F4GRX
the hToken parameter probably can be provided by
https://msdn.microsoft.com/en-us/library/windows/desktop/mt643211(v=vs.85).aspx
Post by David Rowe
Post by Sebastien F4GRX
Another solution based on SDD (stackoverflow driven development) is
http://stackoverflow.com/questions/9542611/how-to-get-the-current-users-home-directory-in-windows
Post by David Rowe
Post by Sebastien F4GRX
Sebastien
Post by David Rowe
Thanks Mathew - I wonder if there is a Win32 system call to return
that
Post by David Rowe
Post by Sebastien F4GRX
Post by David Rowe
path. As well as debug/log information there are also times when the
program records wave files, but the user gets a dialog box to choose
record file locations.
- David
Post by Matthew Pitts
Richard and David,
Windows versions, if installed in C:\Program Files\ (or C:\Program
Files(x86) for 32 bit versions on 64 Bit OS)) should write any needed
data to the C:\Users\<uesrname>\AppData\Roaming\FreeDV\ folder.
This has
Post by David Rowe
Post by Sebastien F4GRX
Post by David Rowe
Post by Matthew Pitts
actually been the preferred method since Windows XP, but has only
been
Post by David Rowe
Post by Sebastien F4GRX
Post by David Rowe
Post by Matthew Pitts
enforced since Windows 8, from what I understand. Doing it this way
avoids the privilege escalation that occurs when running the
software as
Post by David Rowe
Post by Sebastien F4GRX
Post by David Rowe
Post by Matthew Pitts
Administrator, which is just as dangerous (if not more so) than
running
Post by David Rowe
Post by Sebastien F4GRX
Post by David Rowe
Post by Matthew Pitts
applications as root user on Linux.
Matthew Pitts
N8OHU
Post by David Rowe
OK so maybe I need to make sure FreeDV isn't writing any data files, or
if it does write them, they should go to c:\ or some other temp
directory. I think I am just writing a log file as a debug
feature
Post by David Rowe
Post by Sebastien F4GRX
Post by David Rowe
Post by Matthew Pitts
Post by David Rowe
anyway, so it can be disabled.
It also seem like maybe we should install to a different program
files
Post by David Rowe
Post by Sebastien F4GRX
Post by David Rowe
Post by Matthew Pitts
Post by David Rowe
directory on 64bit systems? If we can determine where that is,
then I
Post by David Rowe
Post by Sebastien F4GRX
Post by David Rowe
Post by Matthew Pitts
Post by David Rowe
should be able to adjust the settings for the NSIS installer.
Thanks,
Richard
KF5OIM
------------------------------------------------------------------------------
Post by David Rowe
Post by Sebastien F4GRX
Post by David Rowe
Post by Matthew Pitts
Post by David Rowe
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
------------------------------------------------------------------------------
Post by David Rowe
Post by Sebastien F4GRX
Post by David Rowe
Post by Matthew Pitts
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
------------------------------------------------------------------------------
Post by David Rowe
Post by Sebastien F4GRX
Post by David Rowe
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
------------------------------------------------------------------------------
Post by David Rowe
Post by Sebastien F4GRX
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
------------------------------------------------------------------------------
Post by David Rowe
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
Alan
Evil flourishes when good men do nothing.
Consider the Christmas child.
---------------------------------------------------------------------------
Alan Beard Unix Support Technician from 1984 to today
70 Wedmore Rd. Sun Solaris, AIX, HP/UX, Linux, SCO, MIPS
Emu Heights N.S.W. 2750 Routers, terminal servers, printers, terminals
etc..
+61 2 47353013 (h) Support Programming, shell scripting, "C",
assembler
0414 353013 (mobile) After uni, electronics tech
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
Stuart Longland
2017-05-11 22:14:11 UTC
Permalink
Post by Alan Beard
Transmit mic in.* Nothing on the "Level" bar-graph or the oscilloscope
screen.
"Pulseaudio" from desktop "Settings" shows my mic is working.
I can make an RDP session available for your testing though RDP on Linux
doesn't give mic and speaker.
Very stupid question on my part… is FreeDV set to use PulseAudio for
that part?

If PulseAudio is hogging the sound device and FreeDV is set to use the
sound device directly, it won't get a look in.

I find the sound interface plugged into the transceiver *should NOT* be
accessed via PulseAudio, which means going into the sound settings
applet (e.g. `padevchooser`) and making sure that sound device is
de-selected.

The sound interface representing your "human" interface (e.g. headset or
speaker/mic) *should* go through PulseAudio.

As for RDP; it might be possible to expose a SPICE session instead,
which does carry sound. RDP won't because a Linux RDP server is
basically an X server, and X does not have sound.
--
Stuart Longland (aka Redhatter, VK4MSL)

I haven't lost my mind...
...it's backed up on a tape somewhere.
Alan Beard
2017-05-13 04:40:55 UTC
Permalink
Hi Stuart and all,

Transmit "Mic in" from "pulse" works. I'm txing again.

Now for some On-air.

Yes, this is a compile of 1.2 on Fedora 23 x86_64 from source.

Thanks

Alan VK2ZIW
Post by Stuart Longland
Post by Alan Beard
Transmit mic in.* Nothing on the "Level" bar-graph or the oscilloscope
screen.
"Pulseaudio" from desktop "Settings" shows my mic is working.
I can make an RDP session available for your testing though RDP on Linux
doesn't give mic and speaker.
Very stupid question on my part… is FreeDV set to use PulseAudio
for that part?
If PulseAudio is hogging the sound device and FreeDV is set to use
the sound device directly, it won't get a look in.
I find the sound interface plugged into the transceiver *should NOT*
be accessed via PulseAudio, which means going into the sound
settings applet (e.g. `padevchooser`) and making sure that sound
device is de-selected.
The sound interface representing your "human" interface (e.g.
headset or speaker/mic) *should* go through PulseAudio.
As for RDP; it might be possible to expose a SPICE session instead,
which does carry sound. RDP won't because a Linux RDP server is
basically an X server, and X does not have sound.
--
Stuart Longland (aka Redhatter, VK4MSL)
I haven't lost my mind...
...it's backed up on a tape somewhere.
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
Alan

Evil flourishes when good men do nothing.
Consider the Christmas child.
---------------------------------------------------------------------------
Alan Beard Unix Support Technician from 1984 to today
70 Wedmore Rd. Sun Solaris, AIX, HP/UX, Linux, SCO, MIPS
Emu Heights N.S.W. 2750 Routers, terminal servers, printers, terminals etc..
+61 2 47353013 (h) Support Programming, shell scripting, "C", assembler
0414 353013 (mobile) After uni, electronics tech
David Rowe
2017-05-18 00:31:32 UTC
Permalink
Some initial results for the latest and greatest:

http://www.rowetel.com/?p=5573

- David
Alan Beard
2017-05-19 03:29:02 UTC
Permalink
Hi David,

Can you tell us when 700D is available to test in FreeDV?
(the GIT branch)

And, can we coordinate skeds via VKLOGGER.com ?

Alan VK2ZIW
Post by David Rowe
http://www.rowetel.com/?p=5573
- David
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
Alan

Evil flourishes when good men do nothing.
Consider the Christmas child.
---------------------------------------------------------------------------
Alan Beard Unix Support Technician from 1984 to today
70 Wedmore Rd. Sun Solaris, AIX, HP/UX, Linux, SCO, MIPS
Emu Heights N.S.W. 2750 Routers, terminal servers, printers, terminals etc..
+61 2 47353013 (h) Support Programming, shell scripting, "C", assembler
0414 353013 (mobile) After uni, electronics tech
David Rowe
2017-05-19 05:00:58 UTC
Permalink
Hi Alan,

Pls feel free to set up/coordinate skeds, more FreeDV activity is a good
thing.

One 700D reaches prime time, for sure - there will be an announcement.

Cheers,

David
Post by Alan Beard
Hi David,
Can you tell us when 700D is available to test in FreeDV?
(the GIT branch)
And, can we coordinate skeds via VKLOGGER.com ?
Alan VK2ZIW
Post by David Rowe
http://www.rowetel.com/?p=5573
- David
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
Alan
Evil flourishes when good men do nothing.
Consider the Christmas child.
---------------------------------------------------------------------------
Alan Beard Unix Support Technician from 1984 to today
70 Wedmore Rd. Sun Solaris, AIX, HP/UX, Linux, SCO, MIPS
Emu Heights N.S.W. 2750 Routers, terminal servers, printers, terminals etc..
+61 2 47353013 (h) Support Programming, shell scripting, "C", assembler
0414 353013 (mobile) After uni, electronics tech
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
Richard Shaw
2017-05-12 00:30:22 UTC
Permalink
Fedora 23 has not been supported for some time. I would recommend you
upgrade to Fedora 25 at least where it's in the repos. If you're not ready
for that yet, at least pull the codec2 and FreeDV source RPMS from Fedora
25 and rebuild them for Fedora 23.

Thanks,
Richard
David Rowe
2017-05-11 22:04:39 UTC
Permalink
Hi Jeroen and list,

I have broken up Jeroen's sample below into separate sample files, and
for comparison added the same samples processed by Codec 2 700C and 1300:

http://rowetel.com/downloads/codec2/jeroen_rfc_1300c.tar.gz

This way each sample can be compared to 700C and 1300. BTW the script
to separate all.wav into separate files is:

codec2-dev/scripts/separate_all.sh.

When I compare codecs, I take one (say 3 second) sample and listen to it
processed by different codecs. I have attached a screen shot of how I
organise my file manager for these samples.

What do you all think of Jeroen's 1300 candidate compared to 700C and
1300. Is it superior to the current 1300, or are they roughly the same?
Which samples work well, which ones are problematic?

In this way we look at Jeroens fine work, find out which techniques
work, and where some more work is required.

Crowd-sourced codec evaluation......

Thanks,

David
Post by Jeroen Vreeken
Hi all,
It has been a while, but I did make some steady progress.
(Trying different training strategies can take days at a time)
- I cut out a lot of silence and noises from the training material (like
the breathing). There is no real point in reproducing such noises
perfectly. I also added David's samples and some other nice ones I found
that where different from all I already had. (Also some children voices)
- Instead of three times 256 values (resulting in 24 bits) I used two
tables of 1024 values. (20bits). 24 bits of course give a slightly
better result, but I think the bits are better used now. On average the
error is about 1.7dB lower than 700C.
- Wo and Energy each get an extra bit.
- The last two bits are used for voicing. One of the things I noticed
when listening to audible artifacts was that the combined Wo and voicing
was resulting in Wo being interpolated from the hardcoded non-voiced
default back to the voiced value. This sometimes results in an audible
pitch shift. Having a voice bits seperatly makes this a bit less likely.
http://video.vreeken.net/~dml/patches/all.1300C_10b10b5e7wo2v.wav
73,
Jeroen PE1RXQ
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
Peter Reichelt
2017-05-06 02:02:02 UTC
Permalink
Hi David

I just downloaded 64 bit program from FreeDV.org onto a windows 8.1
machine and FreeDV 1.2 will not start. FreeDV 1.1(32 bit) works.

FreeDV 1.2 (32 bit) is OK.

This is the error message displayed



Problem signature:
Problem Event Name: APPCRASH
Application Name: freedv.exe
Application Version: 0.0.0.0
Application Timestamp: 00000000
Fault Module Name: ntdll.dll
Fault Module Version: 6.3.9600.18438
Fault Module Timestamp: 57ae642e
Exception Code: c0000005
Exception Offset: 000000000003b6a9
OS Version: 6.3.9600.2.0.0.768.101
Locale ID: 3081
Additional Information 1: 2101
Additional Information 2: 2101a04c25076826ce88a8c57a299ae3
Additional Information 3: e88e
Additional Information 4: e88eaca8dbf1b630f3a9f1af99de65ba

Read our privacy statement online:
http://go.microsoft.com/fwlink/?linkid=280262

If the online privacy statement is not available, please read our
privacy statement offline:
C:\Windows\system32\en-US\erofflps.txt


Regards

Peter V5APR

------ Original Message ------
From: "David Rowe" <***@rowetel.com>
To: freetel-***@lists.sourceforge.net; ***@googlegroups.com
Sent: Saturday, 6 May, 2017 At 10:24 AM
Subject: [Freetel-codec2] FreeDV GUI 1.2 on Windows 10

Hi,

Has anyone tried the FreeDV GUI 1.2 software on Windows 10? I have one
report of it "not starting". Not sure if it's the 32 or 64 bit version.
I understand there is also some sort of new Windows security feature
that prevents application program files being installed in the "Program
Files" folder.

Richard and I run mainly Linux so it would be great to get some feedback
from Windows users.

Thanks,

David
Post by Jeroen Vreeken
Hi all,
It has been a while, but I did make some steady progress.
(Trying different training strategies can take days at a time)
- I cut out a lot of silence and noises from the training material
(like
the breathing). There is no real point in reproducing such noises
perfectly. I also added David's samples and some other nice ones I
found
that where different from all I already had. (Also some children
voices)
- Instead of three times 256 values (resulting in 24 bits) I used two
tables of 1024 values. (20bits). 24 bits of course give a slightly
better result, but I think the bits are better used now. On average
the
error is about 1.7dB lower than 700C.
- Wo and Energy each get an extra bit.
- The last two bits are used for voicing. One of the things I noticed
when listening to audible artifacts was that the combined Wo and
voicing
was resulting in Wo being interpolated from the hardcoded non-voiced
default back to the voiced value. This sometimes results in an audible
pitch shift. Having a voice bits seperatly makes this a bit less
likely.
http://video.vreeken.net/~dml/patches/all.1300C_10b10b5e7wo2v.wav
73,
Jeroen PE1RXQ
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
Freetel-***@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
Brady O'Brien
2017-05-06 02:08:59 UTC
Permalink
Stackoverflow says "Exception Code: c000005" is basically a segfault

http://stackoverflow.com/questions/17168982/exception-error-c0000005-in-vc
Post by Peter Reichelt
Hi David
I just downloaded 64 bit program from FreeDV.org onto a windows 8.1
machine and FreeDV 1.2 will not start. FreeDV 1.1(32 bit) works.
FreeDV 1.2 (32 bit) is OK.
This is the error message displayed
Problem Event Name: APPCRASH
Application Name: freedv.exe
Application Version: 0.0.0.0
Application Timestamp: 00000000
Fault Module Name: ntdll.dll
Fault Module Version: 6.3.9600.18438
Fault Module Timestamp: 57ae642e
Exception Code: c0000005
Exception Offset: 000000000003b6a9
OS Version: 6.3.9600.2.0.0.768.101
Locale ID: 3081
Additional Information 1: 2101
Additional Information 2: 2101a04c25076826ce88a8c57a299ae3
Additional Information 3: e88e
Additional Information 4: e88eaca8dbf1b630f3a9f1af99de65ba
http://go.microsoft.com/fwlink/?linkid=280262
If the online privacy statement is not available, please read our
C:\Windows\system32\en-US\erofflps.txt
Regards
Peter V5APR
------ Original Message ------
Sent: Saturday, 6 May, 2017 At 10:24 AM
Subject: [Freetel-codec2] FreeDV GUI 1.2 on Windows 10
Hi,
Has anyone tried the FreeDV GUI 1.2 software on Windows 10? I have one
report of it "not starting". Not sure if it's the 32 or 64 bit version.
I understand there is also some sort of new Windows security feature
that prevents application program files being installed in the "Program
Files" folder.
Richard and I run mainly Linux so it would be great to get some feedback
from Windows users.
Thanks,
David
Post by Jeroen Vreeken
Hi all,
It has been a while, but I did make some steady progress.
(Trying different training strategies can take days at a time)
- I cut out a lot of silence and noises from the training material
(like
the breathing). There is no real point in reproducing such noises
perfectly. I also added David's samples and some other nice ones I
found
that where different from all I already had. (Also some children
voices)
- Instead of three times 256 values (resulting in 24 bits) I used two
tables of 1024 values. (20bits). 24 bits of course give a slightly
better result, but I think the bits are better used now. On average
the
error is about 1.7dB lower than 700C.
- Wo and Energy each get an extra bit.
- The last two bits are used for voicing. One of the things I noticed
when listening to audible artifacts was that the combined Wo and
voicing
was resulting in Wo being interpolated from the hardcoded non-voiced
default back to the voiced value. This sometimes results in an audible
pitch shift. Having a voice bits seperatly makes this a bit less
likely.
http://video.vreeken.net/~dml/patches/all.1300C_10b10b5e7wo2v.wav
73,
Jeroen PE1RXQ
------------------------------------------------------------
------------------
Post by Jeroen Vreeken
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
------------------------------------------------------------
------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
------------------------------------------------------------
------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
Peter Reichelt
2017-05-06 04:05:08 UTC
Permalink
All

The 64 bit version started normaly when right clicking the icon and "run
as administrator" is clicked


Peter VK5APR

------ Original Message ------
From: "Brady O'Brien" <***@gmail.com>
To: freetel-***@lists.sourceforge.net
Sent: Saturday, 6 May, 2017 At 11:38 AM
Subject: Re: [Freetel-codec2] FreeDV GUI 1.2 on Windows 10

Stackoverflow says "Exception Code: c000005" is basically a segfault
http://stackoverflow.com/questions/17168982/exception-error-c0000005-in-vc
<http://stackoverflow.com/questions/17168982/exception-error-c0000005-in-vc>


On Fri, May 5, 2017 at 9:02 PM, Peter Reichelt
<***@bigpond.com <mailto:***@bigpond.com> > wrote:
Hi David

I just downloaded 64 bit program from FreeDV.org onto a windows 8.1
machine and FreeDV 1.2 will not start. FreeDV 1.1(32 bit)  works.

FreeDV 1.2 (32 bit) is OK.

This is the error message displayed



Problem signature:
   Problem Event Name: APPCRASH
   Application Name: freedv.exe
   Application Version: 0.0.0.0
   Application Timestamp: 00000000
   Fault Module Name: ntdll.dll
   Fault Module Version: 6.3.9600.18438
   Fault Module Timestamp: 57ae642e
   Exception Code: c0000005
   Exception Offset: 000000000003b6a9
   OS Version: 6.3.9600.2.0.0.768.101
   Locale ID: 3081
   Additional Information 1: 2101
   Additional Information 2: 2101a04c25076826ce88a8c57a299ae3
   Additional Information 3: e88e
   Additional Information 4: e88eaca8dbf1b630f3a9f1af99de65ba

Read our privacy statement online:
   http://go.microsoft.com/fwlink/?linkid=280262
<http://go.microsoft.com/fwlink/?linkid=280262>

If the online privacy statement is not available, please read our
privacy statement offline:
   C:\Windows\system32\en-US\erofflps.txt


Regards

Peter V5APR


------ Original Message ------
From: "David Rowe" <***@rowetel.com <mailto:***@rowetel.com> >
To: freetel-***@lists.sourceforge.net
<mailto:freetel-***@lists.sourceforge.net> ;
***@googlegroups.com <mailto:***@googlegroups.com>
Sent: Saturday, 6 May, 2017 At 10:24 AM
Subject: [Freetel-codec2] FreeDV GUI 1.2 on Windows 10

Hi,

Has anyone tried the FreeDV GUI 1.2 software on Windows 10?  I have
one
report of it "not starting".  Not sure if it's the 32 or 64 bit
version.
I understand there is also some sort of new Windows security feature
that prevents application program files being installed in the
"Program
Files" folder.

Richard and I run mainly Linux so it would be great to get some
feedback
from Windows users.

Thanks,

David
Post by Jeroen Vreeken
Hi all,
It has been a while, but I did make some steady progress.
(Trying different training strategies can take days at a time)
- I cut out a lot of silence and noises from the training material
(like
the breathing). There is no real point in reproducing such noises
perfectly. I also added David's samples and some other nice ones I
found
that where different from all I already had. (Also some children
voices)
- Instead of three times 256 values (resulting in 24 bits) I used two
tables of 1024 values. (20bits). 24 bits of course give a slightly
better result, but I think the bits are better used now. On average
the
error is about 1.7dB lower than 700C.
- Wo and Energy each get an extra bit.
- The last two bits are used for voicing. One of the things I
noticed
Post by Jeroen Vreeken
when listening to audible artifacts was that the combined Wo and
voicing
was resulting in Wo being interpolated from the hardcoded non-voiced
default back to the voiced value. This sometimes results in an
audible
Post by Jeroen Vreeken
pitch shift. Having a voice bits seperatly makes this a bit less
likely.
http://video.vreeken.net/~dml/patches/all.1300C_10b10b5e7wo2v.wav
<http://video.vreeken.net/~dml/patches/all.1300C_10b10b5e7wo2v.wav>
Post by Jeroen Vreeken
73,
Jeroen PE1RXQ
------------------------------------------------------------------------------
Post by Jeroen Vreeken
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
<http://sdm.link/slashdot>
Post by Jeroen Vreeken
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
<https://lists.sourceforge.net/lists/listinfo/freetel-codec2>
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
<http://sdm.link/slashdot>
_______________________________________________
Freetel-codec2 mailing list
Freetel-***@lists.sourceforge.net
<mailto:Freetel-***@lists.sourceforge.net>
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
<https://lists.sourceforge.net/lists/listinfo/freetel-codec2>





------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
<http://sdm.link/slashdot>
_______________________________________________
Freetel-codec2 mailing list
Freetel-***@lists.sourceforge.net
<mailto:Freetel-***@lists.sourceforge.net>
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
<https://lists.sourceforge.net/lists/listinfo/freetel-codec2>
Jeroen Vreeken
2017-05-11 22:27:48 UTC
Permalink
Post by David Rowe
When I compare codecs, I take one (say 3 second) sample and listen to
it processed by different codecs. I have attached a screen shot of
how I organise my file manager for these samples.
I use a similar approach: I often load two or three files in audacity
and select a few seconds at a time (sometimes just a single word) and
play the different tracks one by one using 'mute' and 'solo'.

When you select a spectrum view some artifacts even become visible (like
a Wo miss)

Regards,
Jeroen
David Rowe
2017-05-12 00:28:47 UTC
Permalink
FB Jeroen - that's a great way to dig into the differences between
samples when working on the algorithms. Some examples of the Audacity
spectrogram view towards the bottom of this post:

http://www.rowetel.com/?p=4454

Hmm, maybe worth writing a blog post on 'Codec 2 hacking', i.e. a worked
example of codec development.

- David
Post by Jeroen Vreeken
Post by David Rowe
When I compare codecs, I take one (say 3 second) sample and listen to
it processed by different codecs. I have attached a screen shot of
how I organise my file manager for these samples.
I use a similar approach: I often load two or three files in audacity
and select a few seconds at a time (sometimes just a single word) and
play the different tracks one by one using 'mute' and 'solo'.
When you select a spectrum view some artifacts even become visible (like
a Wo miss)
Regards,
Jeroen
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
Ross Whenmouth
2017-05-12 10:10:20 UTC
Permalink
Hi,

I've had a listen to Jeroen's samples, and to my ears (which work, but
aren't golden!), 1300C generally sounds better than the old 1300 codec.
The 1300C codec seems to be less "buzzy" on the audio peaks and some
parts of speech sound clearer / less mumbled. The 1300C samples also
seem to have a bit less bass / low frequency content than the 1300
samples (a good thing IMHO).

For example, in cq_ref_all (KA9Q), the phrase "kilowatt alpha" sounds
more distorted / buzzy (clipping?) in 1300, than it does in 1300C. The
"cq cq cq" also sounds more distorted in 1300.

In hts2_all (these days a chicken is a rare dish), 1300 sounds more
mumbled around about the end of "..chicken" going into "is a rare..."
than in 1300C. I haven't looked at an FFT, but the mumble sounds like
the treble suddenly disappears from the audio in 1300 (sounds odd for
female speaker - 1300C seems to maintain its treble and sounds clearer).

In vk5dgr_all, it sounds like the audio is overdriven in both 1300 and
1300C, though the reduced bass level in 1300C and the perceived reduced
duration of the clicks around the audio peaks (clipping?) make the 1300C
sample easier to listen to than the "buzzing" that takes their place in
1300. I think that the word "delta" in this sample is an example of
where 1300C might be able to be improved - "delta" is said very quickly
and to my ears, it sounds almost like the "ta" sound on the end of it is
truncated.

73 ZL2WRW
Ross Whenmouth
Continue reading on narkive:
Loading...