esldude Posted May 25, 2019 Share Posted May 25, 2019 I took the fast file above and slowed it 30 ppm in Audacity and retested. Notice the high frequency noise is much less and no 20 hz hump. Also notice the difference RMS is now 9 db lower. Now this isn't new. I went back to v 26b and it also puts the high frequency noise and the 20 hz hump in there with the greater than 30 ppm speed difference. However, it only managed a poor match rather than good or very good. And the difference RMS with the large ppm difference was worse by 27 db instead of only 9 db worse. So v 30b handles all this better than the prior versions. And always keep in mind: Cognitive biases, like seeing optical illusions are a sign of a normally functioning brain. We all have them, it’s nothing to be ashamed about, but it is something that affects our objective evaluation of reality. Link to comment
esldude Posted May 25, 2019 Share Posted May 25, 2019 Again while working with Archimago's files I'm finding in nearly every case the sample offset can be done manually to 1 or 2 or sometimes more samples and get a better match. As if the sample offset has a problem or something about these files is causing a hiccup. The results improve when you do this. This btw is working on samples I've changed speed upon to get it less than 10 ppm different. It happens with the unchanged version as well. The change in speed by Deltawave seems on the money, but then it applies a wrong offset. When I use other files I'm not having this problem. I've been trimming one or two seconds off the front which helps some. But still the wrong offset is a persistent problem. And always keep in mind: Cognitive biases, like seeing optical illusions are a sign of a normally functioning brain. We all have them, it’s nothing to be ashamed about, but it is something that affects our objective evaluation of reality. Link to comment
esldude Posted May 25, 2019 Share Posted May 25, 2019 37 minutes ago, pkane2001 said: Was this with all the files, or a specific one? Can you please post the Results text from the comparison where this problem occurs? It occurs with Wild World and Le Mal de Vivre. Not with the others. It happens any time you compare files from device A or B with files from device C or D. The difference being those pairs of devices are off by 30 ppm in speed. I'll run a couple and post results in a minute. And always keep in mind: Cognitive biases, like seeing optical illusions are a sign of a normally functioning brain. We all have them, it’s nothing to be ashamed about, but it is something that affects our objective evaluation of reality. Link to comment
esldude Posted May 25, 2019 Share Posted May 25, 2019 16 minutes ago, fas42 said: Just to echo Dennis's findings here - manual correction makes a better fist of things so many times, and it's the offset that is quite obviously out, just from eyeballing in the Matched panel. Merely by adjusting by the displacement observed in the visuals, a far improved alignment is achieved - this should be detectable by the software, and used for fine tuning matching. On another note, have you had any thoughts on trying to compensate for non-uniform group delay, Paul? I don't know what about these files are causing this, but I don't think a single comparison I've made of Arch's files couldn't be improved up, and it was like Frank by looking at the Matched panel and adjust sample offset. I think the closest pair was off by a half sample. DW did report less than excellent matching, but the report was obviously off. Most of them required 1.5 to 2.5 samples adjustment to get fairly close. While DW was reporting matching of a few microseconds. And always keep in mind: Cognitive biases, like seeing optical illusions are a sign of a normally functioning brain. We all have them, it’s nothing to be ashamed about, but it is something that affects our objective evaluation of reality. Link to comment
esldude Posted May 25, 2019 Share Posted May 25, 2019 DeltaWave v1.0.30, 2019-05-25T17:39:22.1212785-05:00 Reference: D - Wild World.flac[L] 8632320 samples 96000Hz 24bits, stereo, MD5=00 Comparison: A - Wild World.flac[L] 8603648 samples 96000Hz 24bits, stereo, MD5=00 Settings: Gain:True, Remove DC:True Non-linear Gain:False EQ FFT Size:65536, EQ Frequency Cut: 0Hz - 0Hz, EQ Threshold: -144dB Correct Drift:True, Precision:30 Upsample:False, Window:Hann Spectrum Window:BlackmanHarris, Spectrum Size:65536 Spectrogram Window:BlackmanHarris, Spectrogram Size:4096, Spectrogram Steps:4096 Dither:False Trim Silence:False Discarding Reference: Start=1s, End=0s Discarding Comparison: Start=1s, End=0s Initial peak values Reference: -3.524dB Comparison: -3.675dB Initial RMS values Reference: -21.166dB Comparison: -21.228dB Null Depth=13.249dB X-Correlation offset: 19436 samples Drift computation quality, #1: Very Good (3.2μs) Trimmed 0 samples ( 0.00ms) front, 0 samples ( 0.00ms end) Final peak values Reference: -3.524dB Comparison: -3.696dB Final RMS values Reference: -21.152dB Comparison: -21.527dB Gain= -0.0141dB (0.9984x) DC=0 Phase offset=202.42604ms (19432.9 samples) Difference (rms) = -40.1dB [-43.41dBA] Correlated Null Depth=49.94dB [46.49dBA] Clock drift: 34.04 ppm Files are NOT a bit-perfect match (match=0.39%) at 16 bits Files are NOT a bit-perfect match (match=0%) at 24 bits Files match @ 49.8314% when reduced to 7.9 bits Phase difference (full bandwidth): 56.9339737829184° 0-10,000Hz: 51.2363600701568° 0-20,000Hz: 46.3402721189724° 0-24,000Hz: 42.7987677217951° 0-44,100Hz: 49.1013869160931° 0-48,000Hz: 56.9339737829184° RMS of the difference of spectra: -85.9040973299296dB gn=1.00162980674891, dc=0, dr=3.4043283E-05, of=19432.8998709808 DONE! Signature: 8f9775968324c5ea0800c85c7cd86010 -------------------------------------------------------------------------------------------------------------------------------- Here are the results page after I slowed the compare file by 30 ppm in Audacity. DeltaWave v1.0.30, 2019-05-25T17:43:24.6607925-05:00 Reference: D - Wild World.flac[L] 8632320 samples 96000Hz 24bits, stereo, MD5=00 Comparison: A - Wild World 30 ppm slo.wav[L] 8607744 samples 96000Hz 24bits, stereo, MD5=00 Settings: Gain:True, Remove DC:True Non-linear Gain:False EQ FFT Size:65536, EQ Frequency Cut: 0Hz - 0Hz, EQ Threshold: -144dB Correct Drift:True, Precision:30 Upsample:False, Window:Hann Spectrum Window:BlackmanHarris, Spectrum Size:65536 Spectrogram Window:BlackmanHarris, Spectrogram Size:4096, Spectrogram Steps:4096 Dither:False Trim Silence:False Discarding Reference: Start=1s, End=0s Discarding Comparison: Start=1s, End=0s Initial peak values Reference: -3.524dB Comparison: -3.676dB Initial RMS values Reference: -21.166dB Comparison: -21.23dB Null Depth=12.872dB X-Correlation offset: 19174 samples Drift computation quality, #1: Very Good (3.19μs) Trimmed 0 samples ( 0.00ms) front, 0 samples ( 0.00ms end) Final peak values Reference: -3.524dB Comparison: -3.619dB Final RMS values Reference: -21.154dB Comparison: -21.16dB Gain= -0.0705dB (0.9919x) DC=0 Phase offset=199.730718ms (19174.149 samples) Difference (rms) = -50dB [-51.37dBA] Correlated Null Depth=67.7dB [59.63dBA] Clock drift: 4.04 ppm Files are NOT a bit-perfect match (match=0.77%) at 16 bits Files are NOT a bit-perfect match (match=0%) at 24 bits Files match @ 50.0236% when reduced to 9.13 bits Phase difference (full bandwidth): 67.9870733957844° 0-10,000Hz: 47.0054070231397° 0-20,000Hz: 53.1728675007677° 0-24,000Hz: 51.0500276878744° 0-44,100Hz: 61.2338399403019° 0-48,000Hz: 67.9870733957844° RMS of the difference of spectra: -99.1337192307866dB gn=1.00815142657064, dc=-5.03102378887611E-07, dr=4.043046E-06, of=19174.1489148799 DONE! Signature: cffdb2191784694eb0b488d37613666a And always keep in mind: Cognitive biases, like seeing optical illusions are a sign of a normally functioning brain. We all have them, it’s nothing to be ashamed about, but it is something that affects our objective evaluation of reality. Link to comment
esldude Posted May 25, 2019 Share Posted May 25, 2019 DeltaWave v1.0.30, 2019-05-25T17:49:47.5576642-05:00 Reference: D - Le Mal de Vivre.flac[L] 11372544 samples 96000Hz 24bits, stereo, MD5=00 Comparison: A - Le Mal de Vivre.flac[L] 11345920 samples 96000Hz 24bits, stereo, MD5=00 Settings: Gain:True, Remove DC:True Non-linear Gain:False EQ FFT Size:65536, EQ Frequency Cut: 0Hz - 0Hz, EQ Threshold: -144dB Correct Drift:True, Precision:30 Upsample:False, Window:Hann Spectrum Window:BlackmanHarris, Spectrum Size:65536 Spectrogram Window:BlackmanHarris, Spectrogram Size:4096, Spectrogram Steps:4096 Dither:False Trim Silence:False Discarding Reference: Start=1s, End=0s Discarding Comparison: Start=1s, End=0s Initial peak values Reference: -5.452dB Comparison: -5.533dB Initial RMS values Reference: -27.759dB Comparison: -27.834dB Null Depth=12.969dB X-Correlation offset: 2243 samples Drift computation quality, #1: Excellent (1.94μs) Trimmed 0 samples ( 0.00ms) front, 0 samples ( 0.00ms end) Final peak values Reference: -5.452dB Comparison: -6.059dB Final RMS values Reference: -27.753dB Comparison: -28.508dB Gain= -0.0216dB (0.9975x) DC=0 Phase offset=23.360081ms (2242.568 samples) Difference (rms) = -41.87dB [-45.68dBA] Correlated Null Depth=49.56dB [47.91dBA] Clock drift: 33.43 ppm Files are NOT a bit-perfect match (match=1.51%) at 16 bits Files are NOT a bit-perfect match (match=0.01%) at 24 bits Files match @ 49.9952% when reduced to 9.77 bits Phase difference (full bandwidth): 14.325946548039° 0-10,000Hz: 30.9893568280198° 0-20,000Hz: 21.9136267540084° 0-24,000Hz: 20.0043175321093° 0-44,100Hz: 15.4674957741854° 0-48,000Hz: 14.325946548039° RMS of the difference of spectra: -90.7331214410073dB gn=1.00248986443164, dc=0, dr=3.3427834E-05, of=2242.5677322559 DONE! Signature: 576ee699452108901c278f3cae331ded ------------------------------------------------------------------------------------------------------------------------------------------------------------- And after the compare file is slowed 30 ppm in Audacity. DeltaWave v1.0.30, 2019-05-25T17:52:04.1875662-05:00 Reference: D - Le Mal de Vivre.flac[L] 11372544 samples 96000Hz 24bits, stereo, MD5=00 Comparison: A - Le Mal de Vivre 30 ppm slo.wav[L] 11347968 samples 96000Hz 24bits, stereo, MD5=00 Settings: Gain:True, Remove DC:True Non-linear Gain:False EQ FFT Size:65536, EQ Frequency Cut: 0Hz - 0Hz, EQ Threshold: -144dB Correct Drift:True, Precision:30 Upsample:False, Window:Hann Spectrum Window:BlackmanHarris, Spectrum Size:65536 Spectrogram Window:BlackmanHarris, Spectrogram Size:4096, Spectrogram Steps:4096 Dither:False Trim Silence:False Discarding Reference: Start=1s, End=0s Discarding Comparison: Start=1s, End=0s Initial peak values Reference: -5.452dB Comparison: -5.533dB Initial RMS values Reference: -27.759dB Comparison: -27.834dB Null Depth=12.37dB X-Correlation offset: 1983 samples Drift computation quality, #1: Excellent (1.96μs) Trimmed 0 samples ( 0.00ms) front, 0 samples ( 0.00ms end) Final peak values Reference: -5.452dB Comparison: -5.458dB Final RMS values Reference: -27.754dB Comparison: -27.754dB Gain= -0.0814dB (0.9907x) DC=0 Phase offset=20.664762ms (1983.817 samples) Difference (rms) = -67.71dB [-70.66dBA] Correlated Null Depth=62.25dB [67.05dBA] Clock drift: 3.43 ppm Files are NOT a bit-perfect match (match=2.91%) at 16 bits Files are NOT a bit-perfect match (match=0.01%) at 24 bits Files match @ 50.0138% when reduced to 11.55 bits Phase difference (full bandwidth): 16.384616820212° 0-10,000Hz: 32.908195994264° 0-20,000Hz: 23.2704649991824° 0-24,000Hz: 21.2429381052507° 0-44,100Hz: 16.5065687228965° 0-48,000Hz: 16.384616820212° RMS of the difference of spectra: -107.768135993824dB gn=1.00942107360986, dc=-7.40238420226983E-08, dr=3.427331E-06, of=1983.8171867867 DONE! Signature: 8eecca28795323158082ffd08726549e pkane2001 1 And always keep in mind: Cognitive biases, like seeing optical illusions are a sign of a normally functioning brain. We all have them, it’s nothing to be ashamed about, but it is something that affects our objective evaluation of reality. Link to comment
Popular Post esldude Posted May 25, 2019 Popular Post Share Posted May 25, 2019 BTW, if you work with device C on Wide World, there was a sudden change in speed difference after the 1:18 mark. So you need to chop off the last 20 seconds at least for it to make any sense. As far as I can tell that is the only file with that issue. And it would be interesting to know how that happened. Of course without DW we'd not know that. fas42 and pkane2001 1 1 And always keep in mind: Cognitive biases, like seeing optical illusions are a sign of a normally functioning brain. We all have them, it’s nothing to be ashamed about, but it is something that affects our objective evaluation of reality. Link to comment
esldude Posted May 27, 2019 Share Posted May 27, 2019 On 5/25/2019 at 8:14 PM, pkane2001 said: Dennis, can you please give me an example of two files where this extra offset is necessary? (or Frank @fas42)? I'd like to make sure I can reproduce what you are seeing. Also, I'm getting a significantly different result than you on the Archimago's files. For example, you posted: For the same two files, I get this: Difference (rms) = -57.51dB [-61.76dBA] Correlated Null Depth=62.68dB [66.65dBA] Not sure what’s different, as I tried to use the same settings as you. And I think I do have the fix for the oscillations you noticed past the sharp filter cutoff. I'll post this in a bit: I don't know what the difference is. I get near your results if I use the 30 ppm slowed file for compare. Nowhere close for the non speed corrected file. I'm including the results page here and checked MD5 checksum. So you can see if we indeed have the same files or not. Also will include a screen shot. I do noticed the MS of offset is very slightly different between yours and mine. DeltaWave v1.0.30, 2019-05-26T20:39:03.6116467-05:00 Reference: D - Crowd Chant.flac[L] 8747008 samples 96000Hz 24bits, stereo, MD5=20e71bf32010376c2b91b42cae62ece4 Comparison: A - Crowd Chant.flac[L] 8761344 samples 96000Hz 24bits, stereo, MD5=f1cdf415085ed847759e4b502a4cf2e2 Settings: Gain:True, Remove DC:True Non-linear Gain:False EQ FFT Size:65536, EQ Frequency Cut: 0Hz - 0Hz, EQ Threshold: -144dB Correct Drift:True, Precision:30 Upsample:False, Window:Hann Spectrum Window:BlackmanHarris, Spectrum Size:65536 Spectrogram Window:BlackmanHarris, Spectrogram Size:4096, Spectrogram Steps:4096 Dither:False Trim Silence:False Discarding Reference: Start=0s, End=0s Discarding Comparison: Start=0s, End=0s Initial peak values Reference: -3.167dB Comparison: -3.441dB Initial RMS values Reference: -16.38dB Comparison: -16.473dB Null Depth=13.067dB X-Correlation offset: -14935 samples Drift computation quality, #1: Excellent (0.62μs) Trimmed 0 samples ( 0.00ms) front, 0 samples ( 0.00ms end) Final peak values Reference: -3.167dB Comparison: -3.418dB Final RMS values Reference: -16.38dB Comparison: -16.725dB Gain= -0.0179dB (0.9979x) DC=0 Phase offset=-155.671577ms (-14944.471 samples) Difference (rms) = -36.46dB [-39.58dBA] Correlated Null Depth=54.44dB [82.79dBA] Clock drift: 33.29 ppm Files are NOT a bit-perfect match (match=0.41%) at 16 bits Files are NOT a bit-perfect match (match=0%) at 24 bits Files match @ 49.9917% when reduced to 7.94 bits Phase difference (full bandwidth): 43.0735971347464° 0-10,000Hz: 64.629845836155° 0-20,000Hz: 45.7120284473142° 0-24,000Hz: 41.7291928068088° 0-44,100Hz: 34.5564258887453° 0-48,000Hz: 43.0735971347464° RMS of the difference of spectra: -84.2579577220821dB gn=1.00206204776414, dc=0, dr=3.3289789E-05, of=-14944.4713477928 DONE! Signature: 7dc1e64a6ac1e672c4bda03833c02735 pkane2001 1 And always keep in mind: Cognitive biases, like seeing optical illusions are a sign of a normally functioning brain. We all have them, it’s nothing to be ashamed about, but it is something that affects our objective evaluation of reality. Link to comment
esldude Posted May 27, 2019 Share Posted May 27, 2019 Results using speed adjusted compare file. I also used auto trim which improved results about 3 db. DeltaWave v1.0.30, 2019-05-26T21:27:04.0694268-05:00 Reference: D - Crowd Chant.flac[L] 8747008 samples 96000Hz 24bits, stereo, MD5=20e71bf32010376c2b91b42cae62ece4 Comparison: A - Crowd Chant 30 ppm slo.wav[L] 8762827 samples 96000Hz 24bits, stereo, MD5=0915995c80b41fecc81190db862eed3a Settings: Gain:True, Remove DC:True Non-linear Gain:False EQ FFT Size:65536, EQ Frequency Cut: 0Hz - 0Hz, EQ Threshold: -144dB Correct Drift:True, Precision:30 Upsample:False, Window:Hann Spectrum Window:BlackmanHarris, Spectrum Size:16384 Spectrogram Window:BlackmanHarris, Spectrogram Size:4096, Spectrogram Steps:4096 Dither:False Trim Silence:True Discarding Reference: Start=0s, End=0s Discarding Comparison: Start=0s, End=0s Initial peak values Reference: -3.167dB Comparison: -3.436dB Initial RMS values Reference: -16.38dB Comparison: -16.473dB Null Depth=13.742dB X-Correlation offset: -14944 samples Trimming 0 samples at start and 1 samples at the end that are below -90.31dB level Drift computation quality, #1: Excellent (0.68μs) Trimmed 93177 samples ( 970.59375ms) front, 0 samples ( 0.00ms end) Final peak values Reference: -3.167dB Comparison: -3.351dB Final RMS values Reference: -16.357dB Comparison: -16.357dB Gain= -0.085dB (0.9903x) DC=0 Phase offset=-155.672959ms (-14944.604 samples) Difference (rms) = -57.14dB [-60.96dBA] Correlated Null Depth=65.23dB [68.82dBA] Clock drift: 3.29 ppm Files are NOT a bit-perfect match (match=0.78%) at 16 bits Files are NOT a bit-perfect match (match=0%) at 24 bits Files match @ 50.0175% when reduced to 9.43 bits Phase difference (full bandwidth): 33.6468912334159° 0-10,000Hz: 65.4363632952387° 0-20,000Hz: 46.2863545911077° 0-24,000Hz: 42.2526792438908° 0-44,100Hz: 33.2056720476031° 0-48,000Hz: 33.6468912334159° RMS of the difference of spectra: -87.86810016035dB gn=1.00983540437225, dc=6.14371468395743E-07, dr=3.286429E-06, of=-14944.6040518008 DONE! Signature: 395c485f162ef9a499460e7ef79c8d79 And always keep in mind: Cognitive biases, like seeing optical illusions are a sign of a normally functioning brain. We all have them, it’s nothing to be ashamed about, but it is something that affects our objective evaluation of reality. Link to comment
esldude Posted May 27, 2019 Share Posted May 27, 2019 Here is a nice little bug I just noticed. Unrelated to the other things in Arch's files I assume. I just ran a match and will show first Original spectra and then Matched spectra. They are slightly different plus the noise up high in the Matched spectrum. Now I do one thing. I check the microseconds box. The Original spectra now shows the Matched spectra. Unchecking microseconds or swapping around doesn't change that. Original Spectra shows Matched spectra until you run a new match. And always keep in mind: Cognitive biases, like seeing optical illusions are a sign of a normally functioning brain. We all have them, it’s nothing to be ashamed about, but it is something that affects our objective evaluation of reality. Link to comment
esldude Posted May 27, 2019 Share Posted May 27, 2019 Related to my last post about mixing up Original and Matched spectra. If you run the match with the microseconds box checked. Everything is fine. If you then uncheck microseconds, it also begins showing the Matched spectra in both Matched and Original spectra. So the bug shows up if you change the microseconds check box no matter which way you approach it. And always keep in mind: Cognitive biases, like seeing optical illusions are a sign of a normally functioning brain. We all have them, it’s nothing to be ashamed about, but it is something that affects our objective evaluation of reality. Link to comment
esldude Posted May 28, 2019 Share Posted May 28, 2019 Now that I've had sufficient BBQ, I've a chance to try this new version out. Looks to be right as rain. Everything looks good so far. Much better results, no weirdness. Offset seems dead on. I can change the offset by one more or one less on the smallest decimal and I get identical results and they are less good than the default. Makes me think offset if dead on. Same for drift rate. Also the results are very nearly identical whether on files that differ by 34 ppm or one I've slowed by 30 ppm. And showing both file paths in the Recent analysis tab is much nicer. Thanks. pkane2001 1 And always keep in mind: Cognitive biases, like seeing optical illusions are a sign of a normally functioning brain. We all have them, it’s nothing to be ashamed about, but it is something that affects our objective evaluation of reality. Link to comment
esldude Posted May 29, 2019 Share Posted May 29, 2019 4 hours ago, fas42 said: Sorry I didn't get back earlier, Paul; busy sorting out a fallen tree, . Haven't given it a thorough run through, but like Dennis, all looks good so far - Recent Analysis layout is fine. Was wondering how long that tree would take you. Glad it wasn't multiple days. Had that problem each of the last two years. Amazing how much wood is in a moderate sized tree once it hits the ground. pkane2001 1 And always keep in mind: Cognitive biases, like seeing optical illusions are a sign of a normally functioning brain. We all have them, it’s nothing to be ashamed about, but it is something that affects our objective evaluation of reality. Link to comment
esldude Posted June 11, 2019 Share Posted June 11, 2019 6 hours ago, pkane2001 said: I tried a variation on the non-linear EQ that also adjusts phase, not just amplitude. Seems promising. Here's the result of matching the two files with non-linear EQ engaged, as it exists in the current .31 version: And here's the result using the amplitude and phase EQ (feature coming in the next version). Seems to produce an 8-9dB better null with phase correction: I often try out the non-linear EQ function to see if the results will be improved. Sometimes they are, sometimes not by much. But I think these settings are a bit out of character for DeltaWave. They compensate for frequency and phase errors that really are true, complex, and possibly audible, errors. While I can certainly add more non-linear correction functions (and I'll do that, just for fun ), I see these as not being useful for the main purpose of what DeltaWave is designed to do: finding differences between two waveforms. Archimago's files are a bit of a problem. I think I've been able to improve all with some manual adjustments. I understand you saying adding non-linear correction isn't the purpose of Deltawave which is to find differences in the waveform. Or as we've said interesting differences. Timing and gain are uninteresting differences. However, the benefit of including them is when you have a poor null it might help you determine why the null isn't better. You do that by correcting and improving it. If EQ does it then you know the main source of the error. If it is phase, then you know it is the source of error. With Archimago's files I get the best null with the Asrock and Sony. The others aren't very good, and it isn't obvious why not. It seems to me when the nulls are poor the timing is sometimes the issue. The poor matching seems to interfere with Deltawave matching the timing as well as when nulls are pretty good. I seem to remember last week I managed to manually improve one of the Archimago comparisons by 11 db with timing adjustments. What would be the issue that would most interfere with Deltawave doing better with the timing in such cases? Timing drift which isn't steady? And always keep in mind: Cognitive biases, like seeing optical illusions are a sign of a normally functioning brain. We all have them, it’s nothing to be ashamed about, but it is something that affects our objective evaluation of reality. Link to comment
esldude Posted June 11, 2019 Share Posted June 11, 2019 11 hours ago, pkane2001 said: The EQ option is most definitely not linear It is linear per frequency bin, but each frequency is adjusted independently of the others, so in effect, this is forcing a correction just like a very fine graphic equalizer would. That reminds me. Knowing it works as you describe if I find it helps some, I do it two or three ways with different size FFT's. I've run across a comparison a couple times that was helped by a rather small FFT in the EQ several decibels more than a larger FFT. And I've seen the reverse. I should have kept up with it, one pair of files was 11 db better at 16 k and only a couple db at any other setting size for the FFT. I do agree with you it is normal for 19 out of 20 comparisons EQ makes no difference or only a small one of 2 db or less. Generally FFT on the smaller side for the EQ seems more likely to help just a little in the files I've been using. pkane2001 1 And always keep in mind: Cognitive biases, like seeing optical illusions are a sign of a normally functioning brain. We all have them, it’s nothing to be ashamed about, but it is something that affects our objective evaluation of reality. Link to comment
esldude Posted June 11, 2019 Share Posted June 11, 2019 10 hours ago, pkane2001 said: Here are some timing error plots that are corrected with non-linear clock drift feature (coming in the next version). First, clock drift after linear correction as it works today: And now, the same exact comparison but with non-linear drift correction engaged: Similar result, using a DAC with resampling turned on and off inside the DAC. First, without non-linear drift correction: Same comparison, but now with non-linear drift correction engaged: Yet I notice in these plots the difference and null depth didn't budge. And always keep in mind: Cognitive biases, like seeing optical illusions are a sign of a normally functioning brain. We all have them, it’s nothing to be ashamed about, but it is something that affects our objective evaluation of reality. Link to comment
esldude Posted June 25, 2019 Share Posted June 25, 2019 @pkane2001 I have a feature request if it isn't too terrible to do. Could we have a setting that allows us to set the lower reach of the spectrograms? Unchecked it does as it currently does, or if we get to set it then it only reaches as low as we want. I've run into a couple instances where going to -200 db for instance lets the noise floor somewhat obscure details of what is happening. If I could pick a higher bottom to the spectrogram and it ignore everything lower some of the things I'm interested in would be clearer to see. You might let us pick -75, -100, 125 and -150 db. That would be workable most of the time. Or if we can just pick any number even better. Of course I don't know if this is a big task to add or not. And always keep in mind: Cognitive biases, like seeing optical illusions are a sign of a normally functioning brain. We all have them, it’s nothing to be ashamed about, but it is something that affects our objective evaluation of reality. Link to comment
esldude Posted June 26, 2019 Share Posted June 26, 2019 4 hours ago, pkane2001 said: So the ability to zoom in to an arbitrary level doesn’t work for this? Do you think a minimum and a maximum might be needed, not just a minimum? Zooming in is often quite useful, but in some cases it isn't enough. Or it isn't best. Audacity lets you set the low end of the spectrogram, and I find it useful. I really wish Audacity would also let you zoom in. Well actually Audacity does let you zoom in. I forgot it as its not exactly convenient the way it works. I've given no thought to whether it would be helpful to set the maximum as well as the minimum. Yeah, I think that might be useful too. Might let you see some of what is happening when you are interested in a lower level rather close to a higher level portion. So is this a big chore to add? It may be something almost no one else cares about or finds useful. I don't want to clutter the software if that is the case. And always keep in mind: Cognitive biases, like seeing optical illusions are a sign of a normally functioning brain. We all have them, it’s nothing to be ashamed about, but it is something that affects our objective evaluation of reality. Link to comment
esldude Posted June 26, 2019 Share Posted June 26, 2019 7 hours ago, pkane2001 said: Shouldn't be hard to add, just need to think of a good place to put the settings. Screens are really crowded as is Well what about the pane where we select FFT size, windowing type and it also has the setting for phase. That has some space. You are doing too good a job adding features maybe. And always keep in mind: Cognitive biases, like seeing optical illusions are a sign of a normally functioning brain. We all have them, it’s nothing to be ashamed about, but it is something that affects our objective evaluation of reality. Link to comment
Popular Post esldude Posted June 28, 2019 Popular Post Share Posted June 28, 2019 4 hours ago, pkane2001 said: Is this something like what you're looking for (min=-110dB, max=-60dB)? No that is not it. Sorry for the misunderstanding. I'm referring to spectrogram 1 and 2 the ones with all the pretty colors. I'll add a screenshot in a minute. This particular one isn't a good example of why it would be useful. I'd like for the lower reaches of the spectrogram to be something I could choose. That can sometimes filter out noise and let you see something otherwise obscured. It also would expand the scale. Like if I set it to -100 db it ignores lower data and expands the scale over a narrower range. rando and pkane2001 2 And always keep in mind: Cognitive biases, like seeing optical illusions are a sign of a normally functioning brain. We all have them, it’s nothing to be ashamed about, but it is something that affects our objective evaluation of reality. Link to comment
esldude Posted July 5, 2019 Share Posted July 5, 2019 6 minutes ago, pkane2001 said: @esldude Dennis, the frequency range setting for spectrograms will need to be in a later release. It's not as simple as I had hoped, as it requires increasing frequency resolution when zooming in, meaning that it will require larger and larger FFTs to be performed. Considering that there are many thousands of FFTs being calculated for each spectrogram, this will slow things down considerably. I'll do some more thinking to see if there's a better way to do this. For now, you can simply chose a larger size FFT in the spectrogram settings to see just how slow it may become. Once you do, this will also let you zoom-in further without losing as much resolution. Thanks for looking at this Paul. And don't worry about it further. It is a useful difference only very rarely. And I can export the files from Deltawave to look at with other software on those rare occasions. So I wouldn't bother with it. What your software does already is something quite special in its overall capabilities. pkane2001 1 And always keep in mind: Cognitive biases, like seeing optical illusions are a sign of a normally functioning brain. We all have them, it’s nothing to be ashamed about, but it is something that affects our objective evaluation of reality. Link to comment
esldude Posted July 5, 2019 Share Posted July 5, 2019 21 minutes ago, pkane2001 said: But if you do need to zoom in, just increase FFT size in settings. Here's a 96kHz file zoomed in on lower 21kHz, FFT size at 8192: The reason that sometimes won't do is because shorter term events get missed by large FFT values. It uses a larger number of samples to generate the FFT and something short enough is no longer resolvable. Lowering the FFT values will catch it, but raises the noise floor so sometimes you obscure what you are looking for. It only comes up very rarely, and only would apply to short events just above, but not too far above the noise floor. So like I said, it is a rare and peculiar thing. So not worth you spending extended time or effort to address because it has such marginal utility. And always keep in mind: Cognitive biases, like seeing optical illusions are a sign of a normally functioning brain. We all have them, it’s nothing to be ashamed about, but it is something that affects our objective evaluation of reality. Link to comment
esldude Posted July 6, 2019 Share Posted July 6, 2019 10 hours ago, pkane2001 said: This is a more fundamental limitation of the Fourier transform. Higher resolution in the frequency domain results in lower resolution in the time domain. Wavelet transform is one way around it, but I'm not quite ready to rewrite most of the algorithms in DW. Maybe in version 2 No, I'm not suggesting any such re-write. If doing so results in some greatly improved results then do it for that reason. Otherwise I don't suggest changing how DW is working at its core right now. And always keep in mind: Cognitive biases, like seeing optical illusions are a sign of a normally functioning brain. We all have them, it’s nothing to be ashamed about, but it is something that affects our objective evaluation of reality. Link to comment
Popular Post esldude Posted July 20, 2019 Popular Post Share Posted July 20, 2019 Haven't had time to do much with the new version until this afternoon. 1.37 looks like a jump in performance. On some of my 8th generation copies vs the original I still get mid 50 db results with default settings. If I engage level EQ, phase EQ and non-linear drift correction the results improve 25 db or so. The residual sounds a bit odd. With 70 db of gain you hear bell like tones and some here and there echoes of the music. Using the loopback file in the loopback thread over at gearslutz, my loopback measure on my Zen Tour goes from 56 db difference to 84 db with the extra functions enabled. Of course un-selecting one item at a time lets me see which makes the most difference. The extra phase graphs are very useful too. I get similar improvements with other loopbacks that are compared. On comparisons of cables the extra compensation makes little to no difference or a slight negative. That is what you would expect. So I would say you have it working much better, able to compensate more precisely for more issues which would let one see what really is causing differences. fas42 and pkane2001 2 And always keep in mind: Cognitive biases, like seeing optical illusions are a sign of a normally functioning brain. We all have them, it’s nothing to be ashamed about, but it is something that affects our objective evaluation of reality. Link to comment
esldude Posted July 21, 2019 Share Posted July 21, 2019 2 hours ago, pkane2001 said: The lowish null values with some of the better DACs/ADCs on the Gearslutz thread were really bothering me. Seems that phase differences accounted for a large portion of them. The phase plot shows the accuracy of the recorded timing much more intimately than a single jitter value. You can see large jitter effects in a chaotically jumping phase result. In some of the better converters, the phase is clean, and linear, with very little variation. Others change as a sinewave, or an exponential curve. I'm sure that most are caused by either ADC or DAC filters (possibly by both in a loop-back recording). I doubt that any of the phase differences of less than a few degrees would cause audible distortions, but a cleaner, more linear phase response certainly speaks to a more accurate sound reproduction, audible or not I found some studies on the audibility of phase, but these were mostly done with single tones, so it's not clear how these translate into non-linear phase distortions of various shapes and sizes, as applied to the whole spectrum of a complex musical passage. The results seem to indicate something I've seen hints of anyway. Unlike what you might expect most of these loopback values seem effected by the low frequency roll off involved. It seems to alter phase enough to give trouble for Diffmaker to match up, and previously Deltawave. What you have it doing now is a big step in the right direction. pkane2001 1 And always keep in mind: Cognitive biases, like seeing optical illusions are a sign of a normally functioning brain. We all have them, it’s nothing to be ashamed about, but it is something that affects our objective evaluation of reality. Link to comment
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now