Jump to content
pkane2001

DeltaWave null-testing audio comparator (beta)

Rate this topic

Recommended Posts

23 minutes ago, modmix said:

Comparing the same file gives "NOT Bit Perfect" with Non-Linear Calibration either EQ selected (see attached image).

No Non-Linear Calibration or Non-linear Drift Correction results in "BIT PERFECT!"

Any explanation?
TIA
Ulli

DeltaWave_Original2_Level EQ.jpg

DeltaWave_Original2_no Non-linear Calibration.jpg

 

Non-linear matching performs multiple FFTs, averages the results, and then applies the difference to the comparison waveform through another FFT. In the process, a little precision is lost resulting in a small mismatch between the reference and comparison (no processing is done on the reference file). But, the difference you reported seems way too large for a tiny loss of precision... :) 

 

I'm about to post a new version that might help with this. Please check it and let me know what you find.

Share this post


Link to post
Share on other sites

Paul, I'm getting some gobbledygook results with 1.0.37 - with the same set of files, and settings as in the post above, I still didn't see what I was hoping to see. Then I noted in the Aligned Spectrum tab that there was a huge displacement between the Reference and Comparison slopes, that prior to what I was after, the beyond 15k signal. Ok, what happens if I reverse the Ref and Comp waves in the processing - which I did by copying and pasting the file paths in the input boxes, swapping them as far as the interface was concerned. But what I got from then matching were quite bizarre graphs, which didn't seem to relate to anything sensible!

 

Can you do this at your end, with everything behaving itself?


Frank

 

http://artofaudioconjuring.blogspot.com/

 

 

Ahhh, Mankind ... Porsche intellect, Trabant emotions ...

Share this post


Link to post
Share on other sites

Level EQ now gives:

  • Files are NOT a bit-perfect match (match=100%) at 16 bits
    Files are NOT a bit-perfect match (match=100%) at 24 bits
    Files match @ 100% when reduced to 12,5 bits
    ---- Phase difference (full bandwidth): 2,52173587518713E-11°
        0-10kHz: 0,00°
        0-20kHz: 0,00°
        0-24kHz: 0,00°
    Timing error (rms jitter): 0sec
    RMS of the difference of spectra: -353,388476384148dB
    gn=1, dc=0, dr=0, of=0

Quite small difference, now.
But: 12,5 bits ?...

PS:
I really like your program 🙂
Looking forward to see further improvements.

Share this post


Link to post
Share on other sites
2 minutes ago, modmix said:

Level EQ now gives:

  • Files are NOT a bit-perfect match (match=100%) at 16 bits
    Files are NOT a bit-perfect match (match=100%) at 24 bits
    Files match @ 100% when reduced to 12,5 bits
    ---- Phase difference (full bandwidth): 2,52173587518713E-11°
        0-10kHz: 0,00°
        0-20kHz: 0,00°
        0-24kHz: 0,00°
    Timing error (rms jitter): 0sec
    RMS of the difference of spectra: -353,388476384148dB
    gn=1, dc=0, dr=0, of=0

Quite small difference, now.
But: 12,5 bits ?...

PS:
I really like your program 🙂
Looking forward to see further improvements.

 

Yes, the 50% bit match computation is confused by the fact that it needs to increase bit size instead of decreasing it! I thought I had fixed this a long time ago, but looks like the issue is back.

 

PS: glad you like it!

Share this post


Link to post
Share on other sites
8 minutes ago, fas42 said:

Paul, I'm getting some gobbledygook results with 1.0.37 - with the same set of files, and settings as in the post above, I still didn't see what I was hoping to see. Then I noted in the Aligned Spectrum tab that there was a huge displacement between the Reference and Comparison slopes, that prior to what I was after, the beyond 15k signal. Ok, what happens if I reverse the Ref and Comp waves in the processing - which I did by copying and pasting the file paths in the input boxes, swapping them as far as the interface was concerned. But what I got from then matching were quite bizarre graphs, which didn't seem to relate to anything sensible!

 

Can you do this at your end, with everything behaving itself?

 

Strange, let me check. It seemed to produce reasonable results when I tested it.

 

By the way there is a swap ref/compare files menu under the file menu that you can use :)

 

Share this post


Link to post
Share on other sites
22 minutes ago, fas42 said:

Paul, I'm getting some gobbledygook results with 1.0.37 - with the same set of files, and settings as in the post above, I still didn't see what I was hoping to see. Then I noted in the Aligned Spectrum tab that there was a huge displacement between the Reference and Comparison slopes, that prior to what I was after, the beyond 15k signal. Ok, what happens if I reverse the Ref and Comp waves in the processing - which I did by copying and pasting the file paths in the input boxes, swapping them as far as the interface was concerned. But what I got from then matching were quite bizarre graphs, which didn't seem to relate to anything sensible!

 

Can you do this at your end, with everything behaving itself?

 

Looks ok to me. Here's what I get (and this is using the 32-bit version):

 

image.thumb.png.00b9bf3a38129205c8daee512c6c20bb.png

 

And now, swapped:

 

image.thumb.png.48fbe40b6468db5b21cb172bdcf723ef.png

Share this post


Link to post
Share on other sites
42 minutes ago, pkane2001 said:

 

Looks ok to me. Here's what I get (and this is using the 32-bit version):

 

 

Hmmm ... I'm getting erratic behaviour - changing the length of the matched waveform gives me results which correspond to yours, and if I go back to the length where I was seeing filtering as if the 500Hz Filter 1 was not actioning, now it is working!

 

Also, swapping the Ref and Comp does behave itself now, whether I use the menu function, or manually move them.

 

Is DW getting into a status where it misbehaves, depending upon precisely what has occurred beforehand? I'll try to get a handle on this, restarting DW when something is abnormal, to see if I can pick a pattern ...again, ^_^.


Frank

 

http://artofaudioconjuring.blogspot.com/

 

 

Ahhh, Mankind ... Porsche intellect, Trabant emotions ...

Share this post


Link to post
Share on other sites

One thought I have just had, is that the memory cleaner utility I'm using is too 'aggressive' for DW, and the program is losing data which it expects not to be touched ... just an idea that popped into my head ....


Frank

 

http://artofaudioconjuring.blogspot.com/

 

 

Ahhh, Mankind ... Porsche intellect, Trabant emotions ...

Share this post


Link to post
Share on other sites
8 minutes ago, fas42 said:

 

Hmmm ... I'm getting erratic behaviour - changing the length of the matched waveform gives me results which correspond to yours, and if I go back to the length where I was seeing filtering as if the 500Hz Filter 1 was not actioning, now it is working!

 

Also, swapping the Ref and Comp does behave itself now, whether I use the menu function, or manually move them.

 

Is DW getting into a status where it misbehaves, depending upon precisely what has occurred beforehand? I'll try to get a handle on this, restarting DW when something is abnormal, to see if I can pick a pattern ...again, ^_^.

 

I assume you're still running the 32-bit version. There's a good chance of out of memory conditions that could happen randomly, even depending on what else is executing at the time, or possibly due to the prior comparisons before running the test that chewed up some of the available memory.

 

You can enable full logging in DW under Help->Logging->Debug. Once you do, try to reproduce the error. When you succeed,  open the log file by invoking Help->Logging->View Log... menu. Save it as a text file, or just post it as an attachment here, and I can take a look.

Share this post


Link to post
Share on other sites

You will be pleased to hear, Paul, that I'm back to 64 bit running :) - the HP laptop is back in operation, after cooking the graphics chip with a hair dryer, and am now giving the restored beast a processing thrashing by using DW64.

 

And noting a couple of things: the group delay curve has a discontinuity, jump in it when processing the Bob Marley samples - altering the FFT, say, adjusts this behaviour, but doesn't make it go away ... and could the locking and unlocking of  plot axes be tweaked, perhaps - sometimes it does exactly what you want; and at other times it stubbornly refuses to behave. An example of the positive behaviour is that if one zooms in on the Delta Waveform, and then locks an unlocked Matched, and unlocks it again, that the latter has also zoomed into precisely where you want it. The irksome behaviour is when locking and then unlocking causes the display to jump between 2 zoom settings, neither of which is right, and which then requires major fiddling to get to what one is actually after.


Frank

 

http://artofaudioconjuring.blogspot.com/

 

 

Ahhh, Mankind ... Porsche intellect, Trabant emotions ...

Share this post


Link to post
Share on other sites

@pkane2001

Hi Paul,

Please do you mind helping with the following questions ?

Print screens joined are dealing with 10 Hz/100 Hz pure tones sampled @10 kHz (3 million samples).

 

1 Manual Corrections:

I started using it since I was not satisfied with the automatic matching. Great feature indeed since I managed to correct phase offsets and drifts. Is it possible to customise the optimise feature? I mean a dichotomy for example, for the chosen parameter: Offset alone, Drift alone, etc. In my test cases the optimisation didn't work since it was acting at both offset and drift when it should not.

BTW correcting phase using zooming in original and matched waveforms is quite painful especially for tiny values where time scale is lost. 😉

2. Default Parameters: 

How to retrieve default parameters in Deltawave setup?

3. Spectrum artefacts:

In average mode there are artefacts increasing with frequency ? How can you explain it? With other audio spectrum software I can manage clean averaged -200 dB floor with 32 bits files, most probably due to better windowing.

The issue is that those artefacts are spoiling the phase and jitter accuracy results.

Spect1.thumb.jpg.93e4d47fc7a445539436799754ca7dcf.jpgSpect2.thumb.jpg.cca6b289737924d1e58829a99487b59c.jpg

4. 360° Phase jumps :

Is there a way to correct such 'fake' phase jumps? Purpose is same as above.

DW_PJ.thumb.jpg.1a7a563a18296a4e9ae58d54e7655aac.jpg 

Thanks.

 

Share this post


Link to post
Share on other sites
16 hours ago, Arpiben said:

@pkane2001

Hi Paul,

Please do you mind helping with the following questions ?

Print screens joined are dealing with 10 Hz/100 Hz pure tones sampled @10 kHz (3 million samples).

 

1 Manual Corrections:

I started using it since I was not satisfied with the automatic matching. Great feature indeed since I managed to correct phase offsets and drifts. Is it possible to customise the optimise feature? I mean a dichotomy for example, for the chosen parameter: Offset alone, Drift alone, etc. In my test cases the optimisation didn't work since it was acting at both offset and drift when it should not.

 

Optimize is just trying a gradient descent search for optimal parameters. I can certainly limit the number of parameters that it tries to optimize for and see if it works differently. You do have the choice of selecting the 'goal' for the optimization in the drop-down.

 

Quote

BTW correcting phase using zooming in original and matched waveforms is quite painful especially for tiny values where time scale is lost. 😉

 

If you click a few times on the 'μs' checkbox you'll get to a state where it is just a filled-in rectangle. In this mode it will display actual sample numbers on the time-axis, so you don't need to worry about nanoseconds display.

image.png.b11ef45778a9ef916c41834a758ce185.png

 

Quote

2. Default Parameters: 

How to retrieve default parameters in Deltawave setup?

 

I'll add an option to restore defaults -- wanted to do it for a while. For now, you just need to open the folder where DeltaWave stores settings and delete (or rename) the default settings file. To find the file, click on Help->Logging->Open Log Folder menu. In the folder, find the two _DeltaWaveDefault.* files and rename them/move them or delete them. This will restore all defaults to DW.

 

Quote

3. Spectrum artefacts:

In average mode there are artefacts increasing with frequency ? How can you explain it? With other audio spectrum software I can manage clean averaged -200 dB floor with 32 bits files, most probably due to better windowing.

The issue is that those artefacts are spoiling the phase and jitter accuracy results.

 

The calculations are done in 64-bit floating point format, window is applied per the selection under settings. FFT size chosen will also have an effect as it changes the resolution and the number of points being averaged. How was this waveform captured? ADC/DAC precision is rarely better than about 20-21 bits, so expecting 32-bit accuracy may be a bit optimistic :)

 

Can you share an example waveform and point me to another software that shows cleaner average that I could compare this to? There could be some precision issue in the libraries I'm using, I've already found a few such issues in the past.

Share this post


Link to post
Share on other sites
On 7/18/2019 at 11:10 PM, fas42 said:

You will be pleased to hear, Paul, that I'm back to 64 bit running :) - the HP laptop is back in operation, after cooking the graphics chip with a hair dryer, and am now giving the restored beast a processing thrashing by using DW64.

 

And noting a couple of things: the group delay curve has a discontinuity, jump in it when processing the Bob Marley samples - altering the FFT, say, adjusts this behaviour, but doesn't make it go away ... and could the locking and unlocking of  plot axes be tweaked, perhaps - sometimes it does exactly what you want; and at other times it stubbornly refuses to behave. An example of the positive behaviour is that if one zooms in on the Delta Waveform, and then locks an unlocked Matched, and unlocks it again, that the latter has also zoomed into precisely where you want it. The irksome behaviour is when locking and then unlocking causes the display to jump between 2 zoom settings, neither of which is right, and which then requires major fiddling to get to what one is actually after.

 

Glad to hear it, Frank!  :) You cooked the GPU and it started working? That's pretty amazing!

 

The phase discontinuity you're talking about occurs in which of the three phase curves? Can you post an example? Some discontinuity can occur if there's a 360 degree jump in phase calculation, similar to Arpiben's last plot (point 4). That is actually not a discontinuity, just an artifact of how phase is computed. I can see if it can be accounted for when unwrap phase option is not selected.

 

I'll try to reproduce the locking/unlocking behavior you are describing. Don't think I've noticed this before.

Share this post


Link to post
Share on other sites
1 hour ago, pkane2001 said:

 

Optimize is just trying a gradient descent search for optimal parameters. I can certainly limit the number of parameters that it tries to optimize for and see if it works differently. You do have the choice of selecting the 'goal' for the optimization in the drop-down.

 

 

If you click a few times on the 'μs' checkbox you'll get to a state where it is just a filled-in rectangle. In this mode it will display actual sample numbers on the time-axis, so you don't need to worry about nanoseconds display.

image.png.b11ef45778a9ef916c41834a758ce185.png

 

 

I'll add an option to restore defaults -- wanted to do it for a while. For now, you just need to open the folder where DeltaWave stores settings and delete (or rename) the default settings file. To find the file, click on Help->Logging->Open Log Folder menu. In the folder, find the two _DeltaWaveDefault.* files and rename them/move them or delete them. This will restore all defaults to DW.

 

 

The calculations are done in 64-bit floating point format, window is applied per the selection under settings. FFT size chosen will also have an effect as it changes the resolution and the number of points being averaged. How was this waveform captured? ADC/DAC precision is rarely better than about 20-21 bits, so expecting 32-bit accuracy may be a bit optimistic :)

 

Can you share an example waveform and point me to another software that shows cleaner average that I could compare this to? There could be some precision issue in the libraries I'm using, I've already found a few such issues in the past.

 

Hi Paul,

 

Yes we do have the choice to select the optimisation goal but we can not force optimisation to act only with a single variable parameter: offset only, drift only, Dc only....Keep in mind that I was looking at pure tones drifts and phase offsets which are not standard files vs audio ones.Therefore I am not complaining about how optimisation actually works.

 

The wave-forms were not captured but generated either by Audacity and saved in 32.bits floating wav or software (Scipy/Mathlab).

Generating a pure 100 Hz in Audacity ( amp=0.5 /100 kSps) is enough to show the spectrum averaging anomalies. 

 

I checked the same above files with MusicScope. You may use the free trial version limited to 30 seconds: https://www.xivero.com/musicscope/. The free version will not give you access to the UHR spectrum ( FFT= 65536 / 0.67Hz/Bin) but you are not losing so much. In principle, for spectrum they use Blackman Harris 7 term FFT windows.

 

With DW I tried different FFT sizes for spectrum with more or less same anomalies in averaged mode. I am expecting something flat with a noise floor level depending on window applied.

 

Well I do know that we are looking at the limits but I have the feeling that DeltaWave can perform even better 😊

 

Share this post


Link to post
Share on other sites
2 hours ago, Arpiben said:

 

Hi Paul,

 

Yes we do have the choice to select the optimisation goal but we can not force optimisation to act only with a single variable parameter: offset only, drift only, Dc only....Keep in mind that I was looking at pure tones drifts and phase offsets which are not standard files vs audio ones.Therefore I am not complaining about how optimisation actually works.

 

The wave-forms were not captured but generated either by Audacity and saved in 32.bits floating wav or software (Scipy/Mathlab).

Generating a pure 100 Hz in Audacity ( amp=0.5 /100 kSps) is enough to show the spectrum averaging anomalies. 

 

I checked the same above files with MusicScope. You may use the free trial version limited to 30 seconds: https://www.xivero.com/musicscope/. The free version will not give you access to the UHR spectrum ( FFT= 65536 / 0.67Hz/Bin) but you are not losing so much. In principle, for spectrum they use Blackman Harris 7 term FFT windows.

 

With DW I tried different FFT sizes for spectrum with more or less same anomalies in averaged mode. I am expecting something flat with a noise floor level depending on window applied.

 

Well I do know that we are looking at the limits but I have the feeling that DeltaWave can perform even better 😊

 

 

 

So I'm not sure if I'm duplicating what you did in Audacity... I generated a 100Hz /0.5 amplitude sine wave sampled at 100kHz. Then, exported it as a 24-bit integer WAV file and as a 32-bit floating point. You can see the result (blue is 24 bit). This seems like quantization noise to me. Spectrum chart setting was 64k FFT with Hann window:

image.thumb.png.5c37df534effcf1f7e0d3b64a9e9ed26.png

 

Same files, but with 512k FFT/Hann window spectrum settings:

image.thumb.png.e3630cb925af684762be5b33792c9f12.png

 

 

Share this post


Link to post
Share on other sites
3 hours ago, Arpiben said:

 

Hi Paul,

 

Yes we do have the choice to select the optimisation goal but we can not force optimisation to act only with a single variable parameter: offset only, drift only, Dc only....Keep in mind that I was looking at pure tones drifts and phase offsets which are not standard files vs audio ones.Therefore I am not complaining about how optimisation actually works.

 

The wave-forms were not captured but generated either by Audacity and saved in 32.bits floating wav or software (Scipy/Mathlab).

Generating a pure 100 Hz in Audacity ( amp=0.5 /100 kSps) is enough to show the spectrum averaging anomalies. 

 

I checked the same above files with MusicScope. You may use the free trial version limited to 30 seconds: https://www.xivero.com/musicscope/. The free version will not give you access to the UHR spectrum ( FFT= 65536 / 0.67Hz/Bin) but you are not losing so much. In principle, for spectrum they use Blackman Harris 7 term FFT windows.

 

With DW I tried different FFT sizes for spectrum with more or less same anomalies in averaged mode. I am expecting something flat with a noise floor level depending on window applied.

 

Well I do know that we are looking at the limits but I have the feeling that DeltaWave can perform even better 😊

 

 

Tried a trial version of MusicScope on 24-bit WAV file. Doesn't seem all that flat to me below -160dB. Are you seeing something different?

 

image.thumb.png.4a755f16f914114533d87445b37b1001.png

 

Share this post


Link to post
Share on other sites
40 minutes ago, pkane2001 said:

 

Tried a trial version of MusicScope on 24-bit WAV file. Doesn't seem all that flat to me below -160dB. Are you seeing something different?

 

image.thumb.png.4a755f16f914114533d87445b37b1001.png

 

 

😊 Paul, using the full rights you would have seen something like the above for the 32 bits version used in DW print screens: 

(I will later verify your 24 bit)

 

1. No average / No Peak mode:

1.thumb.jpg.b3195522c858ed5be4291ddf7a4d82d0.jpg

 

2. Averaged spectrum:

 

2.thumb.jpg.4691084e11945aabfd46d14b04c52b31.jpg

 

You will notice that we don't have here apparent amplitude artefacts increasing with frequency. With DW the amplitude is around 5-6dB close to Nyquist..

 

Rgds.

 

Share this post


Link to post
Share on other sites
1 hour ago, Arpiben said:

 

😊 Paul, using the full rights you would have seen something like the above for the 32 bits version used in DW print screens: 

(I will later verify your 24 bit)

 

1. No average / No Peak mode:

1.thumb.jpg.b3195522c858ed5be4291ddf7a4d82d0.jpg

 

2. Averaged spectrum:

 

2.thumb.jpg.4691084e11945aabfd46d14b04c52b31.jpg

 

You will notice that we don't have here apparent amplitude artefacts increasing with frequency. With DW the amplitude is around 5-6dB close to Nyquist..

 

Rgds.

 

 

5-6dB amplitude does seem like a lot! These appear a lot smaller (well under 1dB) with the 32-bit floating point files, which is also strange. Let me try additional windows and see if they make a difference. I thought that MathNet library I'm using already implemented Blackman-Harris 7 term window, but called it 'BlackmanHarris'. Maybe it's actually the 4-term variant, instead. 

 

Share this post


Link to post
Share on other sites
8 hours ago, pkane2001 said:

 

Glad to hear it, Frank!  :) You cooked the GPU and it started working? That's pretty amazing!

 

Yes, the full story starts here, https://audiophilestyle.com/forums/topic/49659-step-by-step-surgery/?do=findComment&comment=968975.

 

8 hours ago, pkane2001 said:

The phase discontinuity you're talking about occurs in which of the three phase curves? Can you post an example? Some discontinuity can occur if there's a 360 degree jump in phase calculation, similar to Arpiben's last plot (point 4). That is actually not a discontinuity, just an artifact of how phase is computed. I can see if it can be accounted for when unwrap phase option is not selected.

 

I'll try to reproduce the locking/unlocking behavior you are describing. Don't think I've noticed this before.

 

Group delay - thinking about it, it is highly likely the result of DW using two matching curves to adjust the varaible phase, at the botton versus the top of the spectrum; where they 'meet'.


Frank

 

http://artofaudioconjuring.blogspot.com/

 

 

Ahhh, Mankind ... Porsche intellect, Trabant emotions ...

Share this post


Link to post
Share on other sites
38 minutes ago, esldude said:

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.  

 

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.

 

 

Share this post


Link to post
Share on other sites

Before I forget, in the 64 bit environment I did a manual correction, where all I did was add a [email protected] setting - and got crazy results; part of the scenario was that that upsampling to 88.2k was part of the original match, Cache File Data was selected. What I got were impossible null dB numbers, something like 1536dB, and totally nonsense graphs ...

 

Another memory issue? I tried restarting DW, and recreating the circumstances; but this time it behaved itself - put it into the Nuther Day basket :).


Frank

 

http://artofaudioconjuring.blogspot.com/

 

 

Ahhh, Mankind ... Porsche intellect, Trabant emotions ...

Share this post


Link to post
Share on other sites
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. 


To paraphrase Rick James, "sighted listening is a helluva drug".

Share this post


Link to post
Share on other sites

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 account

Sign in

Already have an account? Sign in here.

Sign In Now

×
×
  • Create New...