• Welcome to ASR. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Beta-test: DeltaWave Null Comparison software

It's always given me a value in the past...

Strange. I'm getting one with my testing with the new version, too. If you want, share a reference/comparison files with me, and maybe your settings file, as well (located in C:\ProgramData\DeltaWave\_DeltaWaveDefault.dw). I'll see if I can track this down.

1739123789229.png
 
I'm occasionally getting this error when I try to play delta file, any ideas why? I can "fix" it if I swap between reference and compare files i.e. I still use the same files.

delta.jpg
 
Hmm, strange. That line number in 2.0.16 shows nothing that could produce the error you posted. You said it only happens occasionally? Any specific actions you can think of to try to reproduce it?
It seems to happen with certain files, not occasionally. I have recently recorder files with Audacity, all identical length etc. And comparision of some of them causes the error to appear. I have one file pair that always causes it, I just load them, press match and when finished try to play delta and the error appears.
I have also noticed this: the error appears only when stereo mode is selected, the same files work fine with L, R and L+R. And they work in stereo if reference/compare files are switched.
 
It seems to happen with certain files, not occasionally. I have recently recorder files with Audacity, all identical length etc. And comparision of some of them causes the error to appear. I have one file pair that always causes it, I just load them, press match and when finished try to play delta and the error appears.
I have also noticed this: the error appears only when stereo mode is selected, the same files work fine with L, R and L+R. And they work in stereo if reference/compare files are switched.

If you can share two such files (dropbox.com or box.com or wetransfer.com will let you share large files), I can try to reproduce this.
 
Quick fix v2.0.17 to address the issue reported by @TimoJ: https://app.box.com/s/hlglgy3aj39rjfxu343fho4t0cdrq9v8

The problem was with the difference file playback in stereo mode, where with some input files, one channel turned out to have fewer samples than the other after being trimmed which resulted in 'index out of range' error.
 
Quick fix v2.0.17 to address the issue reported by @TimoJ: https://app.box.com/s/hlglgy3aj39rjfxu343fho4t0cdrq9v8

The problem was with the difference file playback in stereo mode, where with some input files, one channel turned out to have fewer samples than the other after being trimmed which resulted in 'index out of range' error.
I'm getting a server not found message on that link. I was on a VPN. Turning off the VPN let it work. Just in case others run into the issue thought I would note it.
 
Update on installing DeltaWave on Linux (based on @pkane's post here), in this case on a current Linux Mint based on Ubuntu 24.04 noble:

Code:
sudo apt install wine
sudo apt install winbind
env WINEPREFIX=$HOME/winedotnet wineboot --init
env WINEPREFIX=$HOME/winedotnet winetricks --force dotnet461 corefonts
env WINEPREFIX=$HOME/winedotnet wine DeltaWaveSetup.exe

With NVidia-based graphics, you may get a lot of lines like
Code:
0080:err:winediag:is_broken_driver Broken NVIDIA RandR detected, falling back to RandR 1.0. Please consider using the Nouveau driver instead.
during the install but these can be ignored

Finally create and edit laucher, set "command" to
Code:
env WINEPREFIX="/home/XXX/winedotnet" wine-stable "C:\\Program Files (x86)\\PKAudio\\DeltaWave\\DeltaWave.exe"
(replace XXX with your Linux user name)
 
In my latest runs to compare DAC/amp dongles playing the same song I've been seeing more frequent failures of the primary drift correction method, followed by a dialog popping up asking if I wanted to use an alternate method. This second method seems to always run faster and give better results for my material. Is there some way I can force DeltaWave to skip right to using this alternate method and not even try the first one?

But also, regardless of method: can I force some correction-related value manually in order to (hopefully) get an identical result on later re-runs on the same pair of files, like if I write down what the Phase Offset turned out to be last time and try to reuse that one? (I like it as a candidate for a reusable value since it's shown with 6 decimal precision.)
 
In my latest runs to compare DAC/amp dongles playing the same song I've been seeing more frequent failures of the primary drift correction method, followed by a dialog popping up asking if I wanted to use an alternate method. This second method seems to always run faster and give better results for my material. Is there some way I can force DeltaWave to skip right to using this alternate method and not even try the first one?

But also, regardless of method: can I force some correction-related value manually in order to (hopefully) get an identical result on later re-runs on the same pair of files, like if I write down what the Phase Offset turned out to be last time and try to reuse that one? (I like it as a candidate for a reusable value since it's shown with 6 decimal precision.)
The problem with the "simple method" is that it is based on only measuring drift between start and end of the recording, so less accurate but faster. The reason it's being suggested is that other points in between these two produced large error. That indicates that the recording is either noisy, or time is not consistently drifting over the whole recording.

You can apply all the corrections manually from the Process->Manual Adjustments window, but you do need to enter all the values correctly, gain + DC, offset, and drift.
 
A new version of DeltaWave is available for testing, v2.0.19:

https://app.box.com/s/hlglgy3aj39rjfxu343fho4t0cdrq9v8

1. This version adds the option of automatically removing the phase effects of a minimum-phase DC filter by attempting to solve for the filter parameters (discussion). The automatic DC filter calculation and removal is configured under settings:

1745586981621.png


When (auto) frequency is selected, DeltaWave will try to solve for high-pass/DC filter parameters, as part of the Match operation, and automatically remove the effect of any such minimum-phase filter. If a number is entered instead of (auto), the DC filter will be removed with that corner frequency. This will be applied to all comparisons, until "Undo MP DC filter" option is unchecked. Computed filter frequency will be logged in the Results tab for your reference:

1745589180464.png


2. The option for computing the frequency of such a minimum-phase DC filter is also available under Process->Compute DC Filter correction menu. You may want to uncheck the "Undo MP DC filter" option when using this, as the menu will work on the result of the match operation. If this operation already undoes the DC filter, then the menu command wlil not find much left to correct :)

3. This version also adds the option to save audio WAV files as 16- and 24-bit integer PCM samples, and as 32- and 64-bit floating point:
1745587328456.png


Dither will be applied at the bit size selected in settings (or not applied if unselected):

1745587446270.png
 
A new version of DeltaWave is available for testing, v2.0.19:

https://app.box.com/s/hlglgy3aj39rjfxu343fho4t0cdrq9v8

1. This version adds the option of automatically removing the phase effects of a minimum-phase DC filter by attempting to solve for the filter parameters (discussion). The automatic DC filter calculation and removal is configured under settings:

View attachment 446636

When (auto) frequency is selected, DeltaWave will try to solve for high-pass/DC filter parameters, as part of the Match operation, and automatically remove the effect of any such minimum-phase filter. If a number is entered instead of (auto), the DC filter will be removed with that corner frequency. This will be applied to all comparisons, until "Undo MP DC filter" option is unchecked. Computed filter frequency will be logged in the Results tab for your reference:

View attachment 446661

2. The option for computing the frequency of such a minimum-phase DC filter is also available under Process->Compute DC Filter correction menu. You may want to uncheck the "Undo MP DC filter" option when using this, as the menu will work on the result of the match operation. If this operation already undoes the DC filter, then the menu command wlil not find much left to correct :)

3. This version also adds the option to save audio WAV files as 16- and 24-bit integer PCM samples, and as 32- and 64-bit floating point:
View attachment 446643

Dither will be applied at the bit size selected in settings (or not applied if unselected):

View attachment 446645

Hi @pkane !

Thanks again!

This may be an user error, but I can't get the automatic DC filter removal to work:

1746122029703.png


It detects the filter:
1746122247556.png


Still nothing happens:
1746122373282.png




But if I do it manually with the compare file...

1746122623133.png


... and invert the phase...

1746122702065.png


... it works ok:

1746122761111.png


My first thought was that this was due to the decimal separator (Comma), but it works the same way even if I change it to a full stop.

And just like you, I don't want to mess up the reference file, but it works too when testing:

1746123512242.png


Without inverting the phase:

1746123560776.png


The result is the same:

1746123600665.png
 
Hi @pkane !

Thanks again!

This may be an user error, but I can't get the automatic DC filter removal to work:

View attachment 448037

It detects the filter:
View attachment 448038

Still nothing happens:
View attachment 448039



But if I do it manually with the compare file...

View attachment 448044

... and invert the phase...

View attachment 448047

... it works ok:

View attachment 448049

My first thought was that this was due to the decimal separator (Comma), but it works the same way even if I change it to a full stop.

And just like you, I don't want to mess up the reference file, but it works too when testing:

View attachment 448051

Without inverting the phase:

View attachment 448052

The result is the same:

View attachment 448053

I think the phase plot is updated prior to DC filter removal, so it shows original phase. The calculations do seem to reflect the DC filter removal, so that part worked just fine :) I'll make the adjustment to update the plot.
 
Hi pkane, thanks for developing this, it's looks so incredibly useful and promising to shed light on perceived difference in digital audio reproduction..

I also posted on your website but I seem to perhaps have run into a bug?

I seem to have a problem with measuring the R channel exclusively, I always get completely off the chart delta phase results. I did all kinds of cross checking with different dacs, output modes, feeding the R out into the L of my ADC, swapping the channel in my test files and using different reference test files, but everything seems to point to a software bug. Also in stereo mode, I don't get the problem either. Could you please let me know if you can reproduce or if you need debug logging? It never happens on channel L.

Audio data on channel R

1746538955468.png


Same exact data on same ADC input from channel L (multiple DACs)

1746539174432.png


Thanks again
 
Back
Top Bottom