Beta-test: DeltaWave Null Comparison software

[Pluto]

dB scale would certainly make more sense

Thanks.

And while thinking about that area of the software, this might be a cool feature: on the Δ waveform plot (and possibly others), a button to toggle the Y axis sensitivity between the view it currently defaults to (let's call this 'normal') and the same Y axis sensitivity being used on the 'matched' (aligned waveforms) tab.

While, in many cases, this will reduce the plot to nothing more than a straight line across the middle, it will be useful for getting the level of difference artefacts into perspective.

ALSO -- currently the tab names and the description that appears at the top of each plot are (mostly) the same. How about using the ample space at the top of the plot to put a slightly more descriptive version of what each plot is showing i.e. rather than "Delta Waveform" write "The residual waveform after subtracting the correlated waveform from the original" or words to that effect. Once this program gets out into the big wide world, your biggest problem is likely to be failure of comprehension amongst users and the answer to this will be the best possible descriptions of what everything means.

 

[Pluto]

Something like this?

 

[Pluto]

I notice that on those plots in which the X axis is time and Y axis labels are large, it causes the playback position indicator to misalign with the plot. Does this need to be addressed?

 

[pkane]

I notice that on those plots in which the X axis is time and Y axis labels are large, it causes the playback position indicator to misalign with the plot. Does this need to be addressed?

I didn't realize it was aligned with the plot previously...

 

[Pluto]

For most of the plots in which the X-axis represents time, the playback progress bar tracks the time shown on the X-axis accurately, so I presumed it was designed that way. If, for one reason or another, the Y-axis labels are wider on one plot than the others, the zero point of the graph is further to the right to accommodate the extra width of the labels but the start of the playback progress bar remains unchanged, resulting in a misalignment.

Perhaps the simplest answer is to move the X-axis zero datum point for all plots to the same place, slightly further to the right than it is now (to accommodate any wider Y-axis labels), likewise the start of the progress bar which will ensure that the zero datum and the start of the playback progress bar will always align.

Perhaps the alignment between the playback progress bar and (some of) the plots was never intentional? I thought it was a cunning plan

 

[Pluto]

Installation – something worth drawing to your attention

While doing routine maintenance, I discovered a number of DeltaWave entries in my Add/Remove programs list, relating to the number of test versions that have been installed over the past few weeks.

Not understanding the exact mechanisms at work here, I have to ask why new versions of the same program are being treated, by my system, as new software packages rather than updates of an existing one.

 

[pkane]

Installation – something worth drawing to your attention

While doing routine maintenance, I discovered a number of DeltaWave entries in my Add/Remove programs list, relating to the number of test versions that have been installed over the past few weeks.

Not understanding the exact mechanisms at work here, I have to ask why new versions of the same program are being treated, by my system, as new software packages rather than updates of an existing one.

Good question. Bundled installer with Microsoft's development studio is black art, and more often than not it is not clear as to why or what it's doing I'll take a look.

 

[Blumlein 88]

Installation – something worth drawing to your attention

While doing routine maintenance, I discovered a number of DeltaWave entries in my Add/Remove programs list, relating to the number of test versions that have been installed over the past few weeks.

Not understanding the exact mechanisms at work here, I have to ask why new versions of the same program are being treated, by my system, as new software packages rather than updates of an existing one.

I've seen this too. So I uninstalled before installing newer versions.

 

[pkane]

I've seen this too. So I uninstalled before installing newer versions.

Yes, I see that. The installation actually updates the software as expected, but somehow doesn't remove the previous entry from the list of installed packages. Very strange behavior.

 

[rocknss]

Thanks. Looks like you have saved me some time, I was thinking about developing something like this.

 

[RayDunzl]

Well, it worked for a while...

 

[pkane]

Well, it worked for a while...

View attachment 23951

What format are the files in, and how long are they? Try resetting the trim settings to zeros and see if the error goes away.

You can also go to the Help->Logging->View Log menu and post the end of the file, where presumably there is an error message with details.

 

[RayDunzl]

What format are the files in, and how long are they?

24/44.1 - In-room recording via Audacity using UMIK-1, exported to .wav, so probably 16/44.1 at that point.

The files are just under two minutes.

Try resetting the trim settings to zeroes and see if the error goes away.

Restarted the app, changed trim to 0, reselected the files (opposite order)

You can also go to the Help->Logging->View Log menu and post the end of the file, where presumably there is an error message with details.

Log:

2019-03-22 12:22:04.2965|INFO|Wave.WaveForm|DeltaWave 1.0.18.0 starting up
2019-03-22 12:22:04.3825|INFO|Wave.WaveForm|Adding driver: [ASIO] ASIO4ALL v2
2019-03-22 12:22:04.3825|INFO|Wave.WaveForm|Adding driver: [ASIO] miniDSP ASIO Driver
2019-03-22 12:22:04.3975|INFO|Wave.WaveForm|Adding driver: [ASIO] Topping USB Audio Device
2019-03-22 12:22:04.3975|INFO|Wave.WaveForm|Adding driver: [ASIO] UMC ASIO Driver
2019-03-22 12:22:04.4435|INFO|Wave.WaveForm|Adding driver: [WASAPI]{0.0.0.00000000}.{0a0b8dca-0be7-4dbf-b760-f53ab8fd8ba9} | [WASAPI] Speakers (2- TOPPING USB DAC) 48000/32
2019-03-22 12:22:04.4755|INFO|Wave.WaveForm|Adding driver: [WASAPI]{0.0.0.00000000}.{2ecd896b-6393-4e37-9b6e-a2d218958679} | [WASAPI] Speakers (Benchmark DAC2 USB Audio 2.0) 48000/32
2019-03-22 12:22:04.5055|INFO|Wave.WaveForm|Adding driver: [WASAPI]{0.0.0.00000000}.{818a5cd1-dc76-44fb-bf58-74b882a81dd3} | [WASAPI] Speakers (miniDSP 2x4n) 48000/32
2019-03-22 12:22:04.5425|INFO|Wave.WaveForm|Adding driver: [WASAPI]{0.0.0.00000000}.{d215e59e-a420-4909-ac70-7e03986f5581} | [WASAPI] LG TV-4 (NVIDIA High Definition Audio) 48000/32
2019-03-22 12:22:04.5975|INFO|Wave.WaveForm|Adding driver: [WASAPI]{0.0.0.00000000}.{f28658e1-6bcc-4f90-9221-1074ed35e888} | [WASAPI] Speakers (Realtek High Definition Audio) 44100/32
2019-03-22 12:22:04.6265|INFO|Wave.WaveForm|Adding driver: [WASAPI]{0.0.0.00000000}.{feec5e7e-cab0-4899-aa8b-42b3989d04b9} | [WASAPI] Speakers (DTS PlayFi Audio Renderer) 44100/32
2019-03-22 12:22:04.6315|INFO|Wave.WaveForm|Current driver: [WASAPI]{0.0.0.00000000}.{f28658e1-6bcc-4f90-9221-1074ed35e888}
2019-03-22 12:22:04.6765|INFO|Wave.WaveForm|Adding driver: [ASIO] ASIO4ALL v2
2019-03-22 12:22:04.6825|INFO|Wave.WaveForm|Adding driver: [ASIO] miniDSP ASIO Driver
2019-03-22 12:22:04.6825|INFO|Wave.WaveForm|Adding driver: [ASIO] Topping USB Audio Device
2019-03-22 12:22:04.6825|INFO|Wave.WaveForm|Adding driver: [ASIO] UMC ASIO Driver
2019-03-22 12:22:04.7165|INFO|Wave.WaveForm|Adding driver: [WASAPI]{0.0.0.00000000}.{0a0b8dca-0be7-4dbf-b760-f53ab8fd8ba9} | [WASAPI] Speakers (2- TOPPING USB DAC) 48000/32
2019-03-22 12:22:04.7445|INFO|Wave.WaveForm|Adding driver: [WASAPI]{0.0.0.00000000}.{2ecd896b-6393-4e37-9b6e-a2d218958679} | [WASAPI] Speakers (Benchmark DAC2 USB Audio 2.0) 48000/32
2019-03-22 12:22:04.7785|INFO|Wave.WaveForm|Adding driver: [WASAPI]{0.0.0.00000000}.{818a5cd1-dc76-44fb-bf58-74b882a81dd3} | [WASAPI] Speakers (miniDSP 2x4n) 48000/32
2019-03-22 12:22:04.8125|INFO|Wave.WaveForm|Adding driver: [WASAPI]{0.0.0.00000000}.{d215e59e-a420-4909-ac70-7e03986f5581} | [WASAPI] LG TV-4 (NVIDIA High Definition Audio) 48000/32
2019-03-22 12:22:04.8655|INFO|Wave.WaveForm|Adding driver: [WASAPI]{0.0.0.00000000}.{f28658e1-6bcc-4f90-9221-1074ed35e888} | [WASAPI] Speakers (Realtek High Definition Audio) 44100/32
2019-03-22 12:22:04.8935|INFO|Wave.WaveForm|Adding driver: [WASAPI]{0.0.0.00000000}.{feec5e7e-cab0-4899-aa8b-42b3989d04b9} | [WASAPI] Speakers (DTS PlayFi Audio Renderer) 44100/32
2019-03-22 12:22:04.8985|INFO|Wave.WaveForm|Current driver: [WASAPI]{0.0.0.00000000}.{f28658e1-6bcc-4f90-9221-1074ed35e888}
2019-03-22 12:22:07.4725|INFO|Wave.WaveForm|ProcessAll
2019-03-22 12:22:07.4985|ERROR|Wave.WaveForm|Failed to load file: D:\Temp\Frame2 audio.wav

---

Comment:

Trim:

It didn't like comparing that length when it initially did work, so I trimmed it. the complaint seemed to be related to some time difference.

The trim seems to want seconds from beginning and seconds from end. It would make more sense to me to specify a start point and a duration than a distance from each end (assuming that's how it works - seems to be).

Had to keep trimming the the end to get it to a size it would smunge without complaining about the size. Not sure how long that was. Let's see:

Start at +75, end at -35, 120 second file, so about 120 -35 = +85, so, about a ten second compare.

So that worked yesterday, until it didn't, several compares and app restarts later.

My first time trying the software, so you can blame me for all of it.

 

[pkane]

.wav 24/28 - In-room recording via Audacity using UMIK-1, exported to .wav, so probably 16/48 at that point.

The files are just under two minutes.




Restarted the app, changed trim to 0, reselected the files (opposite order)

View attachment 23954



Log:

2019-03-22 12:22:04.2965|INFO|Wave.WaveForm|DeltaWave 1.0.18.0 starting up
2019-03-22 12:22:04.3825|INFO|Wave.WaveForm|Adding driver: [ASIO] ASIO4ALL v2
2019-03-22 12:22:04.3825|INFO|Wave.WaveForm|Adding driver: [ASIO] miniDSP ASIO Driver
2019-03-22 12:22:04.3975|INFO|Wave.WaveForm|Adding driver: [ASIO] Topping USB Audio Device
2019-03-22 12:22:04.3975|INFO|Wave.WaveForm|Adding driver: [ASIO] UMC ASIO Driver
2019-03-22 12:22:04.4435|INFO|Wave.WaveForm|Adding driver: [WASAPI]{0.0.0.00000000}.{0a0b8dca-0be7-4dbf-b760-f53ab8fd8ba9} | [WASAPI] Speakers (2- TOPPING USB DAC) 48000/32
2019-03-22 12:22:04.4755|INFO|Wave.WaveForm|Adding driver: [WASAPI]{0.0.0.00000000}.{2ecd896b-6393-4e37-9b6e-a2d218958679} | [WASAPI] Speakers (Benchmark DAC2 USB Audio 2.0) 48000/32
2019-03-22 12:22:04.5055|INFO|Wave.WaveForm|Adding driver: [WASAPI]{0.0.0.00000000}.{818a5cd1-dc76-44fb-bf58-74b882a81dd3} | [WASAPI] Speakers (miniDSP 2x4n) 48000/32
2019-03-22 12:22:04.5425|INFO|Wave.WaveForm|Adding driver: [WASAPI]{0.0.0.00000000}.{d215e59e-a420-4909-ac70-7e03986f5581} | [WASAPI] LG TV-4 (NVIDIA High Definition Audio) 48000/32
2019-03-22 12:22:04.5975|INFO|Wave.WaveForm|Adding driver: [WASAPI]{0.0.0.00000000}.{f28658e1-6bcc-4f90-9221-1074ed35e888} | [WASAPI] Speakers (Realtek High Definition Audio) 44100/32
2019-03-22 12:22:04.6265|INFO|Wave.WaveForm|Adding driver: [WASAPI]{0.0.0.00000000}.{feec5e7e-cab0-4899-aa8b-42b3989d04b9} | [WASAPI] Speakers (DTS PlayFi Audio Renderer) 44100/32
2019-03-22 12:22:04.6315|INFO|Wave.WaveForm|Current driver: [WASAPI]{0.0.0.00000000}.{f28658e1-6bcc-4f90-9221-1074ed35e888}
2019-03-22 12:22:04.6765|INFO|Wave.WaveForm|Adding driver: [ASIO] ASIO4ALL v2
2019-03-22 12:22:04.6825|INFO|Wave.WaveForm|Adding driver: [ASIO] miniDSP ASIO Driver
2019-03-22 12:22:04.6825|INFO|Wave.WaveForm|Adding driver: [ASIO] Topping USB Audio Device
2019-03-22 12:22:04.6825|INFO|Wave.WaveForm|Adding driver: [ASIO] UMC ASIO Driver
2019-03-22 12:22:04.7165|INFO|Wave.WaveForm|Adding driver: [WASAPI]{0.0.0.00000000}.{0a0b8dca-0be7-4dbf-b760-f53ab8fd8ba9} | [WASAPI] Speakers (2- TOPPING USB DAC) 48000/32
2019-03-22 12:22:04.7445|INFO|Wave.WaveForm|Adding driver: [WASAPI]{0.0.0.00000000}.{2ecd896b-6393-4e37-9b6e-a2d218958679} | [WASAPI] Speakers (Benchmark DAC2 USB Audio 2.0) 48000/32
2019-03-22 12:22:04.7785|INFO|Wave.WaveForm|Adding driver: [WASAPI]{0.0.0.00000000}.{818a5cd1-dc76-44fb-bf58-74b882a81dd3} | [WASAPI] Speakers (miniDSP 2x4n) 48000/32
2019-03-22 12:22:04.8125|INFO|Wave.WaveForm|Adding driver: [WASAPI]{0.0.0.00000000}.{d215e59e-a420-4909-ac70-7e03986f5581} | [WASAPI] LG TV-4 (NVIDIA High Definition Audio) 48000/32
2019-03-22 12:22:04.8655|INFO|Wave.WaveForm|Adding driver: [WASAPI]{0.0.0.00000000}.{f28658e1-6bcc-4f90-9221-1074ed35e888} | [WASAPI] Speakers (Realtek High Definition Audio) 44100/32
2019-03-22 12:22:04.8935|INFO|Wave.WaveForm|Adding driver: [WASAPI]{0.0.0.00000000}.{feec5e7e-cab0-4899-aa8b-42b3989d04b9} | [WASAPI] Speakers (DTS PlayFi Audio Renderer) 44100/32
2019-03-22 12:22:04.8985|INFO|Wave.WaveForm|Current driver: [WASAPI]{0.0.0.00000000}.{f28658e1-6bcc-4f90-9221-1074ed35e888}
2019-03-22 12:22:07.4725|INFO|Wave.WaveForm|ProcessAll
2019-03-22 12:22:07.4985|ERROR|Wave.WaveForm|Failed to load file: D:\Temp\Frame2 audio.wav

---

Comment:

Trim:

It didn't like comparing that length when it initially did work, so I trimmed it. the complaint seemed to be related to some time difference.

The trim seems to want seconds from beginning and seconds from end. It would make more sense to me to specify a start point and a duration than a distance from each end (assuming that's how it works - seems to be).

Had to keep trimming the the end to get it to a size it would smunge without complaining about the size. Not sure how long that was. Let's see:

Start at +75, end at -35, 120 second file, so about 120 -35 = +85, so, about a ten second compare.

So that worked yesterday, until it didn't, several compares and app restarts later.

My first time trying the software, so you can blame me.

Can't blame you, try as I might, as I don't see anything that you did wrong

The issue appears to be in loading the files, so it's not even getting to the point of trying to match them. Can you please try selecting just the Left or Right channel for both files, instead of L+R? Maybe there's somehow a mismatch between the two channels or something else not matching up when doing stereo->mono conversion.

You can also try the latest version (not fully baked yet, but has some fixes for loading different file formats):

https://drive.google.com/open?id=1STW_HC903Lq8x0JKhDfUBkEFCay8fQ1L

And if the problem isn't resolved, it would help me if you could share at least the Frame2 audio.wav file so I can try to reproduce the problem.

 

[pkane]

Comment:

Trim:

It didn't like comparing that length when it initially did work, so I trimmed it. the complaint seemed to be related to some time difference.

The trim seems to want seconds from beginning and seconds from end. It would make more sense to me to specify a start point and a duration than a distance from each end (assuming that's how it works - seems to be).

Had to keep trimming the the end to get it to a size it would smunge without complaining about the size. Not sure how long that was. Let's see:

Start at +75, end at -35, 120 second file, so about 120 -35 = +85, so, about a ten second compare.

So that worked yesterday, until it didn't, several compares and app restarts later.

My first time trying the software, so you can blame me for all of it.

Trim works by removing some some seconds from the front and back of the file. You're right that specifying length might be better, but this was primarily designed to help remove some start-up/end of recording differences where tracks might diverge significantly. Specifying a few seconds to trim from front and back seemed like an easier way to do this way.

Two minutes should not cause DW to complain at all and shouldn't require any trimming. I've compared tracks that are 7-8 minutes in length. Although slow and requires more memory, that should work. Do you remember what the message said when it complained about a 2 minute recording?

 

[RayDunzl]

Do you remember what the message said when it complained about a 2 minute recording?

Go the files loaded (see below)

But got the message as it started working on them:

---

Can you please try selecting just the Left or Right channel for both files, instead of L+R? Maybe there's somehow a mismatch between the two channels or something else not matching up when doing stereo->mono conversion.

The loading problem:

The files are mono, so L+R fails (I guess). Loading just L works now.

Operator Error

---

Files:

https://www.dropbox.com/s/6aneoy7xbcx65yx/Frame1 audio.wav?dl=0

https://www.dropbox.com/s/6wzdq54qtg6bjey/Frame2 audio.wav?dl=0

(links may not work yet)

 

[pkane]

Go the files loaded (see below)

But got the message as it started working on them:

View attachment 23966

---



The loading problem:

The files are mono, so L+R fails (I guess). Loading just L works now.

Operator Error

---

Files:

https://www.dropbox.com/s/6aneoy7xbcx65yx/Frame1 audio.wav?dl=0

https://www.dropbox.com/s/6wzdq54qtg6bjey/Frame2 audio.wav?dl=0

(links may not work yet)

Thanks, Ray. Interesting data, it looks like there are significant (and varying) phase differences between the two tracks. That's the reason you see the residual error message. It means that DW couldn't get the error reduced below 3 samples while trying to correct for clock drift, so the result is probably not valid. When that happens, the best thing to try is to just turn off drift correction and see if you can get a better result. Here's what I get when I do that with your two files:

Final peak values File 1: -10.088dB File 2: -10.47dB
Final RMS values File 1: -27.588dB File 2: -27.754dB

Gain= 0.1551dB (1.018x) Phase offset=-8.520833ms (-409 samples)
Difference (rms) = -31.67dB [-32.91dBA]
Correlated Null Depth=29.95dB [30.59dBA]

Files are NOT a bit-perfect match (match=0.06%) at 16 bits
Files match @ 50% when reduced to 5.65 bits

Phase difference (full bandwidth): 93.2539017914677°
0-10,000Hz: 96.0567326094822°
0-20,000Hz: 99.0272571276742°
0-24,000Hz: 93.2539017914677°
RMS of the difference of spectra: -87.9278096238708dB

 

[RayDunzl]

Interesting data, it looks like there are significant (and varying) phase differences between the two tracks.

Probably due to the ceiling fan.

It's not clocked off the DAC.

 

[Blumlein 88]

There is periodic jitter. There is random jitter. Then there is ceiling fan jitter. Unlike the first two the latter is audible. I think it should henceforth be named after the discoverer of ceiling fan jitter. I'm torn though as to whether it will be Dunzl jitter or maybe just Ray jitter in order to confuse those who think it might be related to ray tracing software or something.

PS the latest version of Deltawave is really nice.

 

[RayDunzl]

Repeat, new recordings, with No Ceiling Fan, a little higher signal to ambient noise level...

Files loaded without the previous "residual drift" error...

The waveform difference is about 3 to 4 times the ambient/mic noise level:

System/ambient/mic noise at beginning of recording:

Spectrum of the noise only:

Looks about the same as the first graphic above.