Both DAC are transparent as measures by Amir, why one not as clean sounding as the other?

[b4nt]

I don't see much of a difference between high-pass filter at 1k and low pass filter at 20k in your results (-57dB vs. -56dB RMS null and -75dB vs -74dB PK metric).

The PKM (plus the delta spectrum) look different with a HP 1k filter.

With the 1k filter, PKM seems to point to the clapping people (the sounds they made).

 

[pkane]

The PKM (plus the delta spectrum) look different with a HP 1k filter.

With the 1k filter, PKM seems to point to the clapping people (the sounds they made).

View attachment 144078




View attachment 144080

Of course the pattern will change -- you removed frequencies below 1KHz. General audibility levels remained about the same, so no huge differences in audibility.

 

[b4nt]

Of course the pattern will change -- you removed frequencies below 1KHz. General audibility levels remained about the same, so no huge differences in audibility.

I tried, 50, 100, then 1K. Because most of the delta seems to be spread there, and to see the DW resuts.

Isn't it strange DACs cannot correctly render that 0-1k FR part?

 

[pkane]

Isn't it strange DACs cannot correctly render that 0-1k FR part?

I'm missing something. How did you conclude this?

 

[b4nt]

I'm missing something. How did you conclude this?

I may missundertand how DW works, or how I shall read the outputs. In the diff spectrum, I see the highest levels of diffs in the 0-1k frequencies ranges. The in the pk metrics ouputs, huge variations, where i assume we shall see none or not so much?

That is now filtered for 50-1k (plus auto trim checked in the prefs). Most of the difference being there.
Plus a match for same inputs files (PKM is so flat).

 

[pkane]

I may missundertand how DW works, or how I shall read the outputs. In the diff spectrum, I see the highest levels of diffs in the 0-1k frequencies ranges. The in the pk metrics ouputs, huge variations, where i assume we shall see none or not so much?

That is now filtered for 50-1k (plus auto trim checked in the prefs). Most of the difference being there.
Plus a match for same inputs files (PKM is so flat).

View attachment 144091


View attachment 144093

View attachment 144095

You're finding things that are obvious. Most content in that track has a higher level at lower frequencies, below 1kHz. Just look at the spectrum of the file. This is the reason the error/delta signal is larger in those frequencies. No surprises here and this doesn't say anything about the DAC or its ability to process 0-1kHz.

 

[b4nt]

@pkane That being the lower FR, for no filter (LP20k only). So I conclude most the diffs susceptible to be heard are in the lower part of FR:

Most being here below -80dB:

That being again for higher FR, were i would only notice the noises made by clapping people:

 

[b4nt]

You're finding things that are obvious. Most content in that track has a higher level at lower frequencies, below 1kHz. Just look at the spectrum of the file. This is the reason the error/delta signal is larger in those frequencies. No surprises here and this doesn't say anything about the DAC or its ability to process 0-1kHz.

Ok, i see.

May I have a way to tweak DW (maybe the levels) to get a lower of better diff between files? Wy do I see so much there in the delta?

PS: I basically opened the files, clicked on match (with auto trim active).

 

[pkane]

@pkane That being the lower FR, for no filter (LP20k only). So I conclude most the diffs susceptible to be heard are in the lower part of FR:

Most being here below -80dB:

View attachment 144105

That being again for higher FR, were i would only notice the noises made by clapping people:

View attachment 144106

Again, only because the level of the original signal is higher below 1kHz. Once you filter these frequencies out, what remains are the higher frequencies that includes audience clapping and noise. This is as expected.

 

[b4nt]

@pkane The main other thing I see there is jitter. May this have an impact for diff calculation?

Problem I see so is that diff reflects the inputs contents. I was more expecting real diffs....

 

[b4nt]

@pkane X16 compared to itself looks much better. D30 to itself also.

Looks then like things go bad whist comparing D30 to X16, and due to jitter or any minimal diffs between them?

I would add happy to see D30 pro plays twice time almost the same.

 

[pkane]

@pkane The main other thing I see there is jitter. May this have an impact for diff calculation?

Problem I see so is that diff reflects the inputs contents. I was more expecting real diffs....


View attachment 144112

Why is this a problem? The error for most devices will, to a large degree, depend on the level of the input. You can, of course, measure the error relative to the input. That's what 'dBr' measurement in PK Metric represents, for example.

 

[pkane]

@pkane X16 compared to itself looks much better. D30 to itself also.

Looks then like things go bad whist comparing D30 to X16, and due to jitter or any minimal diffs between them?

I would add happy to see D30 pro plays twice time almost the same.

View attachment 144113

View attachment 144114

Is there a reason you're repeating all the same measurements and discussions that have already occurred here at least a few times, in this very thread?

 

[b4nt]

Is there a reason you're repeating all the same measurements and discussions that have already occurred here at least a few times, in this very thread?

I would need some clarifications also. I saw a lot more files added here, and was wondering what I would see using them.

So DACs outputs compared to themself seem nice, play twice the same (this is the minimum one could expect). Results are worther comparing them to each other. Or taking the original wave file as ref and comparing a DAC output...

And the only thing I do notice now is maybe related to jitter and clock drifts. Which may add errors to the results (DAC+ADC+DW math). Not saying DW cant do math. But that DACs and ADC shitf/drift add jitter during measurements, which might impact the results DW shows.

Is there anything to tweak in DW for jitter and clocks drifting? What would be required for better matchs, DACs clocks sync?

 

[Blumlein 88]

Jitter and drift are two different things. You could have a clock with zero jitter that runs slightly fast or slow and will drift. Jitter is a sample to sample timing variation. You could have a clock with lots of jitter, but zero clock drift if the average speed is the same as another clock.

Yes, if you have external clocking so both devices use the same clock you won't capture jitter in a loopback. You will still have jitter, but it will effect both clocks the same way at the same time and the recorded result will be no jitter.

 

[pkane]

I would need some clarifications also. I saw a lot more files added here, and was wondering what I would see using them.

So DACs outputs compared to themself seem nice, play twice the same (this is the minimum one could expect). Results are worther comparing them to each other. Or taking the original wave file as ref and comparing a DAC output...

And the only thing I do notice now is maybe related to jitter and clock drifts. Which may add errors to the results (DAC+ADC+DW math). Not saying DW cant do math. But that DACs and ADC shitf/drift add jitter during measurements, which might impact the results DW shows.

Is there anything to tweak in DW for jitter and clocks drifting? What would be required for better matchs, DACs clocks sync?

View attachment 144119

There are many well-known differences between the original and the loop-back file vs. two loop-back files. Some have to do with reconstruction and antialias filters, non-linear distortions, noise, digital processing (like dither), rate conversions, clock differences, analog filtering, and yes, jitter. I'm afraid I'm not equipped to give lessons in fundamental signal processing, but there's plenty of information out there, on the web and in this forum, so try to learn as much as you can before making any conclusions.

 

[Pdxwayne]

Unless experts here object, based on latest USB captures, Node2i Coax captures, and files comparisons after trimming (cut 2 seconds from begin and from end), I am going to conclude the findings as:

*d30pro is not as stable as X16
*d30pro vs orig contains instances where null difference reached potential audible levels, whereas X16 is about 10db better, thus unlikely to have audible difference vs orig.

 

[Jimbob54]

Unless experts here object, based on latest USB captures, Node2i Coax captures, and files comparisons after trimming (cut 2 seconds from begin and from end), I am going to conclude the findings as:

*d30pro is not as stable as X16
*d30pro vs orig contains instances where null difference reached potential audible levels, whereas X16 is about 10db better, thus unlikely to have audible difference vs orig.

And what's the future hold for the d30? And why?

 

[b4nt]

There are many well-known differences between the original and the loop-back file vs. two loop-back files. ... before making any conclusions.

No, I will not make any conclusions. More wondering how much one would have to consider all such paramaters before comparing DACs that way, feeding in and comparing DACs records, and comparing to the original file, expecting up to a perfect match.

Are there any known PK metrics one could reach using common measurement equipements and techniques? Or any recommandations to set up a bench for such activities?

 

[pkane]

No, I will not make any conclusions. More wondering how much one would have to consider all such paramaters before comparing DACs that way, feeding in and comparing DACs records, and comparing to the original file, expecting up to a perfect match.

Are there any known PK metrics one could reach using common measurement equipements and techniques? Or any recommandations to set up a bench for such activities?

Here you can find some pro interfaces and converters analyzed using DeltaWave, including their PK Metric:

https://deltaw.org/gearslutz.html

For doing a more detailed analysis on what each DAC or ADC is doing to the signal, you'll want to start with the basics:

https://www.audiosciencereview.com/forum/index.php?threads/understanding-audio-measurements.2351/
https://www.audiosciencereview.com/...derstanding-digital-audio-measurements.10523/
https://www.audiosciencereview.com/forum/index.php?forums/audio-science-review-video-channel.61/