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

[pkane]

Wouldn't that have a minimal visible impact using REW and multitone for the different inputs? I think PO had done a run for USB, seemed clean... Maybe DW is the ultimate tool

Only known problem I've seen so far is related to spdif, for some solutions and sources having too much jitter: no synch, or synch but resynchs (silences).

Clock drift/difference between devices isn't a real problem, unless you need two devices to be sync'ed up (like in video and audio). This small amount of drift compared to a reference will not result in any audible differences, yet they are (and need to be) measured by DeltaWave in order to compute a more precise null waveform.

 

[b4nt]

Clock drift/difference between devices isn't a real problem

As I understood, clocking problem can come from jitter (not the right time/edge), with audible effects starting in upper FR. This where the main discussions, some years ago. And what you had added showed some very minimal diffs in upper FR, but at -100dB... And I was wondering if multitone would show anything in upper FR, a difference in SNR there, or some spikes, depending of the input.

According to psychoacoustic experts, most or all of us wouldn't be able to distinguish 1005 from 1000Hz. Any very slow shifting around base freq shouldn't be an issue.

 

[pkane]

As I understood, clocking problem can come from jitter (not the right time/edge), with audible effects starting in upper FR. This where the main discussions, some years ago. And what you had added showed some very minimal diffs in upper FR, but at -100dB... And I was wondering if multitone would show anything in upper FR, a difference in SNR there, or some spikes, depending of the input.

According to psychoacoustic experts, most or all of us wouldn't be able to distinguish 1005 from 1000Hz. Any very slow shifting around base freq shouldn't be an issue.

DeltaWave measures all of this in minute detail, including applying perception weights based on frequency. There's no reason to speculate.

 

[b4nt]

DeltaWave measures all of this in minute detail, including applying perception weights based on frequency. There's no reason to speculate.

I'm not speculating, just wondering if it could be measured by other ways also. Any such signals being so at -100dB (out of audible range or masked), unless a input would be very bad.

I found that video, listened to the last 10 minutes, provided me the main answers:

 

[Pdxwayne]

It doesn't point at anything other than the data is different at the beginning, which the spectrogram confirms. Not better, not worse, just different. If DeltaWave can't produce a good match, you need to dig in to understand why and try to correct the settings/parameters to see if you can get a proper result. After the first few seconds, the match is just as good between D30pro and the original as it is between X16 and the original. What does that tell you? Tells me that both DACs are performing at about the same level.

I was curious if the "different at the beginning" issue happened before for d30pro USB captures.

I checked prior USB captures from more than 10 days ago (older laptop did the playback, both laptops plugged in) and I found one capture that also exhibits same kind of different in the beginning. Not both captures, but capture2 of d30pro using filter 2, from July 16. X16 andE30, as-is, at that time, already have excellent fit with the original.

Also, with that older capture, to get "excellent" fit with orig, I needed to trim 3 seconds from the beginning for both d30pro and orig before running the deltawave null process.

d30pro USB capture2, right channel, as-is, vs orig:

Trimmed 1 second get better pkmetric number, but only get "Very Good" fit:

Trimmed 3 seconds, the fit then is "Excellent":

Both older USB E30 and X16 captures, as-is, vs orig got "excellent" fit.
E30 vs Orig:

X16 vs Orig:

 

[Pdxwayne]

I have done Node2i Coax captures.

Download link:
https://we.tl/t-eWrAqqwV7T

d30pro captures, like USB captures, continue to have one random channel that won't null well against orig, unless 1 second is trimmed. X16 and E30 continue to exhibit "excellent" fit with orig and very consistent.

Here are d30pro information.

d30pro vs d30pro left:

d30pro vs d30pro Right:

d30pro capture1 vs orig, Left:

*************************************************************************
d30pro capture1 vs Orig, Left, trimmed 1 second in the beginning

d30pro capture1 vs orig, Right:

d30pro capture2 vs Orig, Left:

D30pro capture 2 vs Orig, Right:

 

[Pdxwayne]

X16 Node2i Coax captures comparisons:

X16 vs X16, Left channel:

X16 vs X16, Right channel:

X16 capture 1 vs Orig, Left:

X16 capture 1 vs Orig, Right:

X16 capture 2 vs Orig, Left:

X16 capture 2 vs Orig, Right:

 

[Pdxwayne]

@Blumlein 88 and @pkane , I played around with USB captured files using trimming as suggested to get "Fit Quality" into "Excellent" quality. I did that by trimming 2 seconds from begin and 2 seconds from end.

By doing so, it seems clear that d30pro has areas in pkmetric that is at audible level, especially capture 2? X16 on the other hand, no audible area, thus better in this regard?


d30pro USB capture1 from today vs Orig, right channel:

d30pro USB capture2 from today vs Orig, right channel:

X16 USB capture 1 from today vs Orig, Right channel:

X16 USB capture 2 from today vs Orig, Right channel:

 

[Pdxwayne]

@Blumlein 88 and @pkane , I did similar trimming exercise with Node2i Coax captures and got similar observations.

d30pro has points of potential audible areas (especially capture 1). X16 seems clearly better?

d30pro Coax capture1 vs Orig, right channel:

d30pro Coax capture 2 vs Orig, right channel:

x16 coax capture 1 vs Orig, Right channel:

X16 coax capture 2 vs Orig, right channel:

 

[b4nt]

@pkane @Pdxwayne I beleive I have f*cked my DW settings.

 

[pkane]

@pkane @Pdxwayne I beleive I have f*cked my DW settings.

View attachment 144060


View attachment 144061

Your settings seem OK. Just click on Reset Axis button near top right to set the scale correctly on the PK Metric chart. Looks like you've zoomed out way too far on the vertical axis.

 

[b4nt]

@pkane And I cannot play the match, I get an error message. I reset things, then un/reinstalled DW. Don't find a way to fix that.

 

[pkane]

@pkane And I cannot play the match, I get an error message. I reset things, then un/reinstalled DW. Don't find a way to fix that.

View attachment 144064

What's the error message?

 

[b4nt]

@pkane Better so. Error message being attached. Who is user ypa?

 

[b4nt]

@pkane And I'm now wondering what could be audible, spectrum of delta being so deep at -80 to -100dB (and mostly spread in lower FR).

 

[pkane]

@pkane Better so. Error message being attached. Who is user ypa?

Click on Help->Logging->View Log menu.

Then copy the text near the end of the file with the error text and post it here.

 

[pkane]

@pkane And I'm now wondering what could be audible, spectrum of delta being so deep at -80 to -100dB (and mostly spread in lower FR).


View attachment 144073

If you want to hear the difference, click on File->Audio Data->Save Delta... file and save it to a WAV file you can play in any player software.

 

[b4nt]

@pkane That is now with a HP filter at 1k (tried other values before). I think there is noise in the diff for lower FR (down to non audible range).

What shall be so lound around second 34? Are those the clapping people?

 

[b4nt]

@pkane I got my sound back. Do not understand what happend. Second 34 seems to be the clapping people (noise, a lot). And I'm wondering if a lot of the audible diffs shown by PK metrics could be related to the lower FR, and related to the hiss in that track. Are all DACs rendering hiss exact the same? What is the incidence off hiss (all that noise) in those measurements and analysis?

 

[pkane]

@pkane I got my sound back. Do not understand what happend. Second 34 seems to be the clapping people (noise, a lot). And I'm wondering if a lot of the audible diffs shown by PK metrics could be related to the lower FR, and related to the hiss in that track. Are all DACs rendering hiss exact the same?

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).