Listening test of 2 power amplifiers - files recorded for download - disclosed

[Balle Clorin]

I can fully agree here. I am not only a fan of my loudspeakers (I have owned them for 30 years) but also a fan of Avantgarde Acoustic. They can bring out such differences the best.

Strange , I think they (Avantgarde)sound very strange and wrong , but maybe all others are just wrong.…or maybe I heard them under wrong circumstances

 

[Blumlein 88]

@pkane @Blumlein 88 May you see what that tool set all features, and how deep it goes in details? Clicking around, I didn't so far find an evident diff between witch1/2.

https://www.sonicvisualiser.org/ (Devised and developed in the Centre for Digital Music, Queen Mary University of London)

This is using BBC tools, for energy plots, I beleive. You would better know how to use and read that. Output seem here different, but for a "valley" plot:

View attachment 147281

I've used this software in the past. It is showing you waveform superimposed upon a spectrogram in the above instance. I'm not sure I see what it would show versus viewing spectrogram 1 and 2 in DW. It is easier of course to scroll around zooming in or out on various segments of the file being examined.

 

[pkane]

@pkane it seems that using a synchronized DAC and ADC in these recordings gives better results in DeltaWave. I assume because then any phase difference in the recordings is an exact number of samples. Before trying any of this I assumed this wouldn't matter too much since the software can deal shifting a fraction of a sample, but in fact it does seem to matter. But maybe I'm not understanding what is happening. Do you have thoughts on this?

@pma your DAC and ADC are synchronous, correct?

@Langston Holland your DAC and ADC are not synchronous, correct?

DeltaWave adjusts phase differences/delay to less than 1/1000 of a sample, so no, whole sample match isn't required. But, DeltaWave doesn't automatically correct for non-linear clock differences, including jitter and other signals modulating the clock. These errors can be removed by using a common clock between the DAC and ADC. You can also turn on the non-linear EQ in DeltaWave as it corrects for some of these errors, also.

 

[b4nt]

@pkane @Blumlein 88 Using another plugin set, I could plot a diff, for tonal changes. But i do not know if this is audible.

https://vamp-plugins.org/plugin-doc/qm-vamp-plugins.html#qm-tonalchange
Tonal Change Detection Function – A function representing the estimated likelihood of a tonal change occurring in each spectral frame.
Tonal Change Positions – The resulting estimated positions of tonal changes.

 

[b4nt]

I'm not sure I see what it would show versus viewing spectrogram 1 and 2 in DW.

The version I am using allows to plot either means, peaks or valley for spectrogram. I'm not knowing exactly what his does.

 

[Blumlein 88]

The version I am using allows to plot either means, peaks or valley for spectrogram. I'm not knowing exactly what his does.

I just installed the latest version to see if it has any new features. I'm not seeing anything helpful beyond what we already know about these files.

 

[b4nt]

I just installed the latest version to see if it has any new features. I'm not seeing anything helpful beyond what we already know about these files.

I was just checking so if those alternate tools (and some plugings) would show some differences, not only in energy/levels but also in tones and so.

BBC plugin now being of 19 Jun 2014, you may then know it.

 

[spacevector]

I have no preference. But I think the files sound a little different due to slight channel balance difference.

I got 25 out of 32. Probability of pure guessing is 0.00105120079 per https://stattrek.com/online-calculator/binomial.aspx.


First try 5/8:
View attachment 146857

Second 5/8:
View attachment 146858


3rd try 7/8:
View attachment 146859


last try 8/8:
View attachment 146860

A true golden ear.

Did you do this on headphones or speakers, @Pdxwayne ?

 

[pjug]

But, DeltaWave doesn't automatically correct for non-linear clock differences, including jitter and other signals modulating the clock. These errors can be removed by using a common clock between the DAC and ADC. .

Thanks for this. Yes, common clock is what I should have said by synchronous. Maybe the bit about exact sample difference in phase matching is not important.

 

[Balle Clorin]

Track2 seem to have more punch and snap of the drums, enjoy that more than no 1, But I may be fooled by the fact that the second track listening to ALWAYS sound best( memory, recognizing effect)
Using Bose headphones, QC15

Frequency content in left channel overlaps completely for track 1 and 2 , no difference

Inverting left channel track 1 and paste-mixing it with left channel track2 gives a residual that sound like the original track just much lower in level, does that means the levels are not exactly the same?. Statistical analysis shows about 0.1db difference

 

[b4nt]

A true golden ear.

Did you do this on headphones or speakers, @Pdxwayne ?

@Pdxwayne But the soft I used to check doesn' detect a balance diff? @pma may later confirm if one of its two amps is imbalanced.

 

[Pdxwayne]

A true golden ear.

Did you do this on headphones or speakers, @Pdxwayne ?

The chain is laptop to Topping E30 to Topping L30 to AKG K371.

I can't really use speakers setup as it would drive my family members crazy.
: )

 

[Pdxwayne]

@Pdxwayne But the soft I used to check doesn' detect a balance diff? @pma may later confirm if one of its two amps is imbalanced.

He did confirm.

See https://www.audiosciencereview.com/...files-recorded-for-download.25646/post-876183

 

[b4nt]

He did confirm.

See https://www.audiosciencereview.com/...files-recorded-for-download.25646/post-876183

So this isn't a diff in sounds or tones

 

[Pdxwayne]

So this isn't a diff in sounds or tones

There could be diff in sounds or tones .....Maybe my chain and my ears are not good enough to distinguish that.
: )

 

[b4nt]

There could be diff in sounds or tones .....Maybe my chain and my ears are not good enough to distinguish that.
: )

pma added an amp may have a phase response diff. I'm not often enough awaken to detect a sound reaches me some ms late

 

[pkane]

The same problem, Audacity only shows sample values and not the reconstructed wave that would correspond with analog output of the amplifier. Remember that the analog signal can be completely reconstructed of its samples if the sampling frequency is at least two times higher than the highest signal frequency. There is no possible discussion on validity of Nyquist theorem to me.

This is the 13kHz sine generated at 96kHz sampling frequency. Audacity would only show the sample values and no sine shape. DW would show sample values interconnected by straight lines and not the sine shape. Anyway the generated sine is perfect at the good DAC output.

View attachment 146905

That made me think if it would be useful to show a sinc-interpolated waveform when zoomed in... Here's what my first attempt looks like (blue is the original DW display of samples connected by lines, white is the interpolated waveform. One large horizontal division is one sample:

 

[Blumlein 88]

That made me think if it would be useful to show a sinc-interpolated waveform when zoomed in... Here's what my first attempt looks like (blue is the original DW display of samples connected by lines, white is the interpolated waveform. One large horizontal division is one sample:

View attachment 147328

It would be nice to have this.

 

[pma]

All - I am out till Sunday. I will reply the posts then .

 

[Langston Holland]

@Langston Holland
Have you already had experience with the following test suite for your device?:
sound_quality_algorithms

This is news to me - thank you! I just sent him an email asking if I can download it after the site seemed to indicate it was going to mail it to me from Germany! : )