The sound quality of DSD?

[SIY]

Do you have a level-matched double blind comparison? When I did this with my mic feed some years back, neither I nor the two people I was with could distinguish direct mic feed from PCM using basic controls.

 

[analogSurviver]

Do you have a level-matched double blind comparison? When I did this with my mic feed some years back, neither I nor the two people I was with could distinguish direct mic feed from PCM using basic controls.

No , I do not have level matched double blind comparison.

ABX using Foobar2000 does NOT allow comparing PCM and DSD files directly; it requires PCM file in order to perform ABX, which necessitates first converting the DSD into PCM - which defeats the very purpose of this particular ABX.

There are other caveats when comparing DSD to PCM in ABX/DBT - well described in previous pages of this thread. Depending on both hardware and software, DSD file level can vary - up to 6 dB - compared to the corresponding PCM file. Even if I did prepare corresponding PCM and DSD files level matched for MY equipment, that would in no way guarantee the levels would remain matched on any set of hardware and software used by others.
I have yet to see hardware and software capable of ABX/DBT of PCM and native DSD using single DAC that allows for good enough level matching and can make switch fast enough and without any clicking/hiccups when changing from PCM to DSD and/or vice versa.
The only way to do it properly would be two computers ( or a single comp with one "zone" each for PCM and DSD), two DACs ( at least one of them would have to be capable of native DSD playback ) , level matched to within less than 0.2 dB, both trough real hardware ABX box.

 

[SIY]

No , I do not have level matched double blind comparison.

ABX using Foobar2000 does NOT allow comparing PCM and DSD files directly; it requires PCM file in order to perform ABX, which necessitates first converting the DSD into PCM - which defeats the very purpose of this particular ABX.

There are other caveats when comparing DSD to PCM in ABX/DBT - well described in previous pages of this thread. Depending on both hardware and software, DSD file level can vary - up to 6 dB - compared to the corresponding PCM file. Even if I did prepare corresponding PCM and DSD files level matched for MY equipment, that would in no way guarantee the levels would remain matched on any set of hardware and software used by others.
I have yet to see hardware and software capable of ABX/DBT of PCM and native DSD using single DAC that allows for good enough level matching and can make switch fast enough and without any clicking/hiccups when changing from PCM to DSD and/or vice versa.
The only way to do it properly would be two computers ( or a single comp with one "zone" each for PCM and DSD), two DACs ( at least one of them would have to be capable of native DSD playback ) , level matched to within less than 0.2 dB, both trough real hardware ABX box.

Start by comparing mic feed to PCM. If they are indistinguishable, you’re done. That’s why I haven’t bothered with other formats.

 

[Frank Dernie]

Start by comparing mic feed to PCM. If they are indistinguishable, you’re done. That’s why I haven’t bothered with other formats.

That is what I did and with the same result. 16/48 recorder output was indistinguishable from the mic feed for me.
My only caveat was maybe my frame of reference was sloppy since I was used to hearing the marked difference there was between the output of the reel-to-reel recorder and microphone feed.

 

[mansr]

Not a single member - at least in this thread - has not even hinted at the fact that she/has ever had access to the live microphone feed and consequent monitoring of both PCM and DSD recording.

That is irrelevant. The great thing about science is that the results of a well-conducted experiment are valid for everybody, even if those who were not there. The limitations of human hearing are well-established, and 24-bit 48 kHz PCM is more than enough to capture everything anyone could possibly perceive. CD quality is cutting it quite close, and there may be corner cases where it isn't fully transparent, though I'd be surprised to find one in an acoustic recording.

 

[watchnerd]

www.NativeDSD is the place to check. As an example:
https://www.nativedsd.com/albums/40020-gustav-mahler-das-lied-von-der-erde

The patient may be thin, but still breathing!

 

[watchnerd]

I am doing it - constantly.

How are you editing the recording without converting to PCM?

My copy of Pro Tools doesn't allow for direct DSD editing via software-only / DAW.

 

[analogSurviver]

How are you editing the recording without converting to PCM?

My copy of Pro Tools doesn't allow for direct DSD editing via software-only / DAW.

I never said DSD can be edited without converting to some form of PCM. The right way to do it is to convert only the smallest possible amount around the intended split - and NOT the entire file.
There are two ways of doing it in different manner - Sonoma and Pyramix.

And since about two or three years ago, there is a software that can determine the precise % of portion of DSD that has gone trough any PCM conversion during its way from raw native DSD recording to master.
100 % ( DSD converted to PCM, mastered as usual, converted back to DSD ) is now considered no-ball.

 

[watchnerd]

I never said DSD can be edited without converting to some form of PCM.

I thought these days it could be done with an out-board interface, like the Avid MTRX.

Hence my question.

 

[analogSurviver]

That is irrelevant. The great thing about science is that the results of a well-conducted experiment are valid for everybody, even if those who were not there. The limitations of human hearing are well-established, and 24-bit 48 kHz PCM is more than enough to capture everything anyone could possibly perceive. CD quality is cutting it quite close, and there may be corner cases where it isn't fully transparent, though I'd be surprised to find one in an acoustic recording.

I beg to differ. If somebody has been there while making the recording and another person has been not, then these two individuals do NOT have the same insight - no matter how well a well conducted experiment is performed. By very definition, the experiment will have to rely on the recording - be it analog or digital, it is one step removed from the original.
And not the real thing - heard by only one of the two persons mentioned above.

Although the limits of human hearing may be well established, nothing of the sort can be said about human perception of sound beyond merely hearing trough ears. The studies regarding human perception of sound trough skin etc might well be in their infancy - and although this is strictly speaking no longer acoustics, it is a part of how people perceive sound. And will - eventually - be explained by branches of science other than acoustics, despite being triggered by acoustic source.

Since when have acoustic instruments lower frequency response extension than anything amplified and reproduced through an electromechanical transducer ? It is only in the last decade there are both mics and end transducers available
to cover approx 2 ( in word : two ) Hz to 100 kHz bandwidth , from recording to playback; and normally none of this stuff is used in any non-acoustic instruments amplification. So, anything amplified is likely to be limited to 10 Hz to approx 30 kHz range - and a great deal less in most real world examples.

The only metric where an amplified instrument can exceed acoustic is the sheer SPL - where 130 dB peaks are not that uncommon.

 

[SIY]

So it would be nice, then, if you could do an experiment with basic controls to see if you can distinguish a direct mic feed from PCM when level matched and double blind. That will give your claims credibility, and relieve you from the task of invoking supernatural mechanisms like “skin hearing.”

 

[mansr]

So it would be nice, then, if you could do an experiment with basic controls to see if you can distinguish a direct mic feed from PCM when level matched and double blind. That will give your claims credibility, and relieve you from the task of invoking supernatural mechanisms like “skin hearing.”

Obviously, you need the right cables for that skin effect to manifest.

 

[analogSurviver]

Start by comparing mic feed to PCM. If they are indistinguishable, you’re done. That’s why I haven’t bothered with other formats.

Transient response will always decide in this case. And PCM - unless ridiculosly high sample rate, like DXD (356 kHz, IIRC ) and above, will lose out to DSD - flat out.

It does depend on the capabilities of the overall system, though ; if amplifiers and end transducers ( be it headphones or speakers ) can not reproduce approx "flat" to
at least 40 kHz, most of the DSD advantage will not be reproduced and thus not audible.
But, this is not the deficiency of the DSD ...

 

[SIY]

Transient response will always decide in this case. And PCM - unless ridiculosly high sample rate, like DXD (356 kHz, IIRC ) and above, will lose out to DSD - flat out.

It does depend on the capabilities of the overall system, though ; if amplifiers and end transducers ( be it headphones or speakers ) can not reproduce approx "flat" to
at least 40 kHz, most of the DSD advantage will not be reproduced and thus not audible.
But, this is not the deficiency of the DSD ...

An experiment would be nicer than a handwave.

 

[mansr]

Transient response will always decide in this case. And PCM - unless ridiculosly high sample rate, like DXD (356 kHz, IIRC ) and above, will lose out to DSD - flat out.

Uh-huh.

 

[analogSurviver]

Obviously, you need the right cables for that skin effect to manifest.

It was meant sensitivity of HUMAN skin to ultrasonic sound - not skin effect in cables, which is rather minor.

Please try not to twist the words and use them entirely out of clearly specified context.

 

[SIY]

It was meant sensitivity of HUMAN skin to ultrasonic sound - not skin effect in cables, which is rather minor.

Please try not to twist the words and use them entirely out of clearly specified context.

It was a joke.

 

[watchnerd]

It was meant sensitivity of HUMAN skin to ultrasonic sound - not skin effect in cables, which is rather minor.

All the studies I've seen on skin hearing were within the audible spectrum.

 

[SIY]

All the studies I've seen on skin hearing were within the audible spectrum.

Not even that. The ones I’ve seen involve puffs of air from certain spoken consonants. I don’t think DSD can reproduce that!

 

[watchnerd]

Not even that. The ones I’ve seen involve puffs of air from certain spoken consonants. I don’t think DSD can reproduce that!

Well, I was assuming the McGurk effect puffs were audible.

I can get skin hearing by shoving my hand in my speaker's reflex port.