What is the cause of "digital glare"?

[Willem]

Or they are unable to even hear them.

 

[Killingbeans]

Most likely, yes

 

[jsrtheta]

In audio, you'll have some critics saying that:
- "this class D amp has digital glare"
- "this ESS DAC chip has Sabre glare"

I wonder what is glare, what causes it and would this glare be related to harmonic distortion that causes the treble to lack control or else?

Thank you!

I never heard of "digital glare" until I bought my first Stereophile. Then I heard it everywhere. Even in my sleep.

 

[JustIntonation]

There is no such thing. It is a made up thing by people who don't know how to perform controlled listening tests. If they did that, they would not hear any such thing!

As far as I can tell you cannot make such a statement in general with "scientific" confidence?
I do know how to perform a controlled listening test. And while I have not done this perfectly with instant switching between DACs I have tried to do it best as I could practically with a 10sec or so interval between two DACs and I could certainly hear important differences to me in such tests which I'm confident would also show up with an instant blind ABX test (which I've also performed where practical in other areas). And this between DACs which measure very well.

As for the ESS K2M Q2M etc variants of their DACs. They all seem to show the IMD "hump". As far as I know the level of this hump falls in a range that is below a leniently set treshold based on (imperfect not really scientific) audibility tests, but above a more strict audibility threashold shown by tests.
So as far as I understand it there is no conclusive evidence that says the ESS hump in their mobile range of DAC chips is below audibility?
(see for instance this post: https://www.audiosciencereview.com/...-thresholds-of-amp-and-dac-measurements.5734/ )

 

[jsrtheta]

As far as I can tell you cannot make such a statement in general with "scientific" confidence?
I do know how to perform a controlled listening test. And while I have not done this perfectly with instant switching between DACs I have tried to do it best as I could practically with a 10sec or so interval between two DACs and I could certainly hear important differences to me in such tests which I'm confident would also show up with an instant blind ABX test (which I've also performed where practical in other areas). And this between DACs which measure very well.

As for the ESS K2M Q2M etc variants of their DACs. They all seem to show the IMD "hump". As far as I know the level of this hump falls in a range that is below a leniently set treshold based on (imperfect not really scientific) audibility tests, but above a more strict audibility threashold shown by tests.
So as far as I understand it there is no conclusive evidence that says the ESS hump in their mobile range of DAC chips is below audibility?
(see for instance this post: https://www.audiosciencereview.com/...-thresholds-of-amp-and-dac-measurements.5734/ )

What you're describing is not really a controlled test. I note that here and elsewhere you argue for the invalidity of DBT. Good luck with that.

 

[JustIntonation]

What you're describing is not really a controlled test. I note that here and elsewhere you argue for the invalidity of DBT. Good luck with that.

Oh you must misunderstand me. I do not argue the invalidity of DBT. I only said that being more familiar with the system / having a lot of time to learn what to listen for would in some circumstances lead to more passed DBT. But this was in a different discussion about a different subject.
As for my own DAC listening tests. Well there wasn't really any other way. Had to switch cables, simple as that. Did it as quickly as possible and keeping my head in the exact same position etc. And was absolutely sure I'd pick out the differences easily had it been a truly blind test. I've picked out much more subtle differences in actual blind tests with ABX on foobar.

 

[gvl]

Well folks, after spending some time with the Khadas Tone Board I tend to think that "Sabre Glare" is a thing. I know the frequency response is flat, but, anyone seen a peak spectrogram comparison for real music music between ESS and other chips? VW diesels did test clean after all.

 

[March Audio]

And how did you test this?

If the FR is flat so will the spectrogram.

Btw the diesel test results were accurate. The engine management system adopted different settings, I gather when no steering input was detected, ie on a rolling road. Also by recognising the fixed driving cycle of the test. (that could be an urban myth but seems reasonable)

 

[pkane]

Well folks, after spending some time with the Khadas Tone Board I tend to think that "Sabre Glare" is a thing. I know the frequency response is flat, but, anyone seen a peak spectrogram comparison for real music music between ESS and other chips? VW diesels did test clean after all.

Don't have Khadas Board, but do have an ESS9028pro DAC. It never sounded like it had digital glare to me.

Here's the spectrum of the difference of the same track being played over ESS and an R2R DACs (mostly around -110dB):

And the spectrograms, for both (hard to see any differences):

ESS9028pro:

R2R DAC:

 

[gvl]

And how did you test this?

If the FR is flat so will the spectrogram.

Btw the diesel test results were accurate. The engine management system adopted different settings, I gather when no steering input was detected, ie on a rolling road. Also by recognising the fixed driving cycle of the test. (that could be an urban myth but seems reasonable)

There can be something in impulse response vs. static FR measurements. I want to see peak spectrograms, they should be more objective than a controlled subjective test, shouldn't they? I'd do it myself but no ADC.

 

[pkane]

There can be something in impulse response vs. static FR measurements. I want to see peak spectrograms, they should be more objective than a controlled subjective test, shouldn't they? I'd do it myself but no ADC.

Are those two spectrograms I posted not what you're looking for?

 

[gvl]

Don't have Khadas Board, but do have an ESS9028pro DAC. It never sounded like it had digital glare to me.

Here's the spectrum of the difference of the same track being played over ESS and an R2R DACs (mostly around -110dB):

View attachment 22189

And the spectrograms, for both (hard to see any differences):

ESS9028pro:
View attachment 22190
R2R DAC:
View attachment 22192

Convincing. There's a theory that not all Sabre DAC implementations have the glare. And Sabre Glare isn't the glare in the general negative sense, but more like emphasis on certain mid to high frequencies and details. It isn't necessarily a bad thing, just sightly different sound. This is my first Sabre based DAC, so who knows maybe it's a brain thing.

 

[gvl]

Are those two spectrograms I posted not what you're looking for?

More or less, I guess spectrum of peaks like below is what I had in mind but the difference is probably same or better.

 

[March Audio]

There can be something in impulse response vs. static FR measurements. I want to see peak spectrograms, they should be more objective than a controlled subjective test, shouldn't they? I'd do it myself but no ADC.

My first response is before we go looking for ghosts lets establish if you really saw one, or did you just have the willies put up you by being in a dark basement

Have you performed a blind and accurately volume controlled listening comparison and consistently identified the Toneboard against a non ess dac?

 

[gvl]

My first response is before we go looking for ghosts lets establish if you really saw one, or did you just have the willies put up you by being in a dark basement

Have you performed a blind and accurately volume controlled listening comparison and consistently identified the Toneboard against a non ess dac?

I did not but it does sound distinctly different from the DACs that went through my hands. Yeah I know, no need to lecture me.... When have the time I'll do a controlled level-matched experiment but it may take a while before that happens. Here is a discussion that suggests that this "glare" may have to do with the transient response of ESS filters rather than frequencies (see KeithL's post down the 1st page): http://emotivalounge.proboards.com/thread/52163/sabre-dac-glare-frequency-range . The Sabre Glare may all be folklore, but it's interesting why many people agree it exists and describe it in similar terms.

 

[tmtomh]

Well folks, after spending some time with the Khadas Tone Board I tend to think that "Sabre Glare" is a thing. I know the frequency response is flat, but, anyone seen a peak spectrogram comparison for real music music between ESS and other chips? VW diesels did test clean after all.

They tested clean because VW cheated. The cars' control software switched operating parameters for the tests compared to normal driving.

 

[gvl]

They tested clean because VW cheated. The cars' control software switched operating parameters for the tests compared to normal driving.

That wasn't the best analogy, I agree.

 

[RayDunzl]

The Sabre Glare may all be folklore, but it's interesting why many people agree it exists and describe it in similar terms.

.

We just need @Blumlein 88 to run a ESSDAC->ADC->ESSDAC loop.

https://www.audiosciencereview.com/...8th-generation-digital-copy.6827/#post-151814

Any aberrations in the reproduction of the music source would be magnified* with each loop. Assuming the ADC doesn't magically undo the "problem" with each pass.

After a few loops the result (if it exists) should be unbearable.

---

*stick a mic out there and record your speaker playback, play that back and record it again and play that back. You'll soon understand the "magnification of error" part if you don't already.

 

[Blumlein 88]

.

We just need @Blumlein 88 to run a ESSDAC->ADC->ESSDAC loop.

https://www.audiosciencereview.com/...8th-generation-digital-copy.6827/#post-151814

Any aberrations in the reproduction of the music source would be magnified* with each loop. Assuming the ADC doesn't magically undo the "problem" with each pass.

After a few loops the result (if it exists) should be unbearable.

---

*stick a mic out there and record your speaker playback, play that back and record it again and play that back. You'll soon understand the "magnification of error" part if you don't already.

I was of course using a March Audio DAC with an ESS 9038. And most comments on hearing a difference (I'm mum on whether they were correct) have been of the variety that the 8th generation copy was softer and took away sparkle, air and dulled the edge of transients. Hmmmm, doesn't sound like a description of 800 % elevated glare does it? Removing sighted bias makes a heck of a difference.

Or to paraphrase Rick James, "sighted listening is a helluva drug!"

 

[andreasmaaan]

I did not but it does sound distinctly different from the DACs that went through my hands. Yeah I know, no need to lecture me.... When have the time I'll do a controlled level-matched experiment but it may take a while before that happens. Here is a discussion that suggests that this "glare" may have to do with the transient response of ESS filters rather than frequencies (see KeithL's post down the 1st page): http://emotivalounge.proboards.com/thread/52163/sabre-dac-glare-frequency-range . The Sabre Glare may all be folklore, but it's interesting why many people agree it exists and describe it in similar terms.

To my knowledge the Sabre chips don't use different filters from other major DAC chips on the market. But in any case, these filters (NOS aside) have no effect in terms of either amplitude or phase response on the frequency band associated with detail (4-7KHz).