Searching for glare...

[Phelonious Ponk]

I thought I'd bring the "PCM glare" discussion over here where the grown-ups can discuss it without being hit by spitballs. Here's what I see/hear so far...

When asked to describe this "glare," most people, even subjectivists, describe harshness in the upper mids and trebles, compounded by complexity in that range, i.e.: choral vocals, string sections, etc. When played back digital rips of vinyl, many subjectivists even seem to agree that digital captures the vinyl sound, and when listening blind...well, I don't recall anyone consistently differentiating the rip from the record, but that could be my photogenic memory at work.

Anecdotal, to be sure, but enough to reinforce what I would rather believe (how's that for brutally honest?), that there is no PCM glare. There is plenty of glare to be found, for sure, but I think it is coming from recordings, crossovers, tweeters...you know, the much more obvious and verifiable stuff than a distortion inherent in digital that somehow cannot be measured. And speaking of measured, if this glare is harshness in the upper mids and trebles, wouldn't it show up in that range in a FR sweep?

Now, I'll wait to hear from those who have the knowledge to provide better than anecdotal evidence and simple logic.

Tim

 

[Thomas savage]

Yes i agree, much less likely to be some mystical happening in the DAC, much more likely to be loud HF being bounced about your room and hanging around messing up the next hit of the symbol or what ever. same for high keys on a piano its can sound loud and brittle live when you sat next to the piano. record it and put it through your system and without some consideration of your room acoustically it will sound like glare. how real do we want our systems? sit 2 feet away from a loud steal string guitar, it can be hard to take.

i am sure the way these frequency ranges get captured and transfer can help, if the recording/mastering is done in a very dead room you can end up with louder HF than you want but i am not recording engineer so cant comment on that.

so i think its more how the whole chain including your room deals with these problem frequencies rather than a issue with PCM. with AC power being a possible other contributor but that's debateable .

 

[Blumlein 88]

You know most condensor microphones have at least a few peaks in the 8khz and above range. I think in some recordings that too could contribute to glare. Not to mention all the processing that goes on with recorded material. Some of those processes are non-linear and generate distortion along with what else they do. As already mentioned I have done rips of LP or reel tape and it sounds like the source it came from. So glare is not endemic to the medium of PCM digital.

On another note, a few of the renowned Mercury recordings have glare on loud parts. Sounds like the old tube gear being overdriven some. Definitely isn't a smooth and calm 'analog' like sound in those exceptionally loud portions.

 

[amirm]

With ruler flat frequency response, digital does not in any way contribute such a distortion. That ruler flat response also means everything is captured all the way up to and above our hearing range. Folks who want a mellow sound with emphasis in lower frequencies as captured by analog gear, may think there is a problem there. But there is none.

Now recordings, speakers and rooms are surely to blame. I rather call that harshness than glare.

 

[Phelonious Ponk]

With ruler flat frequency response, digital does not in any way contribute such a distortion. That ruler flat response also means everything is captured all the way up to and above our hearing range. Folks who want a mellow sound with emphasis in lower frequencies as captured by analog gear, may think there is a problem there. But there is none.

Now recordings, speakers and rooms are surely to blame. I rather call that harshness than glare.

Yes, harshness is probably the more universally understood term. Glare is, I suspect, a term adopted by audiophiles.

Tim

 

[tomelex]

When I listen to vinyl, there is a definite difference in the imaging and the mid to upper treble. However, that part is not glare but one thing to take into consideration IMO is that vinyl seldom reaches above 15 kHz (of actual recorded material, many folks playback adds stuff well above that) or so anyway, and thus vinyl is limited at the bottom and top end. My needle drops to 24/96, stored 24/96 and played back 24/96 did not reveal anything new added, even with headphone listening. Also, there is a definite loss of clarity in the lower bass in vinyl as well, just thought I should add that for completeness.

And folks our age are less likely to even hear stuff that high.

 

[Phelonious Ponk]

When I listen to vinyl, there is a definite difference in the imaging and the mid to upper treble. However, that part is not glare but one thing to take into consideration IMO is that vinyl seldom reaches above 15 kHz (of actual recorded material, many folks playback adds stuff well above that) or so anyway, and thus vinyl is limited at the bottom and top end. My needle drops to 24/96, stored 24/96 and played back 24/96 did not reveal anything new added, even with headphone listening. Also, there is a definite loss of clarity in the lower bass in vinyl as well, just thought I should add that for completeness.

And folks our age are less likely to even hear stuff that high.

You're too kind. I'm sure I haven't heard 15khz in a decade or more.

Tim

 

[Opus111]

. And speaking of measured, if this glare is harshness in the upper mids and trebles, wouldn't it show up in that range in a FR sweep?

Nope - a frequency sweep is done with a sinewave as stimulus, the 'glare' is likely IMD (intermodulation distortion). Added to that its level isn't going to be high enough to affect the level measurably. If its fairly severe it could be -60dB which contributes 0.01dB to the level.

 

[March Audio]

I also think this "glare" has nothing to do with digital.

A bit of fun, below are two test tracks for you to download. One is the original file, one has been re-recorded. Played back by my MDAC and recorded by my TI PCM4222EVM.

So, two opportunities to add digital glare to the recording, from the DAC and the ADC.

Which do you think is which?

http://gofile.me/2vnEF/l5d8GJti7

Have fun

 

[Opus111]

I loaded them both up into Audacity and noticing they weren't the kind of music I prefer to listen to by the heavy compression apparent, looked at the artifacts in the lead-out. They were higher in level on 19 than 20 so I conclude that 19 is the copy.

 

[March Audio]

Thanks for spending your time Opus, but I think you have somewhat missed the point.

Was one of the tracks suffering from more "glare" than the other?

Did the process of digital playback and recording create a sense of glare?

 

[Opus111]

You'd need to repeat the test with some more demanding music - by that I mean with much higher crest factor - for me to be interested in listening.

Glare is most likely to be some form of IMD and IMD gets worse with more discrete tones present. So high crest factor music is the acid test for it.

 

[March Audio]

Can you make a suggestion for a musical choice that would fulfil your criteria?

You can upload a track of your choice to my server if you want. I can PM the details if you are interested.

 

[Opus111]

How would I go about uploading a track? My music is all 44k1 and I noticed your samples were 96k - so will you sample rate convert my music before playing it out through your MDAC?

I'll choose a track with choir and violins from this CD : www.amazon.com/Handel-Messiah-G-F/dp/B002MT3BXA/ref=sr_1_7?s=music&ie=UTF8&qid=1457006116&sr=1-7

 

[March Audio]

I can record at 44, 48, 88, 96, 176, 192kHz and 16 or 24 bit.

That was just a 96kHz 24 bit track I sort of randomly chose for the above test. I dont perform SR conversion, just play the tracks as is.

I will PM you a log in for my server, very simple.

 

[March Audio]

Hi Opus, sent a PM with my NAS details.

 

[Opus111]

Thanks for the help - I've uploaded a 19MB FLAC file from the CD I linked to now.

 

[March Audio]

Cool, Ill re-record it later today

 

[Opus111]

According to your recent post on WBF I'm not going to be able to discern any lower level details due to my room. So should I only evaluate your recording on headphones?

 

[March Audio]

Hi Opus, not quite what I said .

Use both.