Spotify at full CD resolution

[lucretius]

11 hours ago, mevdinc said:

But no integration with 3rd party players such as Audivarna and Roon is a real downer, as I don't know how good their sound engine is.

 

I don't care about the Spotify app's "sound engine".  I just want it to deliver all the bits.

[Jud]

On 3/19/2021 at 11:54 AM, Samuel T Cogley said:

I love Qobuz, I really do.  BUT, their catalog always disappoints in comparison to Spotify or Apple Music, so I have two audio streaming subscriptions.

 

But the $64,000 question: will Spotify ever have a cross-platform way to deliver bit perfect output to a DAC a la Roon?  If it's a mobile-only solution, I'll stick with Qobuz and Apple Music.

 

I don't have a Spotify subscription. What artists/albums do you listen to that they have and Qobuz doesn't?

[Stereo]

2 hours ago, lucretius said:

 

Where have I heard these words before?

How would we know? Why don’t you tell us?

[Stereo]

1 hour ago, Jud said:

 

I don't have a Spotify subscription. What artists/albums do you listen to that they have and Qobuz doesn't?

Not speaking for Samuel but of the ones I’ve found so far that Spotify has that Qobuz doesn’t:

 

a-ha - Cast In Steel

a-ha - Foot Of The Mountain

(also missing a few singles by them)

 

The Donnas - The Donnas

The Donnas - American Teenage Rock ‘N’ Roll Machine

The Donnas - The Donnas Turn 21

 

Good Sleepy - Dessert Before Dinner

Good Sleepy - Okay With The Way It Is, As Is

Good Sleepy - Rest Easy

 

Magnace - two different versions of Her Name (Qobuz has only one of the three)

Magnace - So Far Away

Magnace - Do You

 

Real Steal Soundtrack

 

Eddie Murphy - How Could It Be

 

Amerie - 4AM Mullholland

Amerie - After 4AM

Amerie - A Heart’s For The Breaking

Amerie - Curious

 

Ric Ocasek - Beatitude

Ric Ocasek - Nexterday

 

The Pool - Dance It Down/Jamaica Running

 

Night Ranger - High Road

 

The Bulk Of Golden Earring’s catalog

 

Bob Seger & The Silver Bullet Band’s catalog

 

I’m sure there are several more but I haven’t gotten around to look yet.

[mevdinc]

2 hours ago, Jud said:

 

I don't have a Spotify subscription. What artists/albums do you listen to that they have and Qobuz doesn't?

Too many to mention. :)

[mevdinc]

3 hours ago, lucretius said:

 

I don't care about the Spotify app's "sound engine".  I just want it to deliver all the bits.

Software players receive/process the same bits just like all DACs, but surely they all sound a little bit different from one another.

[lucretius]

6 hours ago, mevdinc said:

Software players receive/process the same bits just like all DACs, but surely they all sound a little bit different from one another.

 

Umm ... the players deliver the bits to the DAC, which does the DA conversion.  Players that do not mangle the bits, i.e. that deliver a bit-perfect stream, all "sound" the same.

[sandyk]

17 minutes ago, lucretius said:

Players that do not mangle the bits, i.e. that deliver a bit-perfect stream, all "sound" the same.

Audiophile (jriver.com)

 

😈

[Jud]

8 hours ago, lucretius said:

 

Umm ... the players deliver the bits to the DAC, which does the DA conversion.  Players that do not mangle the bits, i.e. that deliver a bit-perfect stream, all "sound" the same.

 

I suppose, unless there is some difference in the player processing with regard to noise delivered to the DAC, for instance differences in the ways various players utilize memory. I believe something like this may have been the subject of @manisandher's "Red or Blue Pill" thread, where he was able to get 9 of 10 trials correct for one difference. (I didn't read the entire thread, just isolated snippets, so I'm not totally sure.)

 

But in any case, you have at least a couple of alternatives: Have the player "mangle" the bits and deliver them to the DAC's final analog filter as they came from the player; or have the DAC's internal digital filtering "mangle" them before delivering them to the DAC's final analog filter.

[manisandher]

1 hour ago, Jud said:

I believe something like this may have been the subject of @manisandher's "Red or Blue Pill" thread, where he was able to get 9 of 10 trials correct for one difference. (I didn't read the entire thread, just isolated snippets, so I'm not totally sure.)

 

Yeah, the important point was that the digital input to the DAC was captured in real-time and shown to have remained bit-identical throughout the ABX.

 

Different bit-identical players do sound different. Of this, there is absolutely no doubt in my mind. But if others hear them as all sounding the same, so be it.

 

Mani.

[pkane2001]

41 minutes ago, manisandher said:

 

Yeah, the important point was that the digital input to the DAC was captured in real-time and shown to have remained bit-identical throughout the ABX.

 

Different bit-identical players do sound different. Of this, there is absolutely no doubt in my mind. But if others hear them as all sounding the same, so be it.

 

Mani.

 

Bit identical can sound different when jitter is large enough or noise is large enough, there's no doubt. The red/blue pill test was predicated on an obscure software setting in XXHighEnd configured to an extreme value that even the author didn't recommend using. That, at least in theory, could have had an effect on jitter which wasn't measured.

[Jud]

6 minutes ago, pkane2001 said:

 

Bit identical can sound different when jitter is large enough or noise is large enough, there's no doubt. The red/blue pill test was predicated on an obscure software setting in XXHighEnd configured to an extreme value that even the author didn't recommend using. That, at least in theory, could have had an effect on jitter which wasn't measured.

 

Was the "obscure software setting" SFS?

[pkane2001]

7 minutes ago, Jud said:

 

Was the "obscure software setting" SFS?

 

Don't recall the details, I assume @manisandher does. I remember Peter trying to explain it, but failing. It seemed to have something do with chunking of data in memory, sizes of blocks read from disk, or sizes of blocks being output to audio, or something like that. At least on the surface this sounded like it could affect timing/jitter where the DAC depends on the input from the PC to derive sample clock.

[manisandher]

51 minutes ago, pkane2001 said:

 

Bit identical can sound different when jitter is large enough or noise is large enough, there's no doubt. The red/blue pill test was predicated on an obscure software setting in XXHighEnd configured to an extreme value that even the author didn't recommend using. That, at least in theory, could have had an effect on jitter which wasn't measured.

 

43 minutes ago, Jud said:

Was the "obscure software setting" SFS?

 

38 minutes ago, pkane2001 said:

 

Don't recall the details, I assume @manisandher does. I remember Peter trying to explain it, but failing. It seemed to have something do with chunking of data in memory, sizes of blocks read from disk, or sizes of blocks being output to audio, or something like that. At least on the surface this sounded like it could affect timing/jitter where the DAC depends on the input from the PC to derive sample clock.

 

It was indeed the 'SFS' setting in XXHighEnd. I think "chunking of data from memory" might be the most accurate description of what it adjusts.

 

The two settings we used in the ABX were SFS=0.69 vs. SFS=200. These may be on the "extreme" side from the software's perspective, but importantly, they are bit-identical to each other. (And in actual fact, I've used the 'extreme' SFS=0.69 for years.)

 

Peter's only concern was whether SFS=0.69 (being on the extreme side) had remained bit-identical throughout the ABX. The real-time digital capture proved that it had.

 

XXHighEnd's SFS setting affects every DAC I've ever owned/tried in a totally consistent way. It affects my RME ADI-2 Pro FS R in exactly the same way as it affected the DAC we used in the ABX. I don't think anyone would claim that the RME's USB input is in any way defective.

 

Mani.

[pkane2001]

1 minute ago, manisandher said:

 

 

 

It was indeed the 'SFS' setting in XXHighEnd. I think "chunking of data from memory" might be the most accurate description of what it adjusts.

 

The two settings we used in the ABX were SFS=0.69 vs. SFS=200. These may be on the "extreme" side from the software's perspective, but importantly, they are bit-identical to each other. (And in actual fact, I've used the 'extreme' SFS=0.69 for years.)

 

Peter's only concern was whether SFS=0.69 (being on the extreme side) had remained bit-identical throughout the ABX. The real-time digital capture proved that it had.

 

XXHighEnd's SFS setting affects every DAC I've ever owned/tried in a totally consistent way. It affects my RME ADI-2 Pro FS R in exactly the same way as it affected the DAC we used in the ABX. I don't think anyone would claim that the RME's USB input is in any way defective.

 

Mani.

Mani, it would be interesting to run  J-Test signal through these two settings. Jitter on SPDIF is not unusual and can be caused by severe timing errors in software.

[manisandher]

1 hour ago, pkane2001 said:

Mani, it would be interesting to run  J-Test signal through these two settings. Jitter on SPDIF is not unusual and can be caused by severe timing errors in software.

 

Paul, I don't think it'd help. If a big difference in jitter were measured, it'd just be claimed that the DAC used was crap and not fit for purpose.

 

Instead, I've just taken two captures with a mic of the output of the speakers in my office, using exactly the same XXHighEnd settings we used in the test:

 

In both cases, the playback is bit-identical. And yet, these captures sound clearly different... to me at least.

 

Mani.

[pkane2001]

1 minute ago, manisandher said:

 

Paul, I don't think it'd help. If a big difference in jitter were measured, it'd just be claimed that the DAC used was crap and not fit for purpose.

 

Instead, I've just taken two captures with a mic of the output of the speakers in my office, using exactly the same XXHighEnd settings we used in the test:

 

In both cases, the playback is bit-identical. And yet, these captures sound clearly different... to me at least.

 

Mani.

 

I understand, Mani, but bit-identical doesn't mean it's playing the same analog signal, as you know. The reason I'd like to run the tests is to try to understand the cause, not to claim one thing or another.

[manisandher]

Sure. It'll take me a while to set everything up again. I'll let you know when it's done so that we can run the J-test signal with the two settings.

 

Mani.

[pkane2001]

4 minutes ago, manisandher said:

Sure. It'll take me a while to set everything up again. I'll let you know when it's done so that we can run the J-test signal with the two settings.

 

Mani.

 

Thanks, Mani. Let's continue in your other thread, so we don't take this one off track.

[Jud]

4 hours ago, pkane2001 said:

 

Don't recall the details, I assume @manisandher does. I remember Peter trying to explain it, but failing. It seemed to have something do with chunking of data in memory, sizes of blocks read from disk, or sizes of blocks being output to audio, or something like that. At least on the surface this sounded like it could affect timing/jitter where the DAC depends on the input from the PC to derive sample clock.

 

 

Yep, sounds like SFS.  There have been many, many quite long threads on Peter's forum devoted to that "obscure" setting.

 

The hypothesis is that different noise profiles are created by the app fetching smaller chunks of data more often, versus bigger chunks less often.  There was a discussion several years ago between Gordon Rankin and John Swenson on Computer Audio Asylum about this.

 

Whether there's anything to it I don't know. I do note the DAC Mani used in the test had an async USB input.

[manisandher]

23 minutes ago, Jud said:

I do note the DAC Mani used in the test had an async USB input.

 

Hey Jud, no, not quite.

 

Of course, I would have preferred to use my Phasure DAC for the test, but Mans insisted that we capture the digital input to the DAC in real-time during the ABX (a good call, as it happened). At the time, the only way I could figure out how to do this was to use spdif from the audio PC with a BNC splitter - one output going to the DAC, the other going to the digital recorder. And the only DAC I had to hand with an spdif input was an Altmann NOS DAC. For the test, this was fed a 176.4-upsampled file.

 

If I were to repeat the test today, I'd simply use an async USB DAC with an spdif output. My RME ADI-2 Pro would fit the bill I think.

 

Mani.

[sandyk]

16 hours ago, sandyk said:

Audiophile (jriver.com)

 

😈

Putting aside the anomalies for a moment, caused by the obscure setting SFS in XXHE, did anybody note this reply from Chris as quoted by JRiver ?  These S/W player differences are far from unique to XXHE, with players such as JRiver sounding way better to many members than for example Foobar 2000, and even more so when played from System Memory.

 

Quote

"In January 2010 at CES I listened to a demo using JRMC 14 and was really pleased with the sound. JRMC was running on a Mac laptop with Boot Camp and Windows 7. I compared the sound to the OS X / iTunes partition on the same laptop and was surprised at how much better I like JRMC in that system. I was finally convinced I needed to take JRMC much more seriously. Shortly after arriving home from Las Vegas I arranged a meeting with the people at JRiver and started using the application exclusively."

 

[lucretius]

On 3/21/2021 at 9:51 AM, Jud said:

I suppose, unless there is some difference in the player processing with regard to noise delivered to the DAC, for instance differences in the ways various players utilize memory.

 

Must be a poor DAC that isn't properly isolated from computer noise.  Assuming that a particular DAC is subject to computer noise affecting audible output, then we cannot even ensure this specific DAC will play the same track twice with the same player software, and sound the same both times.  In any case, playing tracks is not a heavy load; there should be no practical difference from one player to the next; unless of course there exists some really horrible player software.  

 

[Currawong]

1 hour ago, lucretius said:

 

Must be a poor DAC that isn't properly isolated from computer noise.  Assuming that a particular DAC is subject to computer noise affecting audible output, then we cannot even ensure this specific DAC will play the same track twice with the same player software, and sound the same both times.  In any case, playing tracks is not a heavy load; there should be no practical difference from one player to the next; unless of course there exists some really horrible player software.  

 

 

Sucker for punishment I am, I'm going to reply.

 

The digital filter used can often distinctly affect the perception of the soundstage. The music waveforms show distinct time domain shifts. An easy way to hear it is with something like a Chord Hugo 2 or Qutest (or better) with their very long, time-domain correct filters, and compare to an iFi Pro iDSD using the very short GTO filter using a high-quality recording with a good stereo image. With the Chord, you get the correct soundstage image, left-to-right and front-to-back. With the iFi, everything is pushed up into your face. The difference is not trivial. 

 

It confuses me why I still see arguments about this when there is a whole science behind how sound is interpreted by the human brain and the effects are trivial to experience with modern digital electronics.

[lucretius]

22 hours ago, Currawong said:

 

Sucker for punishment I am, I'm going to reply.

 

The digital filter used can often distinctly affect the perception of the soundstage. The music waveforms show distinct time domain shifts. An easy way to hear it is with something like a Chord Hugo 2 or Qutest (or better) with their very long, time-domain correct filters, and compare to an iFi Pro iDSD using the very short GTO filter using a high-quality recording with a good stereo image. With the Chord, you get the correct soundstage image, left-to-right and front-to-back. With the iFi, everything is pushed up into your face. The difference is not trivial. 

 

It confuses me why I still see arguments about this when there is a whole science behind how sound is interpreted by the human brain and the effects are trivial to experience with modern digital electronics.

 

The discussion was about using different player software in bit-perfect mode and not about changing DACs or DAC filters.