The answer is.... WHY NOT ??

0 Members and 1 Guest are viewing this topic. Read 6897 times.

AmpDesigner333

  • Industry Participant
  • Posts: 2973
  • Detailed AND Musical
    • Digital Amplifier Company
The answer is.... WHY NOT ??
« on: 8 Apr 2019, 02:00 pm »
16-bit CD resolution reminds me of “surface noise” from vinyl, albeit an order of magnitude lower.  Feed your DAC 16 bits of a very quiet recording with the volume cranked and you’ll see/hear what I mean.  However, higher sample rate relaxes the requirements for digital domain filters and allows nice flat phase past 20kHz.  The debate is on, however, regarding the audibility of either resolution or sample rate (pet peeve — high resolution has nothing to do with sample rate) increases beyond 44/16 when playing music at normal listening levels.  However, MEMORY IS CHEAP!  Consider how much more memory video takes compared to audio.  Why is using more memory (or bandwidth) for music such a big deal in this day and age?  If that’s what it takes to be SURE you’re getting comfortably beyond audible conversion effects, why not ??

Freo-1

Re: The answer is.... WHY NOT ??
« Reply #1 on: 8 Apr 2019, 03:49 pm »

It's too bad that Sony's SACD output is proprietary.  Sony does allow a given vendor to develop a custom interface to mate between the transport and the DAC. 


The reason I bring this up is streaming SACD signals to the DAC is far and away the best sounding setup I've heard in the digital domain.  Case in point:  The McIntosh MCT transport via the DIN connection to a supporting preamp/DAC is among the very best I've heard from digital.  Easily surpasses the CD level on dual layer discs.  Additionally,  the DIN setup is superior to SPDIF on redbook CD's as well.  Jitter is not really an issue.


Another method of getting SACD streamed to a DAC is to employ a audio extractor from the source.  Select the SACD out to PCM, and the audio extractor will supply 22 to 24 bits at 88KHz to the DAC.  I've used this method to listen SACD's with a Devialet 400 with excellent results.  The Devialet does an outstanding job of re-clocking the audio stream, and at -133 noise floor, any hiss is strictly from the source recording.


IMHO, I think that the CD standard should have been 24 bit/48KHz.  That would have been sufficient for most applications.  Not sure how accurate this is, but the story going around at the time of CD release was that the 16 bit/44 KHz limit was the standard of getting Beethoven's 9th on a single CD.  The story goes that the initial spec was to be 48 KHz, but it exceeded the 660 MB limit of the CD. 

JohnR

Re: The answer is.... WHY NOT ??
« Reply #2 on: 8 Apr 2019, 09:33 pm »
You can stream hires from Qobuz.

AmpDesigner333

  • Industry Participant
  • Posts: 2973
  • Detailed AND Musical
    • Digital Amplifier Company
Re: The answer is.... WHY NOT ??
« Reply #3 on: 9 Apr 2019, 02:16 am »
You can stream hires from Qobuz.
Yes, now including the USA.  That’s as of Valentine’s Day 2019, so it’s still relatively unknown here.

AmpDesigner333

  • Industry Participant
  • Posts: 2973
  • Detailed AND Musical
    • Digital Amplifier Company
Re: The answer is.... WHY NOT ??
« Reply #4 on: 9 Apr 2019, 02:25 am »
It's too bad that Sony's SACD output is proprietary.  Sony does allow a given vendor to develop a custom interface to mate between the transport and the DAC. 


The reason I bring this up is streaming SACD signals to the DAC is far and away the best sounding setup I've heard in the digital domain.  Case in point:  The McIntosh MCT transport via the DIN connection to a supporting preamp/DAC is among the very best I've heard from digital.  Easily surpasses the CD level on dual layer discs.  Additionally,  the DIN setup is superior to SPDIF on redbook CD's as well.  Jitter is not really an issue.


Another method of getting SACD streamed to a DAC is to employ a audio extractor from the source.  Select the SACD out to PCM, and the audio extractor will supply 22 to 24 bits at 88KHz to the DAC.  I've used this method to listen SACD's with a Devialet 400 with excellent results.  The Devialet does an outstanding job of re-clocking the audio stream, and at -133 noise floor, any hiss is strictly from the source recording.


IMHO, I think that the CD standard should have been 24 bit/48KHz.  That would have been sufficient for most applications.  Not sure how accurate this is, but the story going around at the time of CD release was that the 16 bit/44 KHz limit was the standard of getting Beethoven's 9th on a single CD.  The story goes that the initial spec was to be 48 KHz, but it exceeded the 660 MB limit of the CD.
SACD/DSD is plagued with issues due to 1-bit modulation.  First of all, you can’t use a modulation index past 0.5 without stability issues.  Second, this is merely skipping the decimation step of a PCM converter, except that PCM modulators are multi-level, allowing deeper modulation and thus realized superior noise performance. Third, in order to edit DSD, you need to convert it to PCM, so you know conversion to PCM is essentially lossless.  Plus, try to find an un-edited SACD.  Fourth, the effective resolution of base rate DSD is only 6-bits at 20kHz!  It increases as the frequency goes lower due to the nature of the modulation. Every halving of the excitation frequency or doubling of the rate gives you one more effective bit, thus 4xDSD gets you 8 effective bits at 20kHz.  So, if you want the best performance audio end-to-end, go with high rate (96kHz/192kHz) 24-bit PCM.
« Last Edit: 9 Apr 2019, 04:14 am by AmpDesigner333 »

AmpDesigner333

  • Industry Participant
  • Posts: 2973
  • Detailed AND Musical
    • Digital Amplifier Company
Re: The answer is.... WHY NOT ??
« Reply #5 on: 9 Apr 2019, 04:26 am »
SACD/DSD is plagued with issues due to 1-bit modulation.  First of all, you can’t use a modulation index past 0.5 without stability issues.  Second, this is merely skipping the decimation step of a PCM converter, except that PCM modulators are multi-level, allowing deeper modulation and thus realized superior noise performance. Third, in order to edit DSD, you need to convert it to PCM, so you know conversion to PCM is essentially lossless.  Plus, try to find an un-edited SACD.  Fourth, the effective resolution of base rate DSD is only 6-bits at 20kHz!  It increases as the frequency goes lower due to the nature of the modulation. Every halving of the excitation frequency or doubling of the rate gives you one more effective bit, thus 4xDSD gets you 8 effective bits at 20kHz.  So, if you want the best performance audio end-to-end, go with high rate (96kHz/192kHz) 24-bit PCM.
I should also mention that pretty much every digital audio player these days (like JRiver) can convert DSD to PCM, and in this case, the realized, measurable audio performance from the DAC is better with 96/24 (or 192/24) than with native DSD format.  Take a look at DAC datasheets and notice the SNR/DNR and THD+N of PCM vs DSD.  Anyway, if you have SACDs, you can also rip them and play them back either way.

Freo-1

Re: The answer is.... WHY NOT ??
« Reply #6 on: 9 Apr 2019, 12:45 pm »

I've done a fair amount of comparison listening between SACD native and SACD/PCM, and native SACD almost always sounds better when played back through systems that support both types of data streams .  For audiophiles that rely on optical media, there is a (still) a good amount of SACD media out there, while DVD audio is dead. 




Freo-1

Re: The answer is.... WHY NOT ??
« Reply #7 on: 9 Apr 2019, 02:48 pm »

Found this article regarding DSD vs. PCM:


https://www.mojo-audio.com/blog/dsd-vs-pcm-myth-vs-truth/


The issue is not as clear cut as argued. 

AmpDesigner333

  • Industry Participant
  • Posts: 2973
  • Detailed AND Musical
    • Digital Amplifier Company
Re: The answer is.... WHY NOT ??
« Reply #8 on: 9 Apr 2019, 05:31 pm »
Found this article regarding DSD vs. PCM:


https://www.mojo-audio.com/blog/dsd-vs-pcm-myth-vs-truth/


The issue is not as clear cut as argued.
Once I got to the jitter* part, I was done with that "article", which is really a sales pitch -- selling way overpriced R2R DACs (antiquated low-performance topology) that don't even have specified THD+N or SNR ratings.  There's no mention of DSD's effective resolution issue!  See my comments earlier in this thread for the "engineering summary" of the difference.  It's brief because it's rather simple.  If you like DSD, fine, but with a truly high performance DAC, convert to PCM for better performance (both measured and heard).

*Well designed modern DACs reclock internally (like Cherry DAC DAC), so jitter should NOT have any effect!

If you're looking for a truly HIGH PERFORMANCE DAC under $1500, here you go:
https://www.cherryamp.com/dac-dac-d-to-a-convertor

"On a cost vs. sound quality basis (alone), Digital Amplifier Company DAC DAC may be the best digital to analog converter to come along in 3 decades; all the way back to the extremely costly & extraordinary $12,000 Stax DAC-X1t Vacuum Tube Output Reference D/A Processor (1989)!" -Jeremy Kipnis, of Kipnis Studios (KSS)™
http://positive-feedback.com/audio-discourse/digital-amplifier-company-dac-dac/

Freo-1

Re: The answer is.... WHY NOT ??
« Reply #9 on: 9 Apr 2019, 08:25 pm »

The article has a number of valid technical points. Don't you find it interesting that a large percentage of the digital recorded catalogue is stored as DSD?


I like both hi-res PCM and DSD.   Both formats have advantages and disadvantages.  Neither one is perfect.  I will say that I have a few 24/192 PCM recordings that are outstanding (especially when using a Devialet 400 to play it back).   


The following summary points are valid:
"High-resolution PCM and DSD formats of comparable resolution are statistically indistinguishable from one another in blind listening tests.
Pure DSD recordings, as pictured in the flow charts used in DSD marketing hype, are almost nonexistent. There are currently very few recording studios that have the ability to edit, mix, or master DSD. High-definition 5-bit and 8-bit PCM (Wide-DSD), are used in recording and post-production editing, mixing, and mastering of nearly all modern DSD recordings.
When a PCM file is played on a native DSD single-bit converter, the single-bit DAC chip has to convert the PCM to DSD in real-time. This is one of the major reasons people claim DSD sounds better than PCM, when in fact, it is just that the chip in most modern single-bit DACs do a poor job of decoding PCM.
DSD64 SACD has roughly 33 times the resolution of a 16-bit 44.1KHz Red Book CD, roughly the same resolution as 24-bit 96KHz PCM recording, and less than half the resolution of a 24-bit 192KHz PCM recording."
The quality of the recording plays a far more significant role than the format or resolution it is distributed in. To increase profits, modern recording studio executives insisted that errors be edited out in post-production, significantly compromising the quality of the original master tapes. "
Another point made was that the DAC should play back the stream in it's native format as opposed to conversion real time.  IMHO, I do not blindly buy the fact that DSD will sound better all the time as PCM.  I would recommend that you listen to a McIntosh setup where the transport is connected to the DAC via the DIN cable passing DSD.  It's VERY good, and a lot better than CD playback. 
There is always some "sales pitch" tied to these types of threads.  Have to separate the wheat from the chaff.  Audio is 80% science and engineering, and 20% art.  We will always argue about the 20% that is art.  8)

AmpDesigner333

  • Industry Participant
  • Posts: 2973
  • Detailed AND Musical
    • Digital Amplifier Company
Re: The answer is.... WHY NOT ??
« Reply #10 on: 11 Apr 2019, 02:35 am »
The article has a number of valid technical points. Don't you find it interesting that a large percentage of the digital recorded catalogue is stored as DSD?


I like both hi-res PCM and DSD.   Both formats have advantages and disadvantages.  Neither one is perfect.  I will say that I have a few 24/192 PCM recordings that are outstanding (especially when using a Devialet 400 to play it back).   


The following summary points are valid:
"High-resolution PCM and DSD formats of comparable resolution are statistically indistinguishable from one another in blind listening tests.
Pure DSD recordings, as pictured in the flow charts used in DSD marketing hype, are almost nonexistent. There are currently very few recording studios that have the ability to edit, mix, or master DSD. High-definition 5-bit and 8-bit PCM (Wide-DSD), are used in recording and post-production editing, mixing, and mastering of nearly all modern DSD recordings.
When a PCM file is played on a native DSD single-bit converter, the single-bit DAC chip has to convert the PCM to DSD in real-time. This is one of the major reasons people claim DSD sounds better than PCM, when in fact, it is just that the chip in most modern single-bit DACs do a poor job of decoding PCM.
DSD64 SACD has roughly 33 times the resolution of a 16-bit 44.1KHz Red Book CD, roughly the same resolution as 24-bit 96KHz PCM recording, and less than half the resolution of a 24-bit 192KHz PCM recording."
The quality of the recording plays a far more significant role than the format or resolution it is distributed in. To increase profits, modern recording studio executives insisted that errors be edited out in post-production, significantly compromising the quality of the original master tapes. "
Another point made was that the DAC should play back the stream in it's native format as opposed to conversion real time.  IMHO, I do not blindly buy the fact that DSD will sound better all the time as PCM.  I would recommend that you listen to a McIntosh setup where the transport is connected to the DAC via the DIN cable passing DSD.  It's VERY good, and a lot better than CD playback. 
There is always some "sales pitch" tied to these types of threads.  Have to separate the wheat from the chaff.  Audio is 80% science and engineering, and 20% art.  We will always argue about the 20% that is art.  8)
Just wanted to mention that “playback in the original format” is neither necessary nor recommended if you have a high performance DAC.  By high performance, I expect more than 120dB SNR and less than 0.001% THD+N at FS.  Fancy interfaces don’t change the data and are also unnecessary. Jitter should be “fixed” in the DAC as well. There’s nothing wrong with SPDIF, especially the coax version.  Lastly, the math concerning EFFECTIVE resolution is simple.  The numbers above are incorrect, and to me represent marketing hyperbole rather than engineering. Just calculate how many samples at the Nyquist frequency. Since the data is one bit for DSD, there’s your resolution!  1024xDSD would be fine, but what’s the point when PCM does even better with less storage requirements, plus at 1024x, the clock speeds get ridiculous !

Freo-1

Re: The answer is.... WHY NOT ??
« Reply #11 on: 11 Apr 2019, 12:50 pm »

The fancy interfaces can and do work very well and do provide a improvement.  They cost more to implement,  Again, as an example, if one actually listens to a demonstration of the McIntosh transport/DAC setup via the DIN, it is pretty easy to notice the improvement with Redbook CD playback using the DIN over SPDIF, while SACD sounds outstanding with the well mastered recordings. 


Since SACD optical media is still readily available, a method to play the media back in its native format is desirable.  If all the media was available as hi-res PCM, then that would be ideal.  Since it's not, SACD is a very viable option for hi res playback, and since audiophiles can't (allegedly) hear a difference in double blind testing between hi res PCM and SACD,  SACD remains a viable option for hi-res playback. 



Converting SACD to PCM when played back on a Devialet sounds incredible, while other setups with different hardware demonstrates that SACD played back in native format does sound better than SACD/PCM.  There is no absolute one is better than the other here.


Here is some more information regarding DSD vs. PCM:


https://headfonics.com/2018/02/dsd-vs-pcm-real-competitors/


Check out the other links provided in the main article.  Also found this from the Arye Audio website:


 
At the highest levels of audio engineers there has been little consensus as to which format is better, but most audiophiles have had first-hand experience with SACDs generally sounding better than DVD-Audio disc. So the prospect of being able to play downloadable DSD files via computer has led to a great deal of excitement, even controversy in the audiophile community.


Read more at https://www.audiostream.com/content/ayres-pcm-dsd-comparison#F0omWVFPqkad08mu.99


So, here's a summary so far:


1) SACD and Hi-Res PCM both sound better than redbook CD's


2) There is some level of disagreement as to merits of hi-res PCM vs. DSD. 






« Last Edit: 11 Apr 2019, 07:07 pm by Freo-1 »

AmpDesigner333

  • Industry Participant
  • Posts: 2973
  • Detailed AND Musical
    • Digital Amplifier Company
Re: The answer is.... WHY NOT ??
« Reply #12 on: 20 Apr 2019, 07:54 pm »
The fancy interfaces can and do work very well and do provide a improvement.  They cost more to implement,  Again, as an example, if one actually listens to a demonstration of the McIntosh transport/DAC setup via the DIN, it is pretty easy to notice the improvement with Redbook CD playback using the DIN over SPDIF, while SACD sounds outstanding with the well mastered recordings. 


Since SACD optical media is still readily available, a method to play the media back in its native format is desirable.  If all the media was available as hi-res PCM, then that would be ideal.  Since it's not, SACD is a very viable option for hi res playback, and since audiophiles can't (allegedly) hear a difference in double blind testing between hi res PCM and SACD,  SACD remains a viable option for hi-res playback. 



Converting SACD to PCM when played back on a Devialet sounds incredible, while other setups with different hardware demonstrates that SACD played back in native format does sound better than SACD/PCM.  There is no absolute one is better than the other here.


Here is some more information regarding DSD vs. PCM:


https://headfonics.com/2018/02/dsd-vs-pcm-real-competitors/


Check out the other links provided in the main article.  Also found this from the Arye Audio website:


 
At the highest levels of audio engineers there has been little consensus as to which format is better, but most audiophiles have had first-hand experience with SACDs generally sounding better than DVD-Audio disc. So the prospect of being able to play downloadable DSD files via computer has led to a great deal of excitement, even controversy in the audiophile community.


Read more at https://www.audiostream.com/content/ayres-pcm-dsd-comparison#F0omWVFPqkad08mu.99


So, here's a summary so far:


1) SACD and Hi-Res PCM both sound better than redbook CD's


2) There is some level of disagreement as to merits of hi-res PCM vs. DSD.
There's a missing link above.

There's so much misinformation out there regarding non-PCM formats, mostly just marketing pushing an agenda.

Bottom line is that with such doubt regarding high-res sounding any better than 44/16 (though listening tests), anything beyond that seems like another money making scheme, although Cherry DAC DAC, Cherry USB, and Cherry DPA all support up to 192/24.  "Why not" was started to show the justification for this.  However, there's no need to keep DSD/SACD in its original format since conversion to PCM doesn't take anything away that can be heard by human ears.

Here's the debate that's raging on (all 30 pages of it!):
https://www.audiosciencereview.com/forum/index.php?threads/high-resolution-audio-does-it-matter.11/

Freo-1

Re: The answer is.... WHY NOT ??
« Reply #13 on: 20 Apr 2019, 09:09 pm »
There's a missing link above.

There's so much misinformation out there regarding non-PCM formats, mostly just marketing pushing an agenda.

Bottom line is that with such doubt regarding high-res sounding any better than 44/16 (though listening tests), anything beyond that seems like another money making scheme, although Cherry DAC DAC, Cherry USB, and Cherry DPA all support up to 192/24.  "Why not" was started to show the justification for this.  However, there's no need to keep DSD/SACD in its original format since conversion to PCM doesn't take anything away that can be heard by human ears.

Here's the debate that's raging on (all 30 pages of it!):
https://www.audiosciencereview.com/forum/index.php?threads/high-resolution-audio-does-it-matter.11/



Good link.  Will take some time to read through it.


One item that jumped out at me is the example of Adel live BD sonics being OK, while the CD is pretty crappy.  I've noticed this with lots of CD vs. DVD, so I always look for a DVD version of artists and performances over CD, as the DVD/BD always seems to sound better than the CD. 


The whole PCM vs. DSD sonics is a debate that to me has no clear answer.  Even with my gear, I get two different sets of outcomes.  When using the McIntosh transport/DAC preamp setup, the DIN connection sounds better across the board, including CD's, over coaxial.  I have tried to play SACD/PCM from a Oppo 105 to the coaxial input of the Mac, and while it sounds good, it's not as good as the DSD direct from the Mac transport to the DAC.  Obviously, the DIN connection must do a better job of handling jitter over coaxial.


Having said all that, when feeding a Devialet 400 a SACD/PCM signal, the results are nothing short of astounding.  The Devialet is just at another level from a playback perspective over pretty much everything else I've come across.


Back to the original question:  To me, YES! Anything above 44KHz/16 bit is a step in the right direction.   

Freo-1

Re: The answer is.... WHY NOT ??
« Reply #14 on: 21 Apr 2019, 02:58 pm »

Tommy,


Thanks for posting your link.  Very informative, learned a lot.  Really enjoyed the talk given by Mark Waldrep.  I then went to Mark's website (Real HD-Audio), and started reading a lot of the blogs.  Turns out his thinking aligns pretty close with my experiences with audio.  I especially enjoyed his take on power cords and interconnect cables. 


I now better understand your issues with DSD.  Unfortunately, I still prefer to get most all of my media via optical storage, so CD, DVD, and SACD are the choices.  Since I now listen to classical more than the other genres, three are still a lot of SACD's of recent vintage for classical that are actual hi-res recordings.  Those are mostly what I get now, and the SACD does sound a bit better than the CD layer.  Depending on the playback hardware used, the SACD in native format sounds better played back in native SACD, the notable exception being the Devialet 400, where SACD/PCM is outstanding. 


What I really found to be informative was the subject of resolution.  His explanation of taking analog master sources and then claiming that the hi-res versions of the remastered audio was somehow better was quite enlightening.  Certainly gives one pause.   

AmpDesigner333

  • Industry Participant
  • Posts: 2973
  • Detailed AND Musical
    • Digital Amplifier Company
Re: The answer is.... WHY NOT ??
« Reply #15 on: 23 Apr 2019, 04:47 pm »
Tommy,


Thanks for posting your link.  Very informative, learned a lot.  Really enjoyed the talk given by Mark Waldrep.  I then went to Mark's website (Real HD-Audio), and started reading a lot of the blogs.  Turns out his thinking aligns pretty close with my experiences with audio.  I especially enjoyed his take on power cords and interconnect cables. 


I now better understand your issues with DSD.  Unfortunately, I still prefer to get most all of my media via optical storage, so CD, DVD, and SACD are the choices.  Since I now listen to classical more than the other genres, three are still a lot of SACD's of recent vintage for classical that are actual hi-res recordings.  Those are mostly what I get now, and the SACD does sound a bit better than the CD layer.  Depending on the playback hardware used, the SACD in native format sounds better played back in native SACD, the notable exception being the Devialet 400, where SACD/PCM is outstanding. 


What I really found to be informative was the subject of resolution.  His explanation of taking analog master sources and then claiming that the hi-res versions of the remastered audio was somehow better was quite enlightening.  Certainly gives one pause.
I agree about studio masters!  How can any recording (analog or digital) be better than the original?  It can't, so the only real benefit must be due to wear or demagnetization.

A related issue is finding out what process was used to make the master.  It takes research, and there's no guarantee you'll even get to the information.  This can be track-by-track, album-by-album, or format-by-format.

The dynamic range database does a good job of telling you which recordings are most dynamic:
http://dr.loudness-war.info/

It's mentioned here, in this example of an LP sounding better than the CD:
https://www.audiocircle.com/index.php?topic=103744.0

-Tommy O

Elizabeth

  • Full Member
  • Posts: 2736
  • So Long, and Thanks for All the Fish
Re: The answer is.... WHY NOT ??
« Reply #16 on: 23 Apr 2019, 09:11 pm »
IMHO, I think that the CD standard should have been 24 bit/48KHz.  That would have been sufficient for most applications.  Not sure how accurate this is, but the story going around at the time of CD release was that the 16 bit/44 KHz limit was the standard of getting Beethoven's 9th on a single CD.  The story goes that the initial spec was to be 48 KHz, but it exceeded the 660 MB limit of the CD.
At the time the standards were being developed. 16/44 was a stretch The developers were reaching for the future.... The early devices were 14 bits because no one could make a 16 bit device yet. The MYTH about 24 bits is total fantasy!... calling it a 'story' is right on. Plus it has to have been made up in the past ten years. NEVER heard that fluff before LOL.
As for The time limit length, yes THAT IS TRUE it was to fit Beethovens 9th on one CD.

OzarkTom

Re: The answer is.... WHY NOT ??
« Reply #17 on: 24 Apr 2019, 12:32 am »
The Bluesound Node 2i I recently bought sounds better than the Sony Hap-Z1es I owned. The Sony played DSD, the Bluesound does not. Does this tell anything?

OzarkTom

Re: The answer is.... WHY NOT ??
« Reply #18 on: 24 Apr 2019, 12:38 am »

It's mentioned here, in this example of an LP sounding better than the CD:
https://www.audiocircle.com/index.php?topic=103744.0

-Tommy O
 

I had a record and CD of the same recording one time. The first song, the record sounded much better. The second song, the CD sounded better. Just engineering?


Freo-1

Re: The answer is.... WHY NOT ??
« Reply #19 on: 24 Apr 2019, 02:35 am »
The Bluesound Node 2i I recently bought sounds better than the Sony Hap-Z1es I owned. The Sony played DSD, the Bluesound does not. Does this tell anything?



It does not necessarily point to DSD.  More likely points to the analog output deltas between the two units.