Mark Waldrep/Sound Liaison/DSD vs PCM part 2,(Mark answers computer audiophile.)

[PAP]

Mark Waldrep/Sound Liaison/DSD vs PCM part 2

 

A few quotes from part one;

From the recordings that I’ve heard from Sound Liaison, they’re doing excellent work. I can only guess that the “buzz” around DSD enticed them into that market with their PCM conversions.I don’t mean to single out Sound Liaison with regards to the pricing of DSD conversions of their high-resolution 96/24 PCM masters. Their recordings are actually spectacular (maybe because they make them at 96 kHz/24-bitPCM?)

I wrote back to Frans and thanked him for his response and I invited him to make his wonderful tracks available through iTrax.com. Mark Waldrep

I've spent a lot of time with Mark Waldrep, and indeed he is vehemently against DSD (In fact, you don't want to be in the same room with Waldrep and Cookie Marenco of Blue Coast Records. It's like a matter/antimatter reaction!)

What I don't hear is any real sonic advantage to 24/192 over 24/96 gmgraves2

DSD is a superior representation of an inherently analog signal than LPCM is. LPCM was developed pretty much as a space saving technique,Paul

Would you rather have the conversion to DSDdone by a chip in your DAC, or with care by Sound Liaison with all the equipment available to them? (Sound Liaison fully discloses the DSD recordings are conversions from the original PCM versions, so no one buys without knowing that.) Jud

Except for the most recent Sound Liaison recording - Improptu - which was recorded in both PCM and DSD. On that DSD download, there is no PCM to DSD conversion. IMPROMPTU Bmoura

 

And if he does not like DSD, then be happy that the PCM recordings from Sound Liaison are relatively cheap. 11$ / 10€ for a sampler with that kind of SQ is very cheap and Mark himself calls the Sound Liaison recordings spectacular, so ???PAP

Blue Coast Records records 99% of it's projects on DSD to the Sonoma a.nd 1% to analog tape. Editing in the Sonoma does not go to a PCM stage the way we use it. We don't mix in the Sonoma. We mix out through an analog console to add effects and levels changes as we hear it. We mix back to DSD128. Cookie

 

why is it OK to record multitrack to tape and then transfer to DSD but

not OK to record multitrack to PCM and then transfer DSD? oso

Perhaps only direct-to-disc vinyl or direct-to-reel recording will provide this purity some are looking for

 

Mark's answer;

530 Comments!

What's going on over at Computer Audiophile? I know many of you are active readers of my site and Chris' CA site. I try to stop by there every so often and see what's brewing. In general, I've found the heat a little high for my sensibilities…and some of the "facts" expressed are more personal opinion than bona fide information. But what could possible prompt the CA community to post 530 comments on my assertion that 96 kHz/24-bit PCM surpasses DSD 64? A sophomore member of CA titled his thread "Mark Waldrep is claiming that PCM 24/96 is superior to DSD" and then went on to copy my entire post from April 11, 2105...the one that took issue with Sound Liaison for upcharging its customers for DSD version of their source 96/24 PCM recordings.

 

He asks a simple question, "Mark Waldrep is claiming that PCM 24/96 is THE format and that DSD should be cheaper than PCM. Comments? Is he right or just promoting his Itraxx pcm recordings?"

 

The simple answer is yes, if you consider the technical specifications between the two formats (never mind that you can't actually produce most recordings in DSD because of the lack of tools). PCM recordings made at 96 kHz/24-bits have greater frequency response and dynamic range than DSD 64 tracks. The noise shaping required to increase the dynamic range beyond 6 dB (the result of 1-bit encoding) in the "audio band" is deposited just above the top of our hearing range. Look at the spectra of a typical DSD 64 track below:

Figure 1 - A spectrograph of an "audiophile" DSD 64 recording. [Click to enlarge]

 

The purple "haze" in the high frequency area of the left hand spectral displays are the ultrasonic noise...put there by the noise shifting. It's supposed to be out of the "audio band" but there is no "audio band" for your playback equipment. Imagine how your electronics and tweeters feel when this type of recording comes along...all analog sounding and warm. You bet. The graph on the right is a big side letter "V". At a point just above 23 kHz, the line steadily rises almost back to where it started. Welcome to DSD 64. Any resemblance to high-end audio is purely coincidental.

 

What about dynamic range? DSD can actually eclipse a CD in dynamics AFTER the noise is plowed out of the way. But it doesn't come close to the theoretical 144 dB that 24-bit PCM can achieve. Those specs are great for recording engineers like me but the fidelity of your system will never benefit from 24-bit DACs or fidelity. There simply are no recordings being released that have that much dynamics and your room couldn't handle it if they were.

 

The answer to the question posed back in April at CA is quite straightforward. How is it possible for Chris' reader to spin out 530 comments arguing about a true statement? I haven't read all of the comments. The ones that I did read focus on the personal preferences of readers for the "warmth and analog sound" erroneously attributed to the DSD format (remember the study that determined that no one could tell the difference between mics split to a PCM rig AND a DSD rig).

 

[ted_b]

I, personally, don't understand the need for setting up a war between factions. To me its like arguing gold vs silver. Who cares; they are both wayyy better than tin or copper. And seldom if ever, is an audio lover faced with a decision as to which one to buy...if the recording is DSD, buy the DSD. If the recording is PCM, buy the PCM. If the original native recorded sample rate is not available, then buy the closest you can (assuming your DAC is somewhat agnostic) What is the reason that we need to establish which is better?? No one in their right mind would debate that converting one to the other is better than the native original, dac sweetspot notwithstanding (except in the rare occurrences where PCM DSP is needed to fix bigger room issues than the lossy tradeoff...and these issues are going away with things like DSD and PCM convolving in HQPlayer, for example)..

 

I think the debate and the cold words that these things end up becoming...is bad for the adoption of higher resolution recording. Buy the closest to the master tape, and then enjoy the wonderful sounds.

[Paul R]

Now isn't that interesting?

 

But what Mr. Waldrep claims in his response, that 96/24K is technically superior to DSD64, is debatable.

 

But substitute DS128 for DSD64 and the comparison is absolutely not true. You will note that the original CA topic was not specifying DSD64. That does change things, slightly. I would suggest it takes PCM at 176.4/24 to equal DSD64.

[Jud]

I, personally, don't understand the need for setting up a war between factions. To me its like arguing gold vs silver.

 

Very much agreed.

 

No one in their right mind would debate that converting one to the other is better than the native original, dac sweetspot notwithstanding

 

My own thought is that the "sweet spot" is going to come into play a high percentage of the time (to put it another way, is your DAC sigma-delta or R2R?), unless you intend to do the conversion with software, in which case the DAC's "sweet spot" will have a great deal to do with what format you convert to.

 

enjoy the wonderful sounds.

 

Again, very much agreed.

[ted_b]

Jud, thanks. Yes, dac sweetspot is going to be a major factor, as is availability of the recording..unless of course this debate is supposed to be among those who record (in which case, let Mark, Barry, Jared, Tom and the other few folks here debate it, and wake me up when it is closed). DSD is not an acceptable recording format for a majority of popular labels, mainly due to the simple fact that it cannot be easily edited and manipulated. But for us audio lovers, that issue is irrelevant except for the paucity of native DSD pop/rock recordings. But for archiving/transferring analog tapes, it seems a lot of traction is in DSD128 and DSD256, regardless of the fact that more of the buying market owns PCM playback gear than DSD playback gear.

 

When Paul R says "equal" I will debate that comment, though. Nothing between PCM and DSD is equal IMHO. Equivalent maybe..in terms of sound quality, but to equate a format that uses samples to a format that uses a modulated stream...is, well, illogical. And in terms of sound quality, it is SOOOO dependent on the rest of the system that this debate is chock full of unmovable variables....i.e not worth debating. In tests that have already occurred (documented on the web), many find 24/192 to sound quite similar to DSD64...but then what can be said for the pool of listeners, or even for each format's higher variant (24/352.8k, for example, or DSD128 or DSD256). It's all so subjective at this sound quality level that it is almost absurd to subject it to hundreds of pages of debate. In my own listening, Norah Jones's DSD box set sounds better than her 24/192 PCM one in several DACs I've heard...and her 24/192 PCM one sounds better in a number of DACs I've heard. And these are different masterings!! Good luck with that. Same goes for those few simultaneous comparo recordings (Massimo Gon's Lizst recording, for example).

 

The pros tend to think that 24/352.8k is required to edit DSD64, so that may tell someone something, too (unless, of course, 24/352.8k was simply picked cuz they could..and they wanted to create a term called DXD for bragging rights purposes...who knows).

[jabbr]

I, personally, don't understand the need for setting up a war between factions. To me its like arguing gold vs silver. Who cares; they are both wayyy better than tin or copper. And seldom if ever, is an audio lover faced with a decision as to which one to buy...if the recording is DSD, buy the DSD. If the recording is PCM, buy the PCM. If the original native recorded sample rate is not available, then buy the closest you can (assuming your DAC is somewhat agnostic) What is the reason that we need to establish which is better?? No one in their right mind would debate that converting one to the other is better than the native original, dac sweetspot notwithstanding (except in the rare occurrences where PCM DSP is needed to fix bigger room issues than the lossy tradeoff...and these issues are going away with things like DSD and PCM convolving in HQPlayer, for example)..

 

I think the debate and the cold words that these things end up becoming...is bad for the adoption of higher resolution recording. Buy the closest to the master tape, and then enjoy the wonderful sounds.

 

+1

 

However the music starts out whether PCM or DSD is less important that proper conversion to whatever is best for the DAC.

 

All else equal I prefer the native resolution of the master -- that is to say if someone or something (eg a chip) is going to do a format conversion I would prefer to do this myself ie using the software I choose.

[loop7]

Just imagine if there was a Michael Fremer type personality preaching either DSD or PCM!

 

We should be focusing on DACs.

[tranz]

Jud, thanks. Yes, dac sweetspot is going to be a major factor, as is availability of the recording..unless of course this debate is supposed to be among those who record (in which case, let Mark, Barry, Jared, Tom and the other few folks here debate it, and wake me up when it is closed). DSD is not an acceptable recording format for a majority of popular labels, mainly due to the simple fact that it cannot be easily edited and manipulated. But for us audio lovers, that issue is irrelevant except for the paucity of native DSD pop/rock recordings. But for archiving/transferring analog tapes, it seems a lot of traction is in DSD128 and DSD256, regardless of the fact that more of the buying market owns PCM playback gear than DSD playback gear.

 

When Paul R says "equal" I will debate that comment, though. Nothing between PCM and DSD is equal IMHO. Equivalent maybe..in terms of sound quality, but to equate a format that uses samples to a format that uses a modulated stream...is, well, illogical. And in terms of sound quality, it is SOOOO dependent on the rest of the system that this debate is chock full of unmovable variables....i.e not worth debating. In tests that have already occurred (documented on the web), many find 24/192 to sound quite similar to DSD64...but then what can be said for the pool of listeners, or even for each format's higher variant (24/352.8k, for example, or DSD128 or DSD256). It's all so subjective at this sound quality level that it is almost absurd to subject it to hundreds of pages of debate. In my own listening, Norah Jones's DSD box set sounds better than her 24/192 PCM one in several DACs I've heard...and her 24/192 PCM one sounds better in a number of DACs I've heard. And these are different masterings!! Good luck with that. Same goes for those few simultaneous comparo recordings (Massimo Gon's Lizst recording, for example).

 

The pros tend to think that 24/352.8k is required to edit DSD64, so that may tell someone something, too (unless, of course, 24/352.8k was simply picked cuz they could..and they wanted to create a term called DXD for bragging rights purposes...who knows).

 

Nice summary. Thanks.

[Jud]

Just imagine if there was a Michael Fremer type personality preaching either DSD or PCM!

 

We should be focusing on DACs.

 

Do an advanced forum search for "Charles Hansen" and "DSD."

[Paul R]

When Paul R says "equal" I will debate that comment, though. Nothing between PCM and DSD is equal IMHO. Equivalent maybe..in terms of sound quality, but to equate a format that uses samples to a format that uses a modulated stream...is, well, illogical. And in terms of sound quality, it is SOOOO dependent on the rest of the system that this debate is chock full of unmovable variables....i.e not worth debating.

 

I will concede that point to you with much grace, though it is actually a minute concession. I happen to agree with you.

 

The wording was me trying to wrap my head around those folks for home LPCM sounds better than DSD and not be insulting to anyone. Especially those I quite disagree with.

 

-Paul

[Allan F]

Do an advanced forum search for "Charles Hansen" and "DSD."

 

You can add "Gilad Tiefenbrun" (of Linn), "Berkeley Audio Design" and "DSD" to that list.

[Fitzcaraldo215]

You can add "Gilad Tiefenbrun" (of Linn), "Berkeley Audio Design" and "DSD" to that list.

 

Yes, but "DSD" is, I believe the famous, or infamous, Teresa, who had maintained her own blog and also appeared in Positive Feedback Online. She argued incessantly in many forums - she was expelled from a number of those - with anyone who disagreed with her about the slam dunk supremacy of the format.

 

PFO now features ted_b, which is a considerable improvement over that strident harpy, who was nothing but a lightweight, but also a tireless zealot, true believer of the worst possible kind. Checking out her postings is really not very worthwhile, believe me. She is simply not rational, technically knowledgable or even musically sophisticated. I have dealt with her numerous times directly in other forums, regrettably, and followed her exploits more than I cared to. I think she may have done the SACD/DSD format more harm than good.

 

And, though we know ted_b personally favors DSD recordings, his responses here are eminently reasonable and inclusive. He genuinely aims to help his fellow audiophiles rather than shriek at them. Were it not for him, I would not have the large library of SACDs that I do in my computer. But, I say this as a firm believer in converted hi rez PCM playback via Room EQ (Dirac) myself.

[YashN]

DSD is not an acceptable recording format for a majority of popular labels, mainly due to the simple fact that it cannot be easily edited and manipulated.

 

One could do as 2L.no and record in DXD rate, do the edits and processing as usual if possible (I can't check in my own setup so someone can tell me if it works in major DAWs and easily available capture cards), and only at the final stage convert to DSD.

 

DSD is very close to how analogue works, so to me the Blue Coast Records way makes sense: part of the production chain is in analogue.

 

Another way is to also use PCM, even lower than DXD, but use a specific setup with specific plugins which do not sound digital at all when used properly, and then either provide the PCM or convert to DSD, provided of course the provenance is mentioned.

[YashN]

But what Mr. Waldrep claims in his response, that 96/24K is technically superior to DSD64

 

The narrowing down of the comparison to DSD64 only is, to me, yet another example of his rather obvious bias towards the format he sells, i.e. 96/24.

[ted_b]

One could do as 2L.no and record in DXD rate, do the edits and processing as usual if possible (I can't check in my own setup so someone can tell me if it works in major DAWs and easily available capture cards), and only at the final stage convert to DSD.

 

DSD is very close to how analogue works, so to me the Blue Coast Records way makes sense: part of the production chain is in analogue.

 

Another way is to also use PCM, even lower than DXD, but use a specific setup with specific plugins which do not sound digital at all when used properly, and then either provide the PCM or convert to DSD, provided of course the provenance is mentioned.

 

Although I am biased, my favorite way to record in DSD is what Jared does. He mixes in the box, in analog, before the ADC, then sends the mix directly to DSD. No editing, period.

 

BTW, I'd MUCH rather have Morten's (2L) 24/352.8k native recordings (aka DXD) than have them converted to DSD. Why do it?

[YashN]

Although I am biased, my favorite way to record in DSD is what Jared does. He mixes in the box, in analog, before the ADC, then sends the mix directly to DSD. No editing, period.

 

BTW, I'd MUCH rather have Morten's (2L) 24/352.8k native recordings (aka DXD) than have them converted to DSD. Why do it?

 

Lots of different ways. I'd probably prefer not editing at all if possible as well for DSD, but I mostly do electronic music myself.

 

As for 2L, ask Morten himself: probably he knows some of his audience may have a focus on DSD and not DXD. Some people do not have access to real-time or offline converter software to do that or have less-capable software.

 

Getting both from the label is a convenience, not the opposite.

 

Forgot to add: it's also dependent on your DAC's sweet-spot.

 

This said, I'll probably need to sit down and compare the two formats again someday: so much has changed in my setup (yet again...)

[YashN]

I think some people can chain several DSD recorders and thus get a true multi-track system like that. Couple with processing in analgoue, that can be a great way to proceed.

[tranz]

Although I am biased, my favorite way to record in DSD is what Jared does. He mixes in the box, in analog, before the ADC, then sends the mix directly to DSD. No editing, period.

 

BTW, I'd MUCH rather have Morten's (2L) 24/352.8k native recordings (aka DXD) than have them converted to DSD. Why do it?

 

Hi Ted,

Did you mean out of the box, i.e. using outboard racks of gear and an analog mixing panel like Cookie does?

[elcorso]

----n my own listening, Norah Jones's DSD box set sounds better than her 24/192 PCM one in several DACs I've heard...and her 24/192 PCM one sounds better in a number of DACs I've heard. And these are different masterings!! Good luck with that. Same goes for those few simultaneous comparo recordings (Massimo Gon's Lizst recording, for example).

------

 

Hi Ted,

 

The same in my case, but, DSD is DSD

 

BTW, any chance of having available Massimo Gon's Liszt recording in NativeDSD download site?

 

I'm behind this recording since a long time and nobody sells the DSD format or the physical SACD!

 

Thanks,

 

Roch

[Hiro]

So Mark Weldrep claims that "96 kHz/24-bit PCM surpasses DSD 64". The question is in what? rounding, ringing, and aliasing?

 

We really should see the wide-band graphs into far ultrasonics that show how the out-of-band PCM noise (better known as aliasing) looks like. You won't find those in PCM marketing materials, but there have been plenty of those posted on Computer Audiophile forum, so they are not hard to find.

[Hiro]

Mark Waldrep/Sound Liaison/DSD vs PCM part 2

 

But it doesn't come close to the theoretical 144 dB that 24-bit PCM can achieve.

 

Oh boy, he's comparing "theoretical" values (no PCM ADC ever achieved) to real life measurements...

 

Though, oddly enough, he later adds that "There simply are no recordings being released that have that much dynamics."

 

Which leads one to wonder why he would trot out this non-argument in the first place...

[oso]

why is it OK to record multitrack to tape and then transfer to DSD but

not OK to record multitrack to PCM and then transfer DSD? oso

 

If somebody would answer this question I'd be much obliged.

[bdiament]

The narrowing down of the comparison to DSD64 only is, to me, yet another example of his rather obvious bias towards the format he sells, i.e. 96/24.

 

Hi YashN,

 

In all fairness, one might consider the fact that it isn't a bias to a format he sells, but simply a matter of him selling a format he likes better. Folks tend to be a little jaded on Internet audio fora. (No! Really?!? ;-})

 

I'm pretty confident that if Mr. Waldrep preferred DSD, that is the format in which he would be recording and that is the format he would be selling. A lot of us make records because we love making records. A the small, independents tend to be labors of love. (Anyone think Mr. Waldrep is getting wealthy doing this?) As such, the proprietors are doing things the way they *want* to do them.

 

Of course, I'm speaking for myself but I have a strong feeling all this applies to Mr. Waldrep and others doing similar things in record-making. So, "obvious bias"? Absolutely. But not at all because he sells it. He sells it because that is how he feels it will be best heard.

 

I know a lot of folks who love the format. Like Mr. Waldrep, I'm not one of them. This is a sonic motivation, not a commercial one. (Soundkeeper, my label, already offers different formats in addition to my favored 24/192 .aif. It does *not* offer flac files as downloads, which would certainly add to the bottom line. Nothing is on iTunes either.) Sometimes folks are in this crazy pursuit because they love music and sound and that is all the motivation they require.

 

Best regards,

Barry

Soundkeeper Recordings

http://www.soundkeeperrecordings.wordpress.com

Barry Diament Audio

[Lighthouse]

For Dr. Waldrep (yes, he has a doctorate degree on music production. We can at least respect his academic knowledge), posting these blog posts certainly do not favor his business since it will anger many DSD advocates who also buy PCM... As Barry says, Dr. Waldrep has no financial gains pushing PCM.

 

I believe he made several comments regarding 24/96 vs 24/192 (basically, he admits his studio equipment is not up for real 24/192, as well as pointing out humans really can't hear diff between 96 and 192.)

 

That said, with almost majority of DSD music available are only DSD64, his opinion is still valid. I must remind you guys that DSD128 and higher formats are truly niche of niche, and has almost nil on larger population. As a owner of a record label and studio, Dr. Waldrep's view is perfectly fine.

 

Personally, until DSD gets actually a decent codec tailored to PC (like FLAC for PCM format,) I don't think I will go deeper into DSD despite I do own some SACDs.

[jabbr]

For Dr. Waldrep (yes, he has a doctorate degree on music production. We can at least respect his academic knowledge), posting these blog posts certainly do not favor his business since it will anger many DSD advocates who also buy PCM... As Barry says, Dr. Waldrep has no financial gains pushing PCM.

 

I believe he made several comments regarding 24/96 vs 24/192 (basically, he admits his studio equipment is not up for real 24/192, as well as pointing out humans really can't hear diff between 96 and 192.)

 

That said, with almost majority of DSD music available are only DSD64, his opinion is still valid. I must remind you guys that DSD128 and higher formats are truly niche of niche, and has almost nil on larger population. As a owner of a record label and studio, Dr. Waldrep's view is perfectly fine.

 

Personally, until DSD gets actually a decent codec tailored to PC (like FLAC for PCM format,) I don't think I will go deeper into DSD despite I do own some SACDs.

 

I'm recording format agnostic ... If was able to choose the recording format then I might develop an opinion but my order of preference is:

 

1,2,3: great music

4: recording & mastering technique

5: high resolution

 

From my own experience one can convert between PCM & DSD and that if this is done in a high enough resolution domain, then artifacts related to conversion will be minimized.