DSD comparison to PCM.

[c1ferrari]

Thanks for the post, Mani.

The converter loopback merits attention.
Rohde & Schwarz. Nice. Would you know which model and options produced this data?

Thanks

Here are the ADC modules:

View attachment 8233

They're about 1.5 times the size of the DAC modules. The PM2 designers claim they are true multi-bit ADCs.

This is what Michael Ritter said in an interview for Mix:

“The actual A/D converter in the Model One runs at 24 bits and 176.4 kHz currently; the Model Two will also convert at 192 kHz. We improve the linearity of our conversion with a high-amplitude broadband dither signal that we mix in with the program in the analog domain. The dither appears to be random, but the system knows at any given instant precisely what the amplitude of that dither signal is. And because we use our own custom, discrete, full-ladder converter with excellent amplitude and phase accuracy, we are able to apply an 'anti-dither' signal, exactly out-of-phase and matched in time, in the digital domain after conversion. That nulls the dither noise out of the signal… If it's going to be a 176.4 or 192kHz DVD-Audio release, then we will not decimate that signal; we use a proprietary filter [non-oversampled] optimized to that sample rate."

This is what Keith O Johnson said in reply to a question I asked him:

“Up conversion is not used at 176.4 kHz and 192 kHz sample rates. Instead, a passive analog filter with group delay correction removes alias products so that all 24 bits of information are captured from each sample by the ADC process. Conversion is made with ladder-pipeline architectures and analog added - DSP subtracted dithering, a first of its kind that is still used for instrumentation. Unlike sigma-delta methods, jitter sensitivity is low and accuracy does not degrade at high program levels.”

In any event, here are some test results of my particular PM2 unit (courtesy of Dave Peck):

ADC Multisine:
View attachment 8234

ADC THD:
View attachment 8235

DAC THD:
View attachment 8237

ADC/DAC Loopback:
View attachment 8236

With all this, can we be certain that the PM2 is indeed a true multi-bit ADC capable of of >16bit resolution? Interested in your thoughts...

Mani.

 

[manisandher]

Oh sure. My Korg uses a PCM4202 which is a single bit ADC, and my Tascam a PCM4222 which is multi-bit ADC. But of course, both use sigma-delta modulation. So all that would need to be updated on Andreas' diagram would be the '2.8MHz/1bit' to the appropriate number of bits. No?

But my real point was that just from looking at the diagram it's clear that chains two and three are going to do less conversions, and hence have the potential to sound better, than chain one. And this has been exactly my experience. However chain three sounds substantially better to me than chain two. I'm happy to accept though that a Grimm ADC and a Meitner/EMM DAC might match or even exceed my PM2 and NOS1. But as it is, the latter combo, which I already possess, sounds close enough to the original analog for me to be perfectly happy.

Mani.

 

[manisandher]

Would you know which model and options produced this data?

Hi Sam, unfortunately no. They were sent to me by Dave Peck who used to work for Pacific Microsonics and then Euphonix. They were measurements that he had taken of my particular machine back in 2003 when he had it (presumably for servicing). Of course, by today's standards 17-18 bit resolution sounds a bit lame. But it's done without any noise-shaping and I think this is one of the keys to its great sound.

Mani.

 

[opus111]

Oh sure. My Korg uses a PCM4202 which is a single bit ADC, and my Tascam a PCM4222 which is multi-bit ADC. But of course, both use sigma-delta modulation. So all that would need to be updated on Andreas' diagram would be the '2.8MHz/1bit' to the appropriate number of bits. No?

Correct, but then its no longer an argument for DSD, just an argument for noise-shaped PCM. And PCM isn't supposed to sound as good as DSD. But at least noise-shaped PCM can be edited, EQ'd and processed in various ways.

But my real point was that just from looking at the diagram it's clear that chains two and three are going to do less conversions, and hence have the potential to sound better, than chain one. And this has been exactly my experience.

Yes I concur - the fewer steps the better. So then it becomes the difference between PCM with a multibit converter, and noise-shaped PCM with a sigma-delta converter. Nowhere does DSD feature. Not quite the result that the DSD promoters want is it?

However chain three sounds substantially better to me than chain two. I'm happy to accept though that a Grimm ADC and a Meitner/EMM DAC might match or even exceed my PM2 and NOS1. But as it is, the latter combo, which I already possess, sounds close enough to the original analog for me to be perfectly happy.

Precisely what the theory would predict. I doubt that a Grimm ADC would leave the dynamics intact as noise modulation is inherent in its single-bit design. But then the NOS1 isn't state of the art for multibit, based as it is on R2R DACs.

 

[manisandher]

I have read great things about Phasure. Congrat's on your choice of SOTA digital.

Yes, I really do think it is SOTA. OK so Bruce didn't like it. And that's fine. He probably used one of the hi-rez files that he prepared for HDTracks, ripped off an SACD which itself was originally recorded in lo-rez PCM. No wonder the NOS1 sounded crap in his hands ;-)

Mani.

 

[manisandher]

But then the NOS1 isn't state of the art for multibit, based as it is on R2R DACs.

Opus, didn't BB do something to reduce glitching in the R2R PCM1704? In any event, the multi-bit choice then comes down to 16 bit with something like a TDA1541A or 24 bits with something like the PCM1704. You pays your money and takes your choice I suppose.

Mani.

 

[opus111]

Yes - good point, I was even now just thinking over that my 'SOTA for multibit' claim needed nuancing to 'SOTA for RBCD'.

BB did come up with their 'CoLinear DAC' architecture yes, which I've commented on elsewhere - it doesn't reduce the glitch just moves it to different output codes (two glitches at the -6dB points, rather than one slightly bigger glitch at zero). This was introduced with the PCM63 which I consider to be the best of BB's offerings based on technical specs. Even that part is bettered by the D20400 on paper. If you really want >16bits and also if you want to run faster than a TDA1541A will go (tops out at 8X OS) then you've got precious little choice. Which is why I intend to develop something for multibit lovers which goes beyond 16bits in future and which hopefully avoids the R2R glitchiness.

 

[manisandher]

Worth noting that the 1st example in the diagram is focused on DSD going to PCM and seems to miss out the ones that do not go to 1-bit recording (sacd/dsd); the studios that are always 24bit/192khz or even 24bit/96khz and possibly 176.4khz.

Orb, when you say "... the studios that are always 24bit/192khz or even 24bit/96khz and possibly 176.4khz", these studios will invariably be using a 5-6-bit sigma-delta ADC. So yes, not 1-bit, but certainly not a 'pure' multi-bit ADC.

IIRC, Linn used to own a Pacific Microsonics Model One - that's how they managed to release their HDCD-encoded CDs. If they still own it, they could indeed record in 'pure' multi-bit PCM.

As I mentioned earlier, my Tascam uses a PCM4222 ADC. This is a 6-bit sigma-delta ADC (apparently one of the better ones), and yet is easily beaten by the PM2 in terms of sound... IMO.

Mani.

 

[manisandher]

Which is why I intend to develop something for multibit lovers which goes beyond 16bits in future and which hopefully avoids the R2R glitchiness.

I very much look forward to that. Will this be something you'll be offering as a commercial product or strictly for DIY? I hope the former but suspect the latter.

Mani.

 

[opus111]

None of my DIY efforts are 'strictly for DIY' - I'm interested in manufacturers taking up the designs and running their products based on them, royalty and license free. They'd only need to pay me a consulting fee if they wanted me to customize them in some way for them. I have no intention of going into mass production myself as I prefer the fun of design and prototype build and test to the apparent chore of organizing production.

 

[Bruce B]

OK so Bruce didn't like it. And that's fine. He probably used one of the hi-rez files that he prepared for HDTracks, ripped off an SACD which itself was originally recorded in lo-rez PCM. No wonder the NOS1 sounded crap in his hands ;-)
Mani.

You're absolutely correct.... I'm busted. The only DSD files I have are from the crap SACD's that were upsampled...

 

[Ki Choi]

You're absolutely correct.... I'm busted. The only DSD files I have are from the crap SACD's that were upsampled...

Wow! I knew it... Matter of fact (now I come to think of it), it was suspicious to see, and I shouldn't have been so gullible to believe the huge wood crate in your studio with the Wilson Audio shipping label was full of Dave Wilson's one-of-kind original recording master tapes.

I took a photo of the wood crate with my phone but can't locate it for the inquiring minds, too bad.

But now I know and should have known that the crates was indeed full of the crap SACDs...

 

[jkeny]

Did some listening tests with 2 friends yesterday. Compared Meitner MA-1 to some S-D & NOS DACs. We played mostly 16/44 WAV or FLAC files from disk via USB or via CD player using SPDIF. They didn't know anything about the Meitner other than it's price so expectations were high. The Meitner internally converts all inputs to DSD AFAIK but it was considered bottom of the pile really - kinda lifeless compared to the other DACs. Didn't get a chance to listen to native DSD yet - hope to do so in the next couple of days.

 

[mep]

I guess I shouldn't have been surprised at how fractured the digital religion is, but I was. We have those that believe that high-sample rate MP3 is all you need. We have those that believe that RBCD is all you need. We have those that believe that hi-rez PCM is all you need. And then we have the real truth, and that is DSD. So along with your favorite digital files, now comes the arguing over the *best* type of DAC. I find it ironic that some people who exclusively listen to digital and pooh-pooh all things analog just for the inferior measurements to the ever more perfect digital measurements that they prefer NOS DACs which are the worst performing DACs technically of all the DACs in the digital wonderland. I sincerely believe that NOS DACs are the tube equivalent of SET amps. They both measure terribly in comparison to their brethren, but yet they are loved for their sound by those who love their sound.

 

[manisandher]

You're absolutely correct.... I'm busted. The only DSD files I have are from the crap SACD's that were upsampled...

When did I say you don't have DSD files derived directly from master tapes? I'm sure you do. But by your own admission, you also have a 1000 DSD album files ripped from crap SACDs:

From RMAF Seminar 2012

Chad Kassem (Analogue Productions): To me the mastering is way more important than DSD or 24/96 or 192 or 88 or all these different numbers that everybody throws around.

Audience member to Bruce: Do you have a sense [of the origin of the recording]?

Bruce: David has the same problem as Chad does. He’s working with the labels and they only give him a CD or a SACD to work with.

Chad: I don’t have that problem. I get the master analog tape. If they don’t give it to me then I guess there’s no problem.

Audience laugh and applaud.

Bruce: Well I guess you should take over David’s position then. So anyway when we first started out, we started out they’ve given him the SACDs, we were ripping the SACDs and we probably ripped a thousand or so until we found out that these weren’t true DSD files and things like that and we were caught with our hands tied and having to go back and check the last year’s work we’d done. And another thing with this, people started looking with Audacity and things like that. You can look at these spectral analyzers and things like that but you have to at least know how to read them because any good mastering engineer can fool anybody up here. I can take a redbook CD file and make you believe it’s a DSD file. So listen with your ears and not your eyes.

A couple of my own thoughts here:
- You were caught with your hands tied by the very people you criticize for using Audacity. If it were't for these people you'd probably still be passing off crap SACD rips as genuine DSD recordings.
- I've got some of the so called 'hi-res' files that you ripped from crap SACDs for HDTracks. You tell us to listen with our ears and not our eyes. When you were doing the ripping, how the hell could you not tell from listening with your ears that these were not true DSD recordings on the SACDs??? It took you a whole year to notice? In fact it was the 'Audacity Cowboys', as you refer to them, who first identified these downloads as being no better than RBCD.

Mani.

 

[Bruce B]

A couple of my own thoughts here:
- You were caught with your hands tied by the very people you criticize for using Audacity. If it were't for these people you'd probably still be passing off crap SACD rips as genuine DSD recordings.
- I've got some of the so called 'hi-res' files that you ripped from crap SACDs for HDTracks. You tell us to listen with our ears and not our eyes. When you were doing the ripping, how the hell could you not tell from listening with your ears that these were not true DSD recordings on the SACDs??? It took you a whole year to notice? In fact it was the 'Audacity Cowboys', as you refer to them, who first identified these downloads as being no better than RBCD.

Mani.

You must have me confused with HDtracks. I am a 3rd party vendor. If you have a beef with HDtracks, stand in line. I don't make decisions at HDtracks though I got the blame for all of this because I was not told or paid to check the files. I've spent the better part of the past few years responding to accusations and provenence questions by customers such as you but have taken time out of my busy schedule to answer these questions. I want an informed public just as much as you but I get my hands slapped by HDtracks saying "I'm not their public relations spokesperson".

You can see on SA-CD.net and other reviews that the same upsampled SACD's were hearlded for their musical fidelity. Just as soon as they were found to be upsampled (BIS and others), then everyone started back tracking saying they sounded like SH*T... that's listening with your eyes.

David got the complaints that these were upsampled and then paid me to go back and start checking the past years work. I don't listen to these SACD's. I make sure the rip was good and then cut/render the files. I run a mastering studio and this was extra busy work that I took on to better the audiophile community. I didn't have the time or manpower to start checking the past 2yr worth of work. I had to put off good mastering jobs to do this. Before you start calling me out for these upsampled files, look in your own backyard. I did not coin the phrase "Audacity Cowboy" BTW.....

 

[dallasjustice]

Keith Johnson lectured a few years ago at RMAF. He covers alot of material, only 30% of which I can comprehend. If you start the video I linked to at 30 minute mark he starts in about jitter and how sdm dacs and ladder dacs are affected by jitter differently. He also shows some nuclear scope shots of jitter.

 

[MarinJim]

I guess I shouldn't have been surprised at how fractured the digital religion is, but I was. We have those that believe that high-sample rate MP3 is all you need. We have those that believe that RBCD is all you need. We have those that believe that hi-rez PCM is all you need. And then we have the real truth, and that is DSD. So along with your favorite digital files, now comes the arguing over the *best* type of DAC. I find it ironic that some people who exclusively listen to digital and pooh-pooh all things analog just for the inferior measurements to the ever more perfect digital measurements that they prefer NOS DACs which are the worst performing DACs technically of all the DACs in the digital wonderland. I sincerely believe that NOS DACs are the tube equivalent of SET amps. They both measure terribly in comparison to their brethren, but yet they are loved for their sound by those who love their sound.

mep, I appreciate your contribution to this forum, and respect your point of view, but you are so wrong.

 

[mep]

MarinJim-I appreciate your contribution to this forum and respect your point of view as well. Please tell me what I am "so wrong" about.

 

[Andre Marc]

I guess I shouldn't have been surprised at how fractured the digital religion is, but I was. We have those that believe that high-sample rate MP3 is all you need. We have those that believe that RBCD is all you need. We have those that believe that hi-rez PCM is all you need. And then we have the real truth, and that is DSD. So along with your favorite digital files, now comes the arguing over the *best* type of DAC. I find it ironic that some people who exclusively listen to digital and pooh-pooh all things analog just for the inferior measurements to the ever more perfect digital measurements that they prefer NOS DACs which are the worst performing DACs technically of all the DACs in the digital wonderland. I sincerely believe that NOS DACs are the tube equivalent of SET amps. They both measure terribly in comparison to their brethren, but yet they are loved for their sound by those who love their sound.

Not sure who you are talking about. I listen to RBCD, high rez PCM, and SACD. Its all good. I don't listen to formats,
I listen to music. There is no fracturing. The only ones discussing this crap are people who spend too much time on forums
and not listening to music. I probably listened to 30 albums while this thread has been droning on.

The vinyl crowd cracks me up when they wax poetic about Lps mastered from digital.