DSD comparison to PCM.

[manisandher]

Hardly you can go on anectdotal information from years past when the MSB DAC has had at least 2 ver. update with numerous firmware changes. Much less the Galaxy/Femto clock architecture that was not even developed by then.

Haha, you could say exactly the same about the gen 1 NOS1 you tried (and dissed) vs. the current 24/768 async-USB version. And furthermore, I suspect by your lack of reply to my earlier questions that you didn't use an activated version of the latest XXHighEnd, configured properly, to do your evaluation of the gen 1 NOS1. Yep, really thorough and a great contribution to anyone reading this thread.

Mani.

 

[opus111]

So he got his opinion/measurements and posted last year and then he said it was "some time ago"... what... 2005? 2003? Hardly you can go on anectdotal information from years past when the MSB DAC has had at least 2 ver. update with numerous firmware changes. Much less the Galaxy/Femto clock architecture that was not even developed by then.

Yes I agree - I'm not basing my view of the current MSB DACs on the one he measured back then. But it is nevertheless interesting that it seems like the architecture used in that DAC was engineered specifically to ace a particular DAC measurement, that of linearity vs level. To the extent that they are using such graphs in marketing now (and I've no idea if they still are, they were last time I looked which was about a year ago after I read this post of Bruno's) then chances are they are still using it in current incarnations. But this is conjecture.

I need to pick music/projects apart under a microscope. I don't need colorations and certainly don't need anything "rolled off".

I agree - the roll-off in most NOS DACs is indeed very off-putting and definitely unsuited to professional use.

I spent the better part of the past 7yr. putting together the most cutting edge studio I can so my clients will get the best possible sound. I don't pick the H/W because it sounds good... I chose it because it was true to the source.. be it live mic feed, analog master tapes or one off direct cut vinyl.

True to the source by ear or by null testing? If you've not tried the latter with DSD, why not give it a whirl?

I truly wish ya'll the best of luck in your search for nirvana with RBCD but I'm moving on. CD's are the past. I need to look in the future and try to guess where the industry is going and be one step ahead. Time is money.... I have enough of one.

CDs are the past but I still prefer a physical media so its CDs for me for the forseeable future because they last longer than flash disks or hard drives. If I had a hard disk from when I first started buying CDs I doubt I'd be able to read it now (assuming it hadn't died in the meantime) without some special purpose hardware - in those days they were using ST506 or ESDI interfaces I think.

 

[NorthStar]

-- MLP, Meridian Lossless Packing, from hi-rez on DVD-Audio; any audio merit among our two main chief leaders (DSD & PCM)?

 

[opus111]

That's like asking if the copy of Excel you're about to buy runs better if delivered in .zip or in .rar. Tangential to the debate.

 

[MarinJim]

I love living in the past. However, I painstakingly love living in the present also. I don't care about the future, until it is the present. I own a 15 year old BMW 540i (supercharged to 700 hp) and 23 year old BMW 535i with 200,00 miles. Both are in pristine condition. I take the same approach to my music system. I wish I never sold my Jon Brown modified Conrad Johnson Premier DAC 9 with upgraded caps. IMHO chips are like chips, can't eat them without good salsa.

 

[PeterSt]

Haha, you could say exactly the same about the gen 1 NOS1 you tried (and dissed) vs. the current 24/768 async-USB version. And furthermore, I suspect by your lack of reply to my earlier questions that you didn't use an activated version of the latest XXHighEnd, configured properly, to do your evaluation of the gen 1 NOS1. Yep, really thorough and a great contribution to anyone reading this thread.

Mani.

With the notice that someone (Mani) is defending his own ears while on the other hand someone (Bruce) is defending his decision, I personally must say that it is only sad when something doesn't get a chance because it was not understood. So of course Bruce did not reply because -by now- what to say ? "sorry I was stupid ?". In any case I can of course tell that Bruce did not use any activated XXHighEnd. So now this little speculation game can be ended.

Also see next post.
Peter

 

[MarinJim]

With the notice that someone (Mani) is defending his own ears while on the other hand someone (Bruce) is defending his decision, I personally must say that it is only sad when something doesn't get a chance because it was not understood. So of course Bruce did not reply because -by now- what to say ? "sorry I was stupid ?". In any case I can of course tell that Bruce did not use any activated XXHighEnd. So now this little speculation game can be ended.

Also see next post.
Peter

When I saw Precision A/V (US distributor of Venture Audio and Weiss), use your DAC, I knew your DAC is SOTA. Congratulations on your success Peter.

 

[PeterSt]

@Bruce

Hi Bruce,

I don't think you will have recognized me as the guy who back in April 2008 asked you for some DXD samples; I had created 24/352.8 playback as a first (through WASAPI) and couldn't do it with the only handful around from 2L. You did (or tried to start up) DXD. Remember ?

I don't know anymore what you used, but I think it was Pyramix. I don't know it from 2L anymore either, but I do know that *after* contacting you I talked to the tech guy at 2L and learned that they were using some dedicated ADC from a brand that never made it to market but somehow was DAD related. Somewhere Switzerland was involved.
Back at the time I was not interested in the formats used on the path to the physical media, but I recall that you transferred to DSD while recording in DXD (which latter is nothing else than 24/352.8 PCM of course).

From theory we should be able to "state" that no other DAC can be better used for DXD playback than a NOS DAC which can play (input) at that rate. As far as I can tell, one exists (two actually, and one of them you dissed or ditched or whatever).
What is your view on this ? any perspective to put this all in ?

The very much to the thread's subject related question is this though :
In my view nothing comes even close to those 2L DXD files. Some hurdle is taken (somehow) and normal "Hires" in any (failed ?) format is nowhere in comparison (btw you may have an opinion on this too). Now :
What ADC could they have been using (or still use) in terms of SDM, R2R-multi-bit etc. etc. ?
What ADC did YOU use at the time ?

Somehow I feel this is important (for general knowlegde and/or future) but I myself combine it with modern audio analysers which are not even capable of more than 192ADC were it about 24 bits, and when I browse for A/D chips I only find noisy ones (and of course SDM only).

Thanks and regards,
Peter

 

[PeterSt]

When I saw Precision A/V (US distributor of Venture Audio and Weiss), use your DAC, I knew your DAC is SOTA. Congratulations on your success Peter.

Haha, thanks. Yes, I noticed your sig.

But don't tell Daniel, because when he's around the corner all is quickly switched. What a game ...
(right now all is manouvred into Chicago )

 

[Orb]

Well Grimm does not make a DAC. This is an A/D converter. It's the best one I've heard so far.... bar none! Every single tape transfer that I do gets run through the Grimm. Even when I blindly offer files to clients, they always choose DSD..... even after a conversion to PCM!

Ah so the DD1 not made yet?
They ever give you any indication when it will be available?
Very interesting and thanks for the heads up.
The AD1 is very interesting in the spec and way they went with DSD.
Cheers
Orb

 

[Orb]

Do you have any links to share - whilst I don't always agree with Bruno, he's definitely worth reading whatever topic he's talking about. Last I heard he thinks the reason people like NOS DACs is the frequency droop, he also wasn't at all impressed with the MSB DAC he connected to his AP - he got around 13bit dynamic range which varied with the signal level as if the internal electronics operated on a gain-ranging basis. Kind of like a floating point DAC.

Bear in mind this goes back to 2003, but it does reflect IMO some aspects of the AD1 (which I posted a link on that included some thoughts about DSD as well).
Hopefully the DD1 will be released with a similar paper.
This is the original thread that has had the interview quoted quite a few times but they miss out the later discussion.
Post #58, however continue reading down afterwards because Gardo post was sent to Bruno and his answer was posted.
http://forums.stevehoffman.tv/threads/sacd-fundamentally-flawed.26075/page-3

Cheers
Orb

 

[opus111]

Love it - classic quote from there is :

I'm only saying that of all the possible solutions that could be formulated for the deficiencies of CD, DSD is the least intelligent one.

 

[Orb]

But following it through fully Opus, including the paper from the link associated with the AD1 earlier; DSD does have a place he says.
Furthermore he mentions streaming is more efficient with DSD rather than other solutions.
That was why I did not want to post the link Opus, because I had the feeling it would be used selectively by each side of the argument
Context and implementation is key IMO to what he says, and reflects the AD1.
If it is that bad they would not develop what Bruce feels is the best AD product, what will be interesting is where this goes with the DD1 (which I admit I thought was available).

Cheers
Orb

 

[opus111]

I hear him saying that SACD sounds better than RBCD so I'm puzzled - then I hear that he has his own DAC when he listens to SACD. If that DAC is using noise shaping (seems most likely as he's no fan of multi-bit) then that explains his preference for DSD. Implementation only gets us so far - when the underlying mathematical basis is broken (as is the case with 1bit noise shaping), its not going to be fixable by fancy implementations.

Incidentally I've read down to post #75 on the third page, is there more worth imbibing? I note there are 21pages in total...

 

[Orb]

I cannot remember as its been quite awhile since I went through that whole thread.
Interesting discussion considering when it started though, and how long the AD1 has been around.
Cheers
Orb

 

[opus111]

I'd not heard of the DD1 until you mentioned it - it seems it'll probably be based on PWM from what he says, with very high order (perhaps 7th?) noise shaping. I'll gladly pitch my multibit hires design up against his PWM one when the time comes - for now I will just tentatively reserve the name 'DD0' for my DAC

I have been persevering with that thread, I note a reference to an AES paper by Reefman of Philips which a poster says addresses the issues raised by Lipshitz - I am delving into it now, attached.

 

[Orb]

Thanks Opus and if notice anything else in the thread please post.
Just to re-iterate, worth putting that thread and especially the variables mentioned by Bruno in context to their more recent whitepaper (in a very loose sense):
http://www.grimmaudio.com/whitepapers/dsd faq.pdf

Cheers
Orb

 

[Bruce B]

Hi Bruce,

I don't think you will have recognized me as the guy who back in April 2008 asked you for some DXD samples; I had created 24/352.8 playback as a first (through WASAPI) and couldn't do it with the only handful around from 2L. You did (or tried to start up) DXD. Remember ?

I don't know anymore what you used, but I think it was Pyramix. I don't know it from 2L anymore either, but I do know that *after* contacting you I talked to the tech guy at 2L and learned that they were using some dedicated ADC from a brand that never made it to market but somehow was DAD related. Somewhere Switzerland was involved.
Back at the time I was not interested in the formats used on the path to the physical media, but I recall that you transferred to DSD while recording in DXD (which latter is nothing else than 24/352.8 PCM of course).Thanks and regards,
Peter

Glad you could join us Peter!

Back then I was using Pyramix and the DAD AX24 (which I stil use today as rebadged as Sphynx 2) and the dCS 905/955.

I feel that if a DAC needs "special" software to show its best, then inherently something is broken. You don't need special software to run any other DAC I know of because if you do, put it in the DAC to begin with. Bricasti and others can do it.

And yes, I use null tests all the time. It's an important tool to use when you "think" something sounds different.

 

[opus111]

Thanks Opus and if notice anything else in the thread please post.

Google turned up a second Derk Reefman paper as well as the earlier one - attached. I've read through the first one now, suffice to say I'm underwhelmed. It overturns nothing that's in the Lipshitz paper - he explicitly says he is steering clear of the theory and only relying on simulations. But he provides scant detail on how he has produced the measurements from the sims he's run. So for example he shows how the noise floor of two SDMs (one dithered the other not) doesn't vary with DC level (except below -118dB on the undithered one) but gives no clue as to how he measured the noise floor figure.

View attachment DerkReefman_DSD_wp-2323.pdf

<edit> In your linked FAQ on DSD, there's a claim there that DSD is more sonically transparent than 24bit/96k. I'm curious what converter hardware this was determined with - do you have any light to shine Orb? Would it be a safe guess that a NOS DAC was not used?

 

[Orb]

In certain context it is probably more transparent such as AD to storage.
Not much help without a DAC to hear this though
TBH one reason I thought the DD1 existed, need to revisit a lot on the subject and read it all carefully to be able to answer your question in detail that would be satisfactory.

Cheers
Orb