SACD fundamentally flawed?

DVD-A does provide for watermarking, not sure if it is being used or not, or by what releases.

There are psychoacoustic watermarking techniques that are format-independent; working on all digital and all analog formats. Sometimes I wonder if any of our CDs and SACDs are already watermarked.

 

Gabe: DSD is not multi-bit; it is Delta Sigma Modulation. The exact same mastering technology is utilized on the final SACD disk!

Alan, I have it on good authority that Sony Music mastered the stereo CD and SACD layers "flat", with no eq. The few disks remastered from multi-channel tapes for 5.1 had to have been eq's to match the sonic signiture of the initial mix, but without bass attenuation. The 6.1 sampler disk with one track from the not yet released Live 1964 does not sound properly eq'd; Perhaps this is why the project was put back a few months?

Richard.

 

Blank

 

Exactly. Anyone can be using this engineer's name and posting on Usenet. Unless I personally know someone and have contacted them in the past elsewhere, I trust nobody on Usenet to be who they say they are.

 

Well DSD is really just a marketing term and in the profesional world is really a 5 bit (or now 8 bit) system - in other words PCM. They don't like to call it PCM, but that's what it is. It is then ported down to a 1 bit PDM system that is the sacd disc that we all use.

DSD is a marketing term and SACD is trade name.

 

Yes, SACD has a fundamental flaw

Isn't he just saying (like many before him) that DSD is a shortcut to a dead end?

DSD's signature flaw comes from its extreme noise shaping that leaves a characteristic sonic signature IN THE AUDIBLE BAND.

You are right, yesman, to find irony in our longing for the old days of RedBook PCM (who could have foreseen it?!) but they seem a lot more neutral than DSD...

Really, isn't it sort of obvious that all DSDs sound THE SAME on the top end?


imho!


FabFourFan
------------
Mike Patrick

 

I'm a bit reluctant to accept this posting a face value as well; although I think reconsidering the assumption that DSD is inherently superior to PCM worthwhile. While each of the writer's points contains at least a kernel of truth, there does seem to be at least a little axe-grinding going on. One example, his first point regarding clicks when switching is technically true although his co-workers presented a paper at the 2001 Amsterdam AES convention ("Editing and switching in 1-bit audio streams," Derk Reefman, Peter Nuijtten ) concluding, "The clicks introduced by switching bitstreams using this algorithm are on average of the order of -90dB. Listening tests have reveiled (SIC) that these effects are not audible by experienced, professional listeners." Some of his other concerns can also be overcome or minimized with the same kind of care and planning while recording/mixing/mastering that good engineers have applied to overcome PCM's and analog's shortcomings for years. Interestingly, he fails to mention many of DSD's real problem areas; limit cycles, improper dithering, etc. (Lipshitz, Vanderkooy, and others do a better job of this than I can so I'll leave it to them).

It's his DSD creation story that caught my attention. While I'm certainly no Sony insider, DSD was originally conceived as an archival format that operated at 5.6MHz. It was my understanding that it was necessary to halve the sampling rate in order to deliver acceptable playing time on the consumer format. 64fs was a compromise made to bring it to market, not part of the original design. Or maybe I'm wrong...

Ultimately, what's important, though, is that DSD sounds pretty darn good; as does hi-res PCM. Which one is better? Since very little music is likely to be issued in both formats, I don't think it really matters- if we want the music, we're going to be buying both until one goes away.

 

Did I miss 5-bit? The Sonoma processed at 8-bit (like Sonic and Sadie), remodulated to 1-bit, and then (once properly synchronized) switched back to the unconverted stream. Merging does the same thing at 32-bit/352kHz. The dCS converters do 5-bit/2.8MHz (or 3.0MHz depending on source rate), but what else?

 

Re: Yes, SACD has a fundamental flaw

IMHO, SACD done right sounds absolutely heavenly.

Let me use an analogy. Back in the days of Dolby-S and MiniDisc, David Ranada ran an article where he showed that the step response and noise in the audible band of MD was a lot worse than Dolby-S.

He missed the point of ATRAC: Sure, there is extra noise, and on a purely theoretical / graphical sense, Dolby-S is better. But all the noise shaping made ATRAC sonically indistinguishable to many reviewers from CD. The same couldn't be said from the Compact Cassette using Dolby-S.

I think that's the case here. If there's noise shaping, yet the format sounds right, then the implementation was succesful, and SACD is a step forward in audio reproduction... I tend to trust my ears more than theory written on paper. MHO.

 

Yes, SACD has a fundamental flaw

But the format DOESN'T SOUND 'RIGHT'!
It imposes an artificially 'smooth' top end on the sound.
Which sounds sort of nice.
At first.

---

So let me ask again a little differently: don't all of the details on your DSDs SOUND THE SAME, as if they are being actively generated by the software?

(But be warned about listening for this - once you catch on to the 'trick' sound of DSD,
you may come to agree that it's the work of the devil, not something from Heaven... )


imho!

FFF
------------
Mike Patrick

 

Re: Yes, SACD has a fundamental flaw

No.

 

Re: Yes, SACD has a fundamental flaw

I'll bite...

Well, uhh... no. I don't think they sound the same on top (for some definition of "top").

Example, the CCR SACD's mastered by Steve sound completely different (tonaly) than the Sony SACDs that I own, which all seem to have a similiar "clinical" sound. (Example: Boston S/T).

Mind you, I don't have the world's most expensive set-up (Sony 775 into Rega amp into JMR speakers).

What's your system config, FFF? (might fill out your profile, hint hint, nudge nudge )

jeff

 

About DSD being the devil's work, yes, of course.

But when other forum members keep repeating that DSD has no audible defects,
and that analog audio is now 'over with' - that always gets me going!


FFF
------------
Mike Patrick

 

Re: Re: Yes, SACD has a fundamental flaw

What you are saying is correct. I did not express myself clearly in my original post.

It's not that simple, unfortunately, to describe what bothers me,
but it's the nagging suspicion that the edge of audibility is entirely suspect,
when you are so dependent on extreme noise shaping in the audible band.
Noise shaping is a perceptual trick.
Somehow, the pro dsd crowd thinks that nothing bad is happening,
because the sample rate is so high, and thus the DSD miracle.

Remember, I agree that DSD mostly sounds soooooo nice!!!

But, maybe if they raised the sampling frequency they wouldn't have such
problems managing the noise, and then everyone would be satisfied.
Oops, I forgot, we're not supposed to mention that DSD could ever be
improved, because IT HAS NO FLAWS. Ha ha. Sorry! (to be spoken like
one of the Philips engineers mentioned in the quoted posting which
started this thread)


As always, just MHO!

FFF
------------
Mike Patrick

 

Let's not confuse the two:

DSD and SACD should not be confused (even though everyone uses the two terms interchangably).

What people call DSD is not the same as what is on an SACD. DSD is usually multibit (5 or 8) PCM and SACD is 1 bit PDM.

 

I've heard that in theory airplanes aren't supposed to fly but they do.SACD sounds better so lets get on with the show and save the music industry.

 

I don' t understand this discussion. (Or maybe I do. )

Who has said DSD has no flaws? Who on earth other than marketing departments would ever buy the "perfect sound" hype?

Does DSD sound better than hi-rez PCM? Many recording and mastering engineers say yes. They're in a position to have heard more controlled and apt comparisons than I have. If I trust their work otherwise, I'm likely to trust their ears on this one. I know my best-mastered DSD discs sound great to my ears.

Does DSD color the top end of the sound? I haven't noticed this. The top end of my SACDs don't all sound the same. The top end of my Tubular Bells sounds different from the top end of my Abraxas sounds different from the top end of my Zombies Greatest Hits sounds different from my Bruckner 9th (Harnoncourt). I do hear greater dimensionality and better decay resolution on SACD than with DVD-A, but I can't do a very meaningful comparison between the two for all the reasons everyone has already stated.

FWIW, I too have a very modest rig, which I have detailed in my profile. What are you listening on, Mike? I ask not to invalidate your opinion but to get some idea of how in- or out-of-sync we might be in terms of our systems.

 

Re: Re: Re: Yes, SACD has a fundamental flaw

I don't think we ever said SACD was flawless. Nothing is. The point is, maybe the flaws aren't as big as some people would like us to believe.

I remember reading an article about someone that said SACD was inherently inferior to a 128K/s MP3, because of its flaws.

I've learned I can't believe everything I read. That's the reason why I now trust my ears.

 

One more time:

Let's not confuse the two:

DSD and SACD should not be confused (even though everyone uses the two terms interchangably).

What people call DSD is not the same as what is on an SACD. DSD is usually multibit (5 or 8) PCM and SACD is 1 bit PDM.

 

Most of the critisism against SACD isn't about DSD, it's about the one bit oversampling process which is the delivery system of SACDs.

DSD is good, I don't think there are any disagreements about that. The issues that have been raised are regarding SACD's 1-bit system.

 

GabeG. I have spent some time with several Philips and Sony folks on the SACD Team. It has been reported in the early press reviews of SACD that the disk was one bit and the DSD master was multi-bit; members of the SACD Team have assured me that this is pure myth. Both the mastering technology and the consumer disk utilize the identical form of Delta Sigma Modulation! The reason that Sony Music and the Sony SACD Team does not label their final mastering project as pure DSD (as some other venders have) is that many projects utilize mastering tools that are either analog or multi-bit/PCM. After mastering the project is stored as a Delta Sigma Modulation project. The use of the term for one bit to describe Delta Sigma Modulation is common but not accurate. Now it could be that every member of the SACD Team is full of bull; but I don't think so. Some of the discussions with the SACD Team and Philips can be searched on http://www.forums/hirez/bbs.html.

Richard.

 

I believe Richard has it correct.

 

Richard, with all due respect, I spend time with several members of the AES and it's pretty public knowledge among members that SACD and DSD are indeed two different things. A couple of years ago this was brought up at an AES meeting (which was pretty well discussed online) and Sony admitted that the archival format (which they label DSD) is indeed multibit PCM and the delivery medium (SACD) is the one bit PDM. Now it could be that they are full of bull, but I don't think so.

 

DSD is multibit PCM??? Isn't DSD PWM?