Whats the Worst thing about SACDs?

That all sounds bad on paper, but honestly, does it make that much of a difference in the real world? Has anyone here done direct comparisons of 96/24 and 192/24 transfers? How about DSD and 48/24 as well?

It seems that most pro/anti-SACD/DVD-A talk is based on what people want to believe, not how they actually sound.

 

Actually, I have heard a difference between 96/24 PCM or 192/24 PCM and DSD. I don't 'want to believe' anything. I own a player for both formats. As far as I'm concerned, they can both continue on or one of them can drop off the face of the Earth, I'm covered either way. I just don't believe that the DVD-A PCM resolutions are as good as DSD.

 

You've heard the same material, mastered by the same person, in all three formats? And you've switched between all three without knowing what was what?

 

Well, without divulging too much information, let me just say, yes, not exactly and yes.

 

I like DVD-A because I think the multi-channel mixes have generally been better than what I've heard on SACD MC layers.

Problem is that I don't like multi-channel mixes...

 

Sony's cardboard slipcovers on their older SACDs -- that's the worst thing.

I'd sure like to listen to my new Stevie Ray disc, but as a collector I don't feel right about destroying the packaging to get to it. A conundrum of sorts.

 

There is really no difference between 24/96 and 24/88.2 from a sonic perspective.

I guess the reason why they chose 24/88.2 for stereo was because they mastered the stereo portion with red book CD in mind. A 24/88.2 master can be easily "downgraded" in the digital domain to 16/44.1 by shaving off 8 of the least significant bits and dropping each other sample to the floor. It is straightforward to do from a technical perspective, and requires no sophisticated math. So I guess this was chosen from a practical point of view, with re-use for red book CD in mind.

You don't need to worry about resolutions. Listen to the product. Choosing the right resolution is a matter of the mastering process and is to some extent determined by the source material and purpose. Different resolutions mean a degree of freedom in the mastering stage.

24 bit is 256 times higher resolution than 16 bit. Even 24/48 is a significant improvement over red book, and gives you twice the playing time per disc compared to 24/96 while giving you most of the sonic benefit. There is empirical evidence that increasing the number of bits from 16 to 20 or 24 is an actual sonic improvement, given some dynamic source material. There is hardly any evidence that increasing the sample rate above 48 kHz makes any difference at all.

 

Well, it's debatable whether increasing the bits offers any audible benefit at all either. Actually, not to start another whole debate here, but it is debatable whether these 'Hi-Rez' formats offer any real, audible benefit over 44.1/16. Even though I have expressed a preference for DSD and SACD over DVD-A, I personally feel that they are all relatively transparent. I don't know if any difference I may be hearing has anything at all to do with the format as opposed to the mastering chain, mastering process, original source or recording etc. My complaint about random resolutions on DVD-A's was really more about knowing what you're getting and simplifying this for the average consumer or even the less than educated 'audiophile', or burgeoning 'audiophile' if you will. Personally, I don't feel they should be marketing me a 48/24 disc as 'Advanced Resolution' when it is barely a higher resolution than a DAT tape that I could have purchased almost 20 years ago.

 

No, I think we've moved forward on this point.

Regards,
Geoff

 

In your opinion, perhaps.

 

"A 24/88.2 master can be easily "downgraded" in the digital domain to 16/44.1 by shaving off 8 of the least significant bits and dropping each other sample to the floor. It is straightforward to do from a technical perspective, and requires no sophisticated math. " (petzi)

Let's hope they don't take that approach. It is true that 88.2kHz can be converted to 44.1kHz by ignoring every other sample, a process called decimation. But "shaving" off bits, called truncation, is nasty business and several well-known algorithms (UV-22, Super Bitmapping, POW-R, etc) have been developed to shorten wordlengths without truncating. That being said however, you are probably correct in assuming that 24/88.2 was selected because of the relative ease with which it can be retasked for red-book CD release.

On to the business at hand. The worst thing about SACD is how little Sony is doing to support it on the production side. To date Sony has yet to release for sale any professional recording products on which to produce true DSD material (Tascam, a company that Sony holds an interest in, did release a stereo DSD-capable recorder). Other companies have released editors, converters, and recorders; until late last year, these were all of the 8-ish track or smaller category. Now one 48-track DSD recorder is available (not from Sony). Sony hasn't even released DSD processing cores for its existing digital products. It seems strange that Sony would float a new format without supporting the means of generating product in that format.

 

Luke, as usual, states the truth.

I have been involved in some non-blind listening tests between differing resolutions of PCM as well as DSD and while I have my own preference, once you get above 24/88, the differences get very small between resolutions (including DSD). Granted, these weren't completely controlled circumstances (I knew which was which), but I think many people are getting worked up over very small differences (in most cases).

- GABe

 

I've heard similar comparisons too. I heard differences with each also.

I heard the same thing. I agree with you on this too.

 

Well I read that it has been proven mathematically that there is no difference between shaving off the least significant bits and attempting to do better by using any special algorithms. That said, science is not what bothers a certain kind of audiophiles.

But whether or not that is true (i.e. if you challenge math), at the very best, any difference that can be made by using any "special" procedure to shorten word length, can only manifest itself in the least significant bit, otherwise it would be a procedure that alterates the sound through distortion (which could be audible)

So in either case, even if math were wrong, if there is a difference at all, it is going to be minimal.

Besides, most audio material consumed these days could be represented by 12 bits just fine, due to lack of dynamics, but that is a different story.

 

I know of careful listening tests that have been made by the Tonmeister community, and the results were that, higher bit depth than 16 resulted in audible improvement when very dynamic source material (which is rare) was used, but not when music with low dynamics was reproduced. Kind of makes sense to me.

As I pointed out, 24 bit is 256 times the resolution of 16 bit. This is a fact. In terms of THD+N it results in a possible improvement of roughly 48 dB.

The same listening tests were made using higher sample rates than 48 kHz, and perceptible phenomena were not found, however, this issue is still under investigation. Please keep in mind that (even Steve says that) almost all audio material hardly reaches 16 kHz on a good day. There are hardly any microphones that record such high frequencies, and even if they could, those microphones have roll-off filter built into them for ultrasonic. So the sound doesn't even make it into the console in the first place.

 

And there lies the rub. Most of the music outside of Classical and perhaps some Jazz does not have very wide dynamics, therefore a higher bit-rate or sampling-rate really is beneficial only in that, when it is interpolated down to 44.1/16 Redbook, that there is less loss of information than would normally occur resulting in a more accurate reproduction.

Yes, I am well aware of this.

This is very true, which is why I've always said that 44.1/16 Redbook was more than accurate enough. No analogue tape contains anything of a resolution that would push the boundaries of a Rebook CD, our ears cannot hear beyond 20KHz and , you are correct, microphones don't really pick up much at these higher frequencies anyhow.

It is when you hear a full digital PCM or pure DSD recording of very dynamic information captured by hypersensitive microphones that you get a more accurate, even if small, picture of the differences between the resolutions and between PCM and DSD. I have heard such recordings, identical in every way expect one was PCM 192/24 and one DSD, or one was 96/24 PCM and the other DSD and there is a real, but barely perceptible difference.

 

"There is really no difference between 24/96 and 24/88.2 from a sonic perspective."

This I will agree with having worked in both extensively. I do draw the line at 24/192 - based on my listening you need to at least have this to match the transient capture of DSD.

 

Keep in mind that the album in question (L.A. Woman) was recorded in Manzarek's house over 30 years ago....

I'm sure the DVD-A sounds comparable to the master on even the lower scale of high end stereos.

I happen to like it, myself.

 

Goldenboy,

Then why is it that at least 90% of my SACDs sound noticeably better at vocal reproductions and bass definition that any of my redbook titles?

I'm not saying CD can't sound amazing, but from your statement I seem to draw that SACD couldn't conceivably offer easily discernible differences vs. standard redbook playback.

 

Not enough titles and I think the prices should come down. And I would say the same thing about DVD-A.

May we get more and cheaper titles in both hi-rez formats, and sooner or later a release in both formats would be nice so we can decide for each and all of ourselves which one we like better.

 

That's not quite true. It wasn't his house.

The Doors workshop was on La Cienaga Boulevard, across from Elektra's "10 percent-discount recording studio" (Ray's words from his book "Light My Fire", p.335)

Here's what Ray says in an interview:

Question from Alex: What was recording LA Woman with Jerry Scheff and Marc Benno like, I heard you guys recorded live?

Ray Manzarek: Yes, Alex, LA Woman is virtually a live recording, with everyone playing at the same time, and Jim in the bathroom as his isolation/singing booth. We recorded the album at our rehearsal place on the corner of La Cienga and Santa Monica. It was an incredible experience. The energy in our small rehearsal room was overflowing and manic. What a great two weeks we had. As you know, Jerry Scheff went on to play with Elvis Presley. All the footage you can see of Elvis Presley in Las Vegas has Jerry Scheff on bass.

Regards,
Geoff

 

Much of it could conceivably be credited to a number of factors: a better master, a better re-mastering job, some SACD's (and DVD-A's) are actually re-mixed, (i.e. Bob Dylan Blonde on Blonde) that do not offer an easy one to one comparison. I think that the DSD D/A conversion process is slightly less likely to introduce any 'mechanical' degradation due to decreased circuitry as opposed to PCM DAC's. In short, I'm am not totally convinced that any and all differences and 'improvements' apparent on SACD's or even DVD-A's is solely the byproduct of the capturing process itself, be it DSD or Hi-Res PCM.

 

So, are you saying that any or all or most of the analogue masters that you work with reach the 20KHz/98db range of Redbook?