Hello everyone,

A lot of audiophiles approves the SACD formats, but some irreducible retorts that the DVD Audio is superior.
And you, of what side are you?

Garf's sig say's it all...

The answer to the SACD vs. DVD-Audio question is: CD -- DigitalMan

Please search and read the FAQ - this has been discussed many, many times here and I don't think another discussion would achieve anything.

Sorry, I didn't know it.

If it helps, I believe there are now hardware players that support both formats.

Thank you indybrett, but I know this domain very well.
Ayre, Linn, Méridian already develops for a long time this principle.

I wanted to know that rightly that uses each among us.

In my opinion, the only way to answer this is to perform blind comparisons. And to be sure that both tracks are coming from the same mastering. It seems that nobody did it.

I approve your reasoning on the principle of this comparison.

But what to say formats audios that constitutes the world of music.
Must one to trust the decoders audios, that identifies and read the formats ( on PC) or to electronics (sources, readers DVD or SACD of lounge) that treat it also.

My sig says another thing about that discussion, actually, from one of the foremost experts here

Personally I strongly favor DVD-A over SACD/DSD on technical grounds. This has been elaborately discussed here before.

Read for example http://sjeng.org/ftp/SACD.pdf

Favorite quote:

Good evening Garf,

Would the quality of a product be bound only to a technical proof?

I don't think it.
It is only my opinion: my ears first then the view.

What I know, an amplifier transistor will never be an lamp amplifier and even less a hybrid amplifier.
The comparison will always be uncertain and function of the individual's entity.

I Would Recommend DVD-A

What part of



didn't you understand?

The goal of the carrier is to reproduce the original signal as faithfully as possible. DSD is very suboptimal in that regard. That's a mathematical fact.

Distortion (like the one from your beloved tube amps) can (and should!) always be added later on, since it's entirely subjective what sounds "better" there.

I admire the clarity of your speech because you give proof on the one hand of truth.
Long life to the développemnt of your codec audio.

To soon

emphasis is mine

I don't know if this is even possible since the processing of DSD and PCM data is so different.

Right now SACD are mostly created from 96/24 masters (at best) and surely the converted DSD signal can not be better.

DVD-A !!

SONY has to be punished for their constant attempts to establish their own 'standards', with us, the users, being the ones paying the bills ...

What's all this talk about optical carriers about? Don't you know that is sooo 20th century...

.........

You would have to know, that many electrical and electronic engineers are not in agreement with this.

In many cases this is a market opinion, and it does not reflect the exact reality...

I doubt much that, due to the limitations of the human ear, you notice a difference between SACD, Dvd-a, and Audio CD (for normal stereo listening).

of course, if you notice differences this probably is as consequence of differents masterings...

Hello Acid Orange Juice,

"You would have to know, that many electrical and electronic engineers are not in agreement with this.
In many cases this is a market opinion, and it does not reflect the exact reality... "


Indeed, if you take the example of the very upscale: Naim Nac552/Nap500, Meridian, Dcs elgar, Ayre or Linn, it is true the analysis of the specter, the passing strip and the dynamics is not really the same.
But the charm of the tube (artisanal manufacture) will remain and will give a heat always little present among the big constructors of electronic high-quality.



" of course, if you notice differences this probably is as consequence of differents masterings... "

The mastering is often questioned in this case of face.
Must i think that an excelllent mastering of the SACD can surpassed a DVD Audio of middle quality.

Finally, this analysis makes myself think about the eternal return (psychoanalysis): the mobile picture of one immobile eternity.
It is strange, but that reflects there strongly: Mastering/Quality, Quality/Mastering.

Heh. An excellent mastering on CD has probably nothing to envy to any SACD or DVD-A.

And, again, technically, DVD-A is far superior to SACD. The point is if that makes DVD-A sound any better than SACD. But then, even when both are technically better than redbook CD, it's not clear that this can make them sound any better than good old CD. It's all about mastering.

The key advantage to SACD is that it can easily be distributed as a hybrid CD. I haven't seen any listening-test being conducted that clearly favored one over the other, so unless the music is targeted at bats, I think the world would be better off with SACD becoming the standard.

Just my 2¢.

What bills?

Hello guada2 ;

My opinions are based in my own experience as electronic engineer and as scientist; not by publicity or by opinions of third parts, that, others repeat like parrots without having idea of the subject which they are treating.

Their arguments are not more than their own and subjective opinions, and they are not based on facts, evidence or science; like either in engineering.

I have 4 years designing amplifiers of high fidelity. If you think that you will obtain better quality with an expensive amplifier of tubes, when you can obtain the same sound with an amplifier of transistors well designed (by a much smaller cost), then, you are free to waste your money...



... again ...

Your opinions are not based on facts or scientific evidence; they are based only in magical suppositions..

you did not pay attention to this:


It's a fact, not a supposition..
It's well known that the audio CD is completely overkill for the human ear. The normal audio CD have more resolution that the human ear is able to perceive. SACD and DVD-A have more resolution that the normal audio CD.. but.. for what?. It's not necessary for your ears.. maybe if you are a dog probably you hear the difference...
Of course, this is for the case of normal stereo listening.. For multichannel is another story..

To see if there are indeed audible differences between DVD-A and CD, would a valid test be to make a recording @ 96khz/24 bit WAV (a dvd-a spec), resample to a 44.1khz 16 bit CD spec WAV, and ABX the two, or would the 96khz recording be invalid because it wasnt first compressed in the DVD-A MLP format? Would differences become apparent through dithering noise?

Hello Acid Orange Juice,

"Their arguments are not more than their own subjective and expressed an opinion, and they are not based one facts, evidence however science; like either in engineering.
I haggard 4 years designing amplifiers of high fidelity. Yew you think that you will obtain better quality with year expensive to amplify of tubes, when you can obtain the same sound with year to amplify of transistors well designed (by has much smaller cost), then, you are free to waste your money... "


I don't put in doubt your experience, But to think that one DCS ELgar Plus/Verdi/Purcell doesn't make the weight, I am not sure. In short, it was only a parenthesis.

Indeed, I join you on a point: these are not the best electronic that pull the best resonant restitution. Because you know it all as much than me, that the quality of a source raises several criterias of tests.
After some years of comparatives and purchases of materials, I resolved to this stereo system:

Source: Linn Ikemi
Integrated amp: Icos Elsberg 270
New loudspeakers: Thiel CS2.4
Case of alimentaion sector: MPC
Cables HP: Synergistic Research
Cables of modulation: Esprit

It is not the dearest, but I am pleased of its capacity to answer best to the system that I searched for.

For what concerns video I reserve the right of all disclosure: scaler + spotlight + player DVD + préamp + amplifier multicannal + loudspeakers 5.1.

I just noticed, that DVD Forum has (or will) specified DVDPlus or DualDisc format, which is a double sided disc, one side could be DVD Audio, other side is CD. Check this:
http://www.disctronics.co.uk/technology/dv...dvdaud_spec.htm

I think this will help somehow spread of DVD Audio discs.

I prefer DVD Audio format!

Mine: CDDA.

Industry: The side they think will be the hardest the be put on P2P networks. And like Garf said, DSD being completely inefficient AND different it is a godsend. Sound quality? Oh yeah... that..................

Guys... you don't see the BIG picture...

Unfortunately, it looks like the SACD is "winning" the war as the next audio format. In records stores, I've seen the DVD-Audio section shrink and wither, while more and more albums are released on (hybrid) SACDs.

As witnessed many times in the past, the technically superior product doesn't necessarily win...

A few years back I used to think DVD-Audio would win, because all PCs would be able to read them with the right software (and a 24/96 sound card, granted), while consumers would need a new player to read SACDs. It may have been precisely the reason why music publishers turned away from DVD-Audio, with the fear that the protection on DVD-Audio would some day be broken as easily as CSS was on DVD-Video.

It's only a matter of time before someone breaks SACD's protection schemes, though. If you can create it, you can also destroy it.

Actually the hybrid CD/SACD disks have an interesting "protection", since a computer drive doesn't "see" the SACD layer, only the CD layer.

That's just one of the intellectual property protection schemes built into SACD. Another is physical (visible) watermarking of the disc, as well as data encryption. Since there is no such thing as an SACD "burner," there's less likelihood that an SACD disc will be trivially pirated, as is the case with the DVD family.

As to which encoding method is "better," that’s a tough and lengthy discussion…

First off, any reference to Lipshitz and Vanderkooy’s paper on the inability to optimally dither DSD data is inappropriate and outdated if you pay attention to current technology. As one of the few people who has taken the time to conduct extensive tests with both 192 kHz LPCM and single speed DSD with live sources (my company <www.sonicstudio.com> pioneered and manufactures both PCM and DSD professional production systems), I can say that they are subjectively different but both equally valid.

As was pointed out in earlier postings, it's the details of the formats that really highlights the advantages of one over the other. The SACD format was designed, from the start, as backwards compatible with the Red Book spec. The Dual Disc "format" is a bolt-on afterthought that lacks compatibility with CD players due to the physical characteristics (thickness) of the discs.

DVD-Audio also cannot deliver the highest fidelity LPCM in multichannel and is limited to 192 k stereo. All other factors being equal, 96 kHz LPCM does not sound as close to the source as DSD does, which means you are saddled with lesser quality 5.1 when listening to a DVD-A release.

Production-wise, both 192 k LPCM and DSD make great origination formats; material can be transcoded into any distribution format from either source file, and is typically done in "the real world" to accomodate the vagaries of the marketplace. 48 or 96 k sources just don't contain enough information to produce a subjectively lossless upsample.

Since the DVD Forum has seen fit to not mandate various player features that would make the consumer experience more enjoyable (persistant memory of user settings, group selection controls, complete front panel control of the player), DVD-A users have to have a TV just to listen to a title. As the “multimedia” capabilities of DVD-A, I’ve never found the slideshow feature to be compelling, especially considering the severe limitations placed on the author by the format’s limitations.

Frankly, for the average consumer, there is little in the DVD-A format that isn’t already delivered more easily and more compatibly by the DVD-Video format. Badly produced audio, whether DSD or 192 k LPCM, will always sound worse than expertly and carefully produced 48k/24. Simply looking at numbers on a spec sheet tell you nothing about the intricacies of a complete media delivery system. Read a lot of information from different sources, weight your budetary options (“universal” players are now cheap but titles in both formats are overpriced) and, more importantly, listen critically before making a decision.

I thought that sacd goes lossy if you put hard-to-encode hi-res multichannel content on it.

(btw: That part of yours smells like a terms-of-service rule #8 violation)

Are you implying that 96/24 isn't sufficient enough to transport sound transparently for us humans ? (I seriously doubt that)

Please elaborate on how you came to this conlcusion (and on the conlcusion itselt).


AFAIK the DSD stream isn't stored lossy on the SACD -- no matter what.


SebastianG

The 192kHz claim from the poster you have responded seems absurd to me. Forget 192kHz(96Khz bandwidth) -- at this point no confirmed perceptual tests have concluded that anything exceeding a [1]16kHz bandwidth is needed to be transparent for musical program playback for humans. The accepted JAES standard on this matter still stands as of this day.

-Chris

[1] Which Bandwidth Is Necessary for Optimal Sound Transmission?
G. PLENGE, H. JAKUBOWSKI, AND P. SCHONE
JAES, Volume 28 Number 3 pp. 114-119; March 1980

Do you have detailed perceptual studies avaliable for our analysis?




As the other poster noted, you have violated TOS of this forum. Audibility claims, when made in this fashion, must be accompanied by at least preliminary DBT results or solid correlation to a valid perceptual study.

-Chris

Allright. Then it must be the MLP (dvd-a) which does lossy when lossless isn't small enough.

That was my understanding too. Mind you, it'd be easier to know the answer if it wasn't such a closed protected system. Way to go, Sony.

http://www.hydrogenaudio.org/forums/index....ndpost&p=106542

DVD-A and SACD sales are declining (it's early for the "media of the future" )
HA has proved that even PCM is overkill for the human ear.
Talking about multi-channel, ac3 and DTS is widespread. For me DTS is fare superior and difficult to differentiate from the source. I didn’t make a real scientific test though. Just a friend changing the source.

I think you mean that 96kHz 24bit linear PCM is overkill for the human ear. PCM merely refers to the way the signal is sampled and quantized - with no specification of sample rate or bit depth. For example, 22kHz 8bit linear PCM is certainly not overkill for music signals (it is for speech, though).
If it's not a double blind test, then making any sort of conclusion from it puts you on fairly shaky ground. However, I agree with you that multichannel is the way things are going (people have mixed opinions about this) but I don't see why, with rapidly increasing storage space, lossy multichannel formats should have any advantage over 6 channel LPCM.

Sorry I didn't mention it, I was talking about 16/44 PCM

Yep, that's why I have started the following topic:DTS/ac3 tests, Is it worth it?

Outside the size benefits (which still an advantage even with big hdd, ...), with ac3/DTS you have the compatibility with all the DVD player...

I doubt you will get much support for the opinion that 16/44 PCM is overkill for the human ear. Perhaps "adequate" would be better - because that's exactly what it is - enough, but not too much.

Agreed, compatability is a big plus. It will be interesting to see the outcome of your surround listening tests. Unfortunately I don't have a decent 5.1 system, so I can't participate.

Well, it's the principle of lossy encoder. You can’t say the difference between a lossy at around 200 kb/s and a PCM 16/44 at 1411 kb/s. It means pcm 16/44 is overkill for the human ear. I doubt you will get much support to claim the contrary here at HA.
Outside the compatibility, the benefits of pcm being lossless, meaning you have the choice of the lossy encoder without transcoding artefact. You still have lossless compression at around 800 kb/s.

I haven't read that paper, but a clean 16kHz low pass is clearly not transparent to many younger listeners for much source material. This was demonstrated during the tuning of the --alt-presets, and even before during work on the old --r3mix preset - filtering at 19.5kHz was audible to one person who took part in those tests!

Cheers,
David.

The testing in the JAES article, which was eventually peer reviewed and accepted as a standard, used test signals that were designed to test respective fixed phase and spectrum content, simulating analysed music(with boosted high frequency spectrum in order to increase sensativity ). The conditions of playback were conistant for all test subjects(half of which were audio professionals - the results of both groups were segregrated). It is true, that perhaps very young people could score significantly better in the musical program HF testing, but since very young people were not used, it remains to be a theory.

I don't know specifically which factors/variables were used during this testing of which you refer. I suspect it may have been the standard *remote* testing as is standard during the codec developments on hydrogenaudio.org. In the case of remote listening, a simple matter of a person testing at much higher SPL levels for the test than is realsiticly feasible during actual listening may greatly influence the test results. They might use a device that has a very inaccurate frequency response anamoly that greatly enhances the ability to hear a specific band.These are just two examples, and of course there are others that limit the practical usefulness in such a remote testing method. I recognize this remote method is certainly useful for guidance of codec designs , especially considering the alternative of no group perceptual testing -- but the lack of careful consistant test conditions and fraud suspectibility(I.E.; depending on the test subject's honesty) reduces the merit and reliability of such testing. It could be that the test results of the testing to which you refer were accurate -- but please understand my skeptiscm based on the factors present in such testing.

-Chris

one need only try to ABX musical content with --aps to realise the tests are statistically significant... i can't ABX preset standard except on problem samples (and these are mp3 flaws, not bandwidth related problems), but i can pretty consistently ABX --preset standard -Y, which is basically a 16khz lowpass (except where short-blocks are triggered).

as far as DVD-A vs SACD, that's a no-brainer. the processing is all done in LPCM, so why not leave it there?

i'm not yet a multichannel convert. for most, it's simply a fad. look at the average 5.1 setup in people's homes - the speakers are in all the wrong places, often obscured completely by couches or ornaments or whatever, pointing in bizarre angles, etc etc. to say nothing of the much-too-high subwoofer crossovers, and ludicrously exaggerated bass. most systems are set up to fit the decor, not for good sound.

i'm sure i'll change my opinion if i get a really good 5.1 system, but right now i'll settle for an above-average 2.0 system (movies still sound stunning on anything capable of decent dynamics and flattish response).

All those problems are certainly possible (or probable!) in such tests - but these "problems" also apply to normal people listening to normal music under what they consider to be "normal" circumstances.

In other words, if playing back the music louder than "normal", or using poorer transducers than "ideal" causes the 16kHz low pass to be audible, then it's audible - and will be audible to some listeners some of the time in what they consider to be "normal" listening.

If we forget deafening loudness levels for a moment, all the other factors you mention are only partly relevant anyway - does it really matter if someone detects a difference by listening with a pair of headphones that boost HF by 6dB? This simply means that, on a different track with 6dB more HF, they'd still identify the problem on a pair of headphones with a flat frequency response.

What's more, taking a "normal" recording, and boosting the HF by 6dB, may well give you much less HF than is found on some recent electronic or experimental music.

I agree that it would be useful to test people like Garf, Dibrom etc under ideal conditions, but I'd suggest that the results we already have show that 16kHz is too low - unless you believe all audible differences apparently due to content above 16kHz were due to ITD in the equipment producing actual changes below 16kHz in the output.

Cheers,
David.

At the moment, the conclusions that have been drawn in reference to encoder testing and HF that you refer to are valid only for the set of circumstances -- but isolated perceptual data, in a controlled setting, must be aquisitioned if you want to be certain of that variable's specific contribution. Several conditions could also be tested to account for the speculations we have both made RE: HF content, etc.. However, when I am forced to pick between some tests not isolating factors or establishing a set of controls vs. a peer reviewed JAES article/research, it's no contest as to which one I will consider as more credible/useful. It is possible, of course, that some very young people with very HF ability could detect the difference with music that has unusual power distribution into the HF spectra. The article also concluded that(with the even a slightly lower point filtering that was slightly audible) that a lack of the highest frequencies was not detected as a sound quality decrease in blind tests.

Could you link to the test files used in this ABX test to which you refer? I would, as a matter of seperate interest, be interested in taking a look and listen.

Thanks.

-Chris

Too bad that the benefit of surround is dependant on competant mixing of the surround channels by the same illustrious folks who bring you the *excellent* average 2 channel recordings of today. Cough. Cough.

-Chris

I think each the SACD and DVD-A camps make claims that its format is insistinguishable in tests from live feed while the other isn't. At least I have heard both claims. I am not sure who has tested this with CD, or how else your result came to be "well known". Take a reference ADC to convert to 16/44 and reference DAC and compare to live analog or else take 24/192 recordings and convert to 16/44 to compare.