DVD Audio vs SACD (1 Viewer)

[John-Miles]

Hi I am new to this forum, and i was hoping I could get some information on what the difference is between the two. I have heard soem say that SACD is a lower quality than DVD Audio, however I have also heard that SACD is the same thing as a master recording form a studio and therefore the best you can get aside from being there live.

I would love some feedback on this, either responses, or tell me where to get this information.

Cheers

John

 

[John Garcia]

Everything I've heard is that SACD is HIGHER quality than DVD-A. Every recording is only as good as it's mastering... I'm leaning towards SACD myself.

Try asking this in the Music forum, as there is a fair amount of discussion on it there.

 

[John Tillman]

There has been some outstanding discussions on the formats in the past couple of months. Search the Hardware & Music sections for the threads.

 

[Jeff Adams]

It's really hard to say. I was looking for a great progressive scan dvd player when I came across the JVC XV-723GD and as a bonus it also had dvd-audio. I have really enjoyed dvd-audio. It is really amazing. I only have 4 dvd-audio discs and love them. Metallica and Steely Dan have never sounded better! With how inexpensive dvd players are with one of the two formats included, it's a sin to not be enjoying one of these 2 formats. I just hope an affordable dvd player with both formats comes out soon. The only thing I am frustrated about is the lack of software for bot formats. At least the lack of software I want to buy. I would like to see some new releases coming out. The Red Hot Chili Peppers would be sweet on dvd-audio!

 

[Michael_T]

Having both a DVD-Audio player and two SACD players (one stereo and one multichannel) I have to say that SACD wins hands down in overall sonic quality.

I have DVD-Audio titles that rival SACD in sound quality, but the smoothness and overall sonic signiture of 90% of SACDs released is something to behold.

Many people take issue with DVD-Audio and the Verance Watermark that is supposedly applied to most DVD-Audio discs to prevent digital piracy. They say the watermark is audible (I have yet to hear it, but then I am not listening for it at all) and is reason not to support DVD-Audio.

I say, that if you find software to your liking that happens to be released on DVD-Audio - go for it. If your musical tastes tend to run toward those titles released on SACD - then that should be your choice. (At this point in time is appears that Warners will not be releasing any SACDs and Sony will not be releasing any DVD-Audio discs. Yet other labels seem to be releasing both, or will release both, although many labels (such as Universal) have said they will support SACD.)

If you could only chose between one, and wanted the best sounding medium, then SACD would have to be your choice.

 

[Jagan Seshadri]

Here's the summary as I know it, analytically speaking, and I apologize in advance for being a bit dry here
Note: To get to the point, without the technical explanations, go to "Debunking SACD/DVD-A".
SACD is only better than DVD-A if the SACD is made from a "live-to-DSD-Tape". This is because the recording process simply transcibes the a highly sampled 1-bit data stream (known as PDM: Pulse Density Modulation, or "DSD" in Sonyspeak) from the recording studio's analog-to-digital converter. This is the big deal about SACD: maintaining recording purity by avoiding a lot of processing (the less calculations, the less round-off/overflow errors happen).
DVD-Audio, on the other hand, uses a lower sampling rate but every sample contains more information, but this recording technique (used in CD's and known as PCM: Pulse Code Modulation) tends to go through more calculations:
- Conversion from 1-bit studio signal to PCM
- Conversion from PCM into analog
The differences between DVD-A PCM and CD PCM:
- DVD-A PCM uses very descriptive 24-bit samples (as opposed to 16-bit samples)
- DVD-A has 6 channels instead of the two on a CD (of course)
- DVD-A channels are sampled more often, resulting in a wider frequency response, which is very good since the CD's response didn't have a lot of safety margin built in, whereas DVD-A has ample safety margin (as does SACD).
The differences between SACD (PDM or DSD) and DVD-A (PCM):
- DVD-A PCM must be sent through a DAC to get converted back to analog (a trickier conversion to do)
- DVD-A PCM 24-bit samples are taken 96000 or 192000 times each second
- The entire frequency band of DVD-A has a dynamic range (signal-to-quantization noise ratio) of approximately 144dB (beyond the thermal noise of microphones themselves!)
- SACD/PDM can be put through an analog low-pass filter to demodulate it into an audible-frequency analog signal (purer conversion)
samples per second
- SACD's 1-bit modulation scheme give more dynamic range (more PCM-equivalent bits) to low frequency inputs and less dynamic range (less PCM-equivalent bits) for higher frequencies.
Debunking SACD/DVD-A:
- For multitrack recordings, multiple 1-bit tracks must be combined. As far as I know, to combine multiple uni-bit signals you need a multi-bit representation, which is the very antithesis of SACD's purity argument. Hence my earlier statement about SACD being superior only for live-to-DSD recordings
- To theoretically mimic a CD's PCM signal in true 1-bit resolution would require 2,890,137,600 samples per second (44100*2^16) per channel, so SACD's 1-bit sampling rate (at about one thousandth of that rate) is a tradeoff. (Correct me if my math is wrong on that statement).
- DVD-A's claim of 144dB dynamic range is theoretically true but practically is false since the noise of microphones and electronics only allows for 120dB at best typically unless you happen to live in liquid nitrogen
However, the extra range (i.e. bits) allows for the necessary PCM processing to happen without creating noise beyond what you already would experience in any piece of electronics.
So what was your question again?
Both formats are superior to CD's, and are multichannel too! As for which one makes more sense, DVD-A makes more sense technically, whereas SACD makes more sense to fulfill Sony's goal to keep royalties pouring in (after all, their royalty agreement for CD's expired). The timing of SACD is no coincidence, but have they learned from their Beta and MiniDisc endeavors?
Technicalities and political maneuvers aside, best thing is to trust your ears. One problem is that no recording exists in both formats to do a fair comparison (with the possible exception of Telarc's 1812 Overture??).
For the record, I'm onside with DVD-A.
-JNS

 

[KeithH]

I agree with Michael here. SACD is the better-sounding medium, but some DVD-Audio titles are very good, and DVD-Audio still beats CD. I have bought into both formats (one DVD-Audio player and three SACD players), and SACD continues to amaze me. The clarity, depth of the soundstage, etc. is downright incredible. DVD-Audio while good, doesn't match SACD.

 

[Howard_S]

Wow. This is getting me to think about SACD. Which is like THE SACD player to get right now?

 

[Jaehoon Heo]

I just want to add a win hand on DVD-Audio. I worked on that area for a couple of years, and I know that DVD-Audio is superior to SACD theoretically, and can be better in actuality, too.

DVD-Audio utilizes efficient PCM coding(at least better than bitstream DSD). DSD(technically PDM) is a derivative from PCM through noise shaping. DSD is good for conversion, but quite inefficient in storing audio information. Current SACD recordings are made from PCM original master to avoid the inefficiency of analog-to-DSD recording. Sony/Philips admitted this.

In an AES convention this year, Dr. Lipshwitz(not sure for spell) proved that DSD recording has inherent and unavoidable problem in quality due to its noise-shaping. PCM's superiority in information efficiency(thus more detail and higher performance(SNR/DNR...)) to bitstream coding is well known and even claimed by renowned ARA(Audio Renaissance Association) and Dr. Hawksford.

I also admit that SACD recordings made from high-resolution PCM master(this may be from analog master.) can be very good. But let's think about this: noise-shaping from PCM master to bitstream can be done NOT ONLY in studio like SACD, BUT ALSO in players like DVD-A/CD. Most DVD-Audio players use delta-sigma type DAC's and they do it.

Then who will do it better? Studio's sophisticated and adaptive noise-shaping can outperform low to mid-end DAC in player. However recent semiconductor technology allows very linear multi-bit delta-sigma DAC performing very good. So, SACD approach is just LIMITING the possibility of better playback of original master!!!!!

Actually, some high-end DAC(still below $20/pc) from Analog Devices, TI, Wolfson, etc., or dCS module(I think you know this highest-end studio DAC maker and Elgar) simply outperforms the DSD specification. Some DVD-A player may be underperforming, but DVD-A format has absolutely more potential.

I know that many people thinks SACD recording/players sound better. One reason is that just the SACD studios are making acoustically better recording.

Another reason is that all DVD-A players are not built as good as the SACD players are. Even low-end SACD players have more budget to invest for sound quality than mid-priced DVD-A/V players. Most player-to-player comparisons are not fair.

The other possible reason is the watermarking scheme. Watermarking is reportedly damaging the sound quality quite much, and in this case Sony's watermarking scheme seem to be better than one used in DVD-Audio format. But Bob Stewart of Meridian mentioned that watermarking need not be applied throughout the music, just some part of it. And watermarking is even optional. Many minor labels don't do it. I hope the major labels to stop applying watermarking, as I don't see any good mean of copy protection by adding watermark noise into audio.

 

[John Morris]

I've got both a pair of DVD-A players (RP-91 & LA-85) and a single new SACD player (sony 555)and I decided to use my new SACD player as my refernce CD player. I did not want to use a switcher for my 5.1 audio and so I just decided on what I heard. It was not scientific, nor was it fair, I just got a feeling about which 5.1 music I liked better. The final straw was that without doubt, regular CDs sounded better, to me , when played back by the SACD player.

 

[RicP]

Jagan:

Your entire argument against SACD falls apart when you realize that Pro-DSD recording uses 8-bit DSD, not 1-bit.

SACD uses 1-bit DSD on the consumer end, but 8-bit DSD in the pro Audio recording arena.

One reason is that just the SACD studios are making acoustically better recording.

Ludicrous! How in the World do you explain the dramatic increase in sound quality on SACD's where the performance was recorded 20 years ago? Did they make the recording "acoustically better" (whatever that means) 20 years ago because they knew that they'd need to for the DSD conversion 20 years later? That's a ridiculous argument.

I'd be glad to argue the theory and whatnot, but one listen to an SACD recording "live to DSD" or not, and the astounding quality is readily apparent.

 

[John Morris]

Did they make the recording "acoustically better" (whatever that means) 20 years ago because they knew that they'd need to for the DSD conversion 20 years later? That's a ridiculous argument.

I just listened to the Earth Wind and Fire SACD tonight, and compared it to my EWF album, and I'll tell you, this may be the first CD that I fully preferred to the vinyl version.

 

[John Kotches]

Jagan,

Can we discuss some of the points you've made in a little more detail? It's some definite food for thought on my end.

DVD-A's claim of 144dB dynamic range is theoretically true but practically is false since the noise of microphones and electronics only allows for 120dB at best typically unless you happen to live in liquid nitrogen

As I said earlier, SACD is already delivering at maximum potential. DVD-A has about 4dB left to improve. It's hard to say at this time what improvements might come in the signal chain.

Another item not mentioned (and I am no DSP engineer):

The ability to do digital crossovers via DSP arithmetic is impossible to do in a DSD environment. As best as I can determine it is impossible to do a correlation between the time (1-bit DSD samples) and frequency (spectral content) domains of the DSD data set.

Interesting points made.

Regards,

 

[Jaehoon Heo]

RicP,

Please note that I don't want to violate you. As far as I know, most of commercial SACD recordings are made from PCM masters. Maybe I am wrong. But using noise-shaped signal as the master will create great cost and pain to do any kind of digital signal processing. Imagine adding two DSD streams or doing some filtering. They may use 8-bit 64fs bitstream at AD conversion, but I believe the master is in PCM format for easy mixing and various processing. So, the bottom line is that even using 8-bit ADC, the master quality will be about the same as the master for DVD-A, theoretically.

In PCM side, high-resolution ADC's like dCS904 use 4~5 bit 64fs ADC. They may use 8-bit like in DSD recording. There I see little possible difference at the master tape stage.

Then the difference will be made by noise-shaping in DSD studio, and DVD-Audio player.

DSD's playback principle is using 1-bit(2-level) quantizer, as originally intended. DVD-Audio players can do anything as the designer designed in the chip. Most DAC's for DVD-Audio use 3~6 bit DAC with some kind of dynamic average weighting technique to minimize multi-bit distortion. Theoretically, using more quantization levels in quantizer provides better SNR with the same level of side effect of noise-shaping, or same SNR with lower level of side-effect of noise-shaping. Side effect of noise-shaping, I think, is well adressed by many audio theory specialists like Dr. Lipshwitz.

Another bad thing about DSD in playback is the difficulty to apply some digital signal processing. Future audio equipments are expected to be almost fully-digital. We will want to apply many kinds of digital signal processing like room acoustic compensation, digital volume control, sound field processing etc., to the outcome of SACD/DVD-A player. DSD is not just in the right format, and need to be converted to PCM for that processing. DSD will complicate the digital audio chain.

 

[PatrickM]

A couple of things that nobody has brought up yet is the fact that the majority of SACD titles people are comparing to DVD-A titles are all stereo recordings. The majority of DVD-A's out there only have 24/96 stereo tracks. There are very few 24/192 stereo discs out there right now and I would hold off judgement until we hear more 24/192 discs.

Secondly, the types of SACD and DVD-A players out there vary greatly. The majority of SACD players began in the high end, audio only, audiophile world like the SCD-1 and are migrating slowly down to mid-fi and even inexpensive lower end units with the recent addition of the low end Sony unit.

The DVD-A players on the other hand are all DVD players first and DVD-A players as a feature (albeit a big feature). Unfortunately the DVD-A standard forced the inclusion of video which is something that audiophiles just don't like (at least most of them) which means we will probably never see a high end, audio only, audiophile DVD-A player which makes comparisons a little more difficult.

I actually don't have an idea of which is better because I've yet to buy an SACD player (although I really want one) but I do have a DVD-A player because it was just part of the RP-91 I bought and I am enjoying the DVD-A titles I've purchased and as someone said before, it kicks ass on most Redbook CD's.

Patrick

 

[John Morris]

Okay, although these latest posts have my head spinning, lets see if I understand?

- DVD-A by its' nature can be a higher fidelity playback of an original master but is not inherently now.

Then, without considering formats, why is my Sony SACD player a much better sounding unit, than my RP-91 when playing back regular CDs using their analog outputs? Is it that the DVD-A format has the potential to be better, but the actual components used to play back these formats are currently tilted in the other direction, favoring SACD?

 

[Jaehoon Heo]

RicP,

One more comment to you: Some recordings made 20 year ago may be acoustically better than SOME recordings made recently. And you should think that you didn't compare the SACD with DVD-A with same master. We seem to point to different points.

 

[Jaehoon Heo]

John,

Could you tell me what SACD player you are using?

 

[RicP]

Jaehoon,
Most of your argument seems to revolve around the supposed difficulty in performing DSP on a DSD stream.
I think that you're making a large assumption that somehow there are not Pro-DSD tools for performing DSP on a DSD stream, when Sony says that there are such tools not only in studios now, but more being developed.
The Vanderkooy/Lipschitz paper was adequately refuted by Sony at the last AES meeting as well.
There are noise shaping issues with DSD, but do they somehow render the format useless or crippled? Absolutely not.

And you should think that you didn't compare the SACD with DVD-A with same master.

And you have compared the two directly? The Telarc 1812 overture is available in both DVD-A and SACD from the same master, and most reviewers seem to prefer the SACD almost unanimously.
Have you really listened to SACD?

 

[John Morris]

Jaehoon: I am using a Sony SCD-C555es unit.