SACD - is it better?

Posted by: bam on 10 May 2001

Does SACD sound better than CD? Any first hand experience to share?
Reading a website recently it stated that SACD records 1-bit samples at 2.8MHz. CD records 16-bit samples at 44.1kHz. By my calculations this means that the sample values of CD can change by 65536 amplitude quanta between two successive samples but SACD can only change by 63 in the same time.
Is SACD going to destroy yet more musical information? Is this a cynical deception by Sony et al or do they really think the difference cannot be heard? Will it replace CD? Are my facts wrong?
BAM
Posted on: 10 May 2001 by Sproggle
Given what you've said,

CD provides 16 x 44.1 = 705.6 kbits / second, while

SACD provides 1 x 28000 = 28000 kbits / second.

What it does with those bits is another matter, but it seems reasonable, to me, to assume that SACD will eventually be a lot better than CD.

--Jeremy

[This message was edited by Sproggle on THURSDAY 10 May 2001 at 19:31.]

Posted on: 10 May 2001 by Martin Payne
the DSD layer on the SACD I heard certainly was certainly better than the compatibility layer played on the same Sony player.

Having heard the player in a couple of systems on CD I'm pretty sure that the CD replay is well behind that of a CDS-II.

Juan, can you remember the model number?

cheers, Martin

Posted on: 10 May 2001 by bam
My understanding is that CD samples are signed 16 bit numbers that linearly represent the analogue waveform amplitude.

Imagine a 20kHz sinewave at full amplitude - in this case the successive sample values (as there are just over 2 samples per cycle) must very nearly change by full scale. You may argue that this will almost never happen because the recording engineer will not record such high levels, but let's say the sinewave is at 1/10 full amplitude - successive samples will vary by some 6500 quanta.

Since the delta modulated SACD method can only record 63 level changes in the same time then how can the same amount of information be contained? In fact, wouldn't it be the case that the delta modulation would require the analogue signal to be low-pass filtered to reduce the high frequency amplitudes to avoid slewing problems? I'm reasoning that the high frequencies will see much poorer amplitude granularity than the low frequencies.

On another tack...
"As an example: if instead of 1 16-bit sample, you have 2 15-bit samples, providing they have been separately obtained, you can average them to give effectively 1 16-bit sample. So the two are equivalent from an information point of view."

So is it correct to extrapolate this to 1 bit by increasing the number of samples by 2^16 or 65536, thus requiring 44100*66636 samples/sec or 2.9 billion samples/sec to retain the same information in a single bit sample stream?

BAM

Posted on: 11 May 2001 by Sproggle
quote:
Imagine a 20kHz sinewave at full amplitude - in this case the successive sample values (as there are just over 2 samples per cycle) must very nearly change by full scale.

Ouch! Sounds painful!

Surely the result of sampling such a sine wave at 44.1 kHz would be a triangle-ish wave of varying amplitude?

--Jeremy

[This message was edited by Sproggle on FRIDAY 11 May 2001 at 12:40.]

Posted on: 11 May 2001 by bam
I found a web site last night that has a concise, relatively understandable explanation of how delta and delta-sigma modulation works:
1-bit converters
Posted on: 11 May 2001 by Derek Wright
From what I remember in a recent copy of the Musical Fidelity Club newsletter they commented on SACD and DVDa - indicating that they would not be producing players for these mediums - comments included that the copy protection / watermark process really dumbed down the SACD sound quality (they likened it to MP3) and they commented on the hens teeth amount of software.

Re DVD-A they commented that you need a video screen to access the controls to play the medium correctly -

AS I have now binned the newsletter _ I may have remembered inaccccurately in which case apologies - no deliberate offence intended etc.

Jonathan - you have some MF kit - did you get the newsletter.

Derek W

Posted on: 11 May 2001 by Derek Wright
Yes I did listen to the disk - I was more interested in the music - I will listen to it again with more a more critical ear, however I will catalog it in my list of CDs

I received a MF CD several years ago and I thought that one was excellent.

I have found the MF CDs more accesssible than the Naim sampler CD that I received with the CDSii - It was not until I got the brace of 135s warmed up to replace the A370 did I find the Naim sampler CD to be worth listening to. - so I better keep my head down while the rocks of derision fly overhead. <g>

Derek W

Posted on: 11 May 2001 by Tony Lockhart
I think the main reason that cd is so successful is that vinyl had so many perceived handling and wear problems, and the silver disc has almost none of these. The man on the street will not see or hear any advantage in the new formats. For most people, cd is much more than good enough. For them vinyl was a pain. I don't think 'audiophiles' are the bulk buyers of cd music.
Tony
Posted on: 11 May 2001 by Martin Payne
quote:
comments included that the copy protection / watermark process really dumbed down the SACD sound quality

AFAIK there is no copy encryption on SACD.

DVD-A includes such a feature, and the launch date of the medium was put back by six months after a group of hackers broke the scheme even before any players were released!

I believe there is a smattering of DVD-A players around, but they don't have decryption capabilites - I think there was some sort of pledge to upgrade them once the scheme was finalised.

The encryption and 'MLP' (non-lossy compression) do worry me as they move us further away from a 'minimalist' system.

However, SACD is about as minimilaist as it can get, and to me it just feels like a cynical attempt to cash in on a new technology. In theory DVD-A has a greater ultimate potential than SACD, for the reasons mentioned above.

cheers, Martin

Posted on: 12 May 2001 by Top Cat
Technically SACD is miles better than CD IMHO, but the players don't seem to be there yet. However, given that it took ten years before I heard a CD player that I could live with, and how much CD players have subsequently come on in the last 8 years or so, I am willing to believe that SACD could be mind-blowing given the right equipment. Problem is, most amps aren't capable of doing the ultrasonic stuff, so I imagine that a lot of the benefits would be lost with any amps which roll-off at 20KHz, although I imagine the likes of Naim and so on will be OK with this (being detailed and airy up there anyway)

John