PS3 SACD Playback

[The Seventh Taylor]

I don't see why not, if H264 at high bitrate takes up 3 SPEs (apparently, and that's the most processor intensive decoding it will have to do), then there's loads of free room for other tasks.

The heaviest A/V processing task it gets to do is converting DSD (the SACD audio format) to high-resolution PCM. This takes 5 SPEs according to Sony. Check out this article: http://pc.watch.impress.co.jp/docs/2006/1117/mobile356.htm

The main reason why SACDs sounds so much better than regular CDs on the proper equipment has more to do with production/mastering than the technical merits of the format.

The main reason lies not in the signal-to-noise ratio, not even in the increased dynamic range or the higher resolution, but in the far better accuracy in the time domain. With 'red book' CD audio (44.1 kHz PCM), unless eveything in the chain from mastering to playback equipment is excellent, the audio often "sticks to the speakers": you can hear it coming from the left speaker and right speaker but there's almost nothing in between. SACD uses 64 times the sampling frequency of CD and the result is a far better reproduction of the sound stage.

I can hear no difference switching between the CD and SACD layers on hybrid SACDs.

I have done many SACD demos and found that even people (like me) without 'golden ears' can quite easily appreciate the difference between stereo SACD and CD. I've even had a person deaf on one ear who noticed a clear difference.

When you switch to multichannel SACD the difference is overwhelming. This is where the real benefit for me lies.

For more info about SACD on PS3 have a look at this FAQ: http://www.ps3sacd.com/faq.html
There's also a general FAQ about SACD: http://www.ps3sacd.com/sacdfaq.html

 

[Carl B]

I moved this post into its own thread from the original bumped thread because thread necromancy like that is discouraged.

I'd point out also Seventh Taylor that the article you reference for London-Boy is actually included within the first post of the thread you had bumped, and indeed that thread was the original source for the spread of the story. (thanks go to One)

 

[Gubbi]

With 'red book' CD audio (44.1 kHz PCM), unless eveything in the chain from mastering to playback equipment is excellent, the audio often "sticks to the speakers": you can hear it coming from the left speaker and right speaker but there's almost nothing in between. SACD uses 64 times the sampling frequency of CD and the result is a far better reproduction of the sound stage.

I'm not sure I'm buying this. Recording is done almost exclusively in PCM. Downsampling to 44KHz shouldn't destroy phase information the way you describe it. And can't imagine jitter would be a problem on modern equipment (ie. PCM signal data is buffered)

Cheers

 

[-tkf-]

I'm not sure I'm buying this. Recording is done almost exclusively in PCM. Downsampling to 44KHz shouldn't destroy phase information the way you describe it. And can't imagine jitter would be a problem on modern equipment (ie. PCM signal data is buffered)

Cheers

I think Jitter and Downsampling is still considered some of the great enemies of high quality audio.

On AVS forum the guy that produced the HiDef movie "Natures Journey" expains the great lenght he went through to avoid to much "tampering" with the Audio. I´ll point you to the Blu-Ray thread of the Disc since the HD-DVD thread was fast filled with HD-DVD supported being angry over several issues. Most of the issues seemed to come from the Blu-Ray version spotting a Lossless Audio track and a Higher Bitrate transfer of the movie. In any case this is a true HiDef Disc done to the outmost perfection.

http://www.avsforum.com/avs-vb/showthread.php?t=868185

 

[Bohdy]

The main reason lies not in the signal-to-noise ratio, not even in the increased dynamic range or the higher resolution, but in the far better accuracy in the time domain. With 'red book' CD audio (44.1 kHz PCM), unless eveything in the chain from mastering to playback equipment is excellent, the audio often "sticks to the speakers": you can hear it coming from the left speaker and right speaker but there's almost nothing in between. SACD uses 64 times the sampling frequency of CD and the result is a far better reproduction of the sound stage.

Is that right? You seem to be talking about the surround mix, which is more to do with the mastering/recoding process than the sampling rate.

Besides that, from what I understand of Sampling Theory, 44.1khz should be high enough to avoid aliasing in the time domain for any frequencies that humans can discern, thus the "64 times the sampling rate" claim is not very significant.

 

[randycat99]

To add to the "jitter" debunking, I would also comment that it seems quite unlikely that the current state of technology should still yield problems with jitter in the kHz range, let alone sub-Mhz range, when we are already routinely dealing with consistent timings and clockrates on digital devices that operate on the order of GHz. If we couldn't address/mitigate/eliminate jitter issues at a mere 40 kHz, we wouldn't have a chance of getting the underhood workings (millions of discrete logic units) in a 3-4 GHz CPU running consistently and deterministically.

 

[Rainbow Man]

But Randy, you don't get it...

The human ear is th emost sensitive measuring instrument ever! Especially in a non-blind test situation.

if you don't have equipment costing at least $5000 apiece and platinum-plated triple-insulated, phazon-coated audio cables with arrows indicating the flow of current, you just can't have good sound! The jitter will literally KILL you!

Peace.

 

[-tkf-]

But Randy, you don't get it...

The human ear is th emost sensitive measuring instrument ever! Especially in a non-blind test situation.

if you don't have equipment costing at least $5000 apiece and platinum-plated triple-insulated, phazon-coated audio cables with arrows indicating the flow of current, you just can't have good sound! The jitter will literally KILL you!

Peace.

And some just don´t care about sound at all..

 

[randycat99]

Personally, I like to shoot heroin to get that audio nirvana sensation as my system plays. That way I am addressing jitter and hysteresis electron flow right at the ear nerve, in lieu of spending $500/ft on silver core audio cable!

 

[-tkf-]

To add to the "jitter" debunking, I would also comment that it seems quite unlikely that the current state of technology should still yield problems with jitter in the kHz range, let alone sub-Mhz range, when we are already routinely dealing with consistent timings and clockrates on digital devices that operate on the order of GHz. If we couldn't address/mitigate/eliminate jitter issues at a mere 40 kHz, we wouldn't have a chance of getting the underhood workings (millions of discrete logic units) in a 3-4 GHz CPU running consistently and deterministically.

Sure

http://jitter.de/english/sound.html

 

[randycat99]

You could just as well post a link to the "crying Brittany dude". Basically, the site shows how they had to put a cable 10x longer than the standard length in order to develop an anomaly on the order of nanoseconds. Additionally, they fail to acknowledge that if the edge of a pulse still has enough definition, it still gets read as a digital "1" just as well as a "perfectly" shaped pulse. A "1" is still a "1", regardless of the relative quality of the "1". That's the whole point of digital.

So, again, it looks pretty well debunked, afaik. Stick with a standard length cable (not a whopping 10 m) and your pulse envelopes will look the way they are supposed to look. Consequentially, the d/a convertor will create the waveform as it was intended. Jitter will be entirely up to the clock stability at this stage, not the envelope quality of the digital pulses further up the chain.

 

[-tkf-]

You could just as well post a link to the "crying Brittany dude". Basically, the site shows how they had to put a cable 10x longer than the standard length in order to develop an anomaly on the order of nanoseconds. Additionally, they fail to acknowledge that if the edge of a pulse still has enough definition, it still gets read as a digital "1" just as well as a "perfectly" shaped pulse. A "1" is still a "1", regardless of the relative quality of the "1". That's the whole point of digital.

So, again, it looks pretty well debunked, afaik. Stick with a standard length cable (not a whopping 10 m) and your pulse envelopes will look the way they are supposed to look. Consequentially, the d/a convertor will create the waveform as it was intended. Jitter will be entirely up to the clock stability at this stage, not the envelope quality of the digital pulses further up the chain.

Of course Jitter is just a fantasy, there isn´t a jitter problem. There never where and there will be. MP3 is all we need.

 

[randycat99]

Did I make any reference to mp3 being the end all/be all solution? That's a total non-sequitor.

It doesn't change the fact that "jitter" continues to be this mysterious boogyman conjured to attribute audio problems that just don't exist. It doesn't change the fact that if we had such problems clocking out signals in the kHz range, there's no way we could possibly get GHz processors off the ground. That's just a non-dismissable smoking gun to the "jitter" argument.

 

[-tkf-]

Did I make any reference to mp3 being the end all/be all solution? That's a total non-sequitor.

It doesn't change the fact that "jitter" continues to be this mysterious boogyman conjured to attribute audio problems that just don't exist. It doesn't change the fact that if we had such problems clocking out signals in the kHz range, there's no way we could possibly get GHz processors off the ground. That's just a non-dismissable smoking gun to the "jitter" argument.

Yeah it´s pretty weird that we have 2ghz CPUs and that jitter is still considered a audio problem by purists and can be meassured as well. But in a world where MP3 is just as good as uncompressed PCM i´m not surpissed you dimiss it.

 

[Rainbow Man]

And some just don´t care about sound at all..

"Caring" isn't the same as being smart.

Or even rational for that matter.




Just to grab some arbitrary numbers purely out of thin air, if we say that to reach 80% 'sound quality' (imaginary measurement) it'll cost you $2000.

Then to reach 90% it'll cost you $6000. Huge increase in dollars for a very small, likely unnoticeable performance increase.

And an audiophile system that in reality may only reach 88% will cost you $11500 and up because of the overhyped and baloney pseudo-science materials and construction techniques used.

Yet human hearing can only appreciate 70% 'sound quality'! You're spending thousands on equipment that can only be appreciated by dogs and bats!

And don't give me that BS about "oh no but with my superior appreciation of expensive HiFi equipment (and large wallet) I can hear a difference!" because no. You're not some kind of ubermensch. You're just a guy with a strong belief, like a Mac fanperson for example!

Peace.

 

[Pete]

http://jitter.de/english/sound.html

How objective is this site? From the English home page:

"www.jitter.de is published by Altmann Micro Machines, a german manufacturer of audio equipment, phone +49-2129-54260."

The page you linked shows how the Altmann Ultra-Precision Clock Injector (a "high-end consumer audio" device, per its data sheet) corrects the jitter in the signal from a cheap consumer electronics item (portable CD player) output over a decidedly nonstandard 10m cable.

I don't know much about jitter, but that page doesn't seem to present a sparkling indictment of perceptible jitter in even low-end devices.

And it's (more than) a bit of a troll to accuse those skeptical of jitter polluting sound as being so unsophisticated as to be content with an MP3 over even a CD, let alone SACDs.

 

[-tkf-]

How objective is this site? From the English home page:

"www.jitter.de is published by Altmann Micro Machines, a german manufacturer of audio equipment, phone +49-2129-54260."

The page you linked shows how the Altmann Ultra-Precision Clock Injector (a "high-end consumer audio" device, per its data sheet) corrects the jitter in the signal from a cheap consumer electronics item (portable CD player) output over a decidedly nonstandard 10m cable.

I don't know much about jitter, but that page doesn't seem to present a sparkling indictment of perceptible jitter in even low-end devices.

And it's (more than) a bit of a troll to accuse those skeptical of jitter polluting sound as being so unsophisticated as to be content with an MP3 over even a CD, let alone SACDs.

The page was an example, if your interested you can use Google and spend a few months reading up on the subject. The first time i read about Jitter was when the DVD players were introduced, there was alot of controversity in the Magazine calles Widescreen Review and alot of discussion about the subject, it wasn´t sponsored by a manufactor btw. The conclusion back then was "yeah we have a problem".

My MP3 example was just in line with the other trollish remarks about "expensive cables with the right color" sounding better. It´s all about where the personal limit is, some say they can hear jitter and some semy deaf people can´t hear the difference between MP3 and PCM.

Do i care about jitter? no not really, do i have special cables? no do i use a MP3 player? yes.

And thanks for pointing me out as being the troll while ignoring the others.

if you don't have equipment costing at least $5000 apiece and platinum-plated triple-insulated, phazon-coated audio cables with arrows indicating the flow of current, you just can't have good sound! The jitter will literally KILL you!

Personally, I like to shoot heroin to get that audio nirvana sensation as my system plays. That way I am addressing jitter and hysteresis electron flow right at the ear nerve, in lieu of spending $500/ft on silver core audio cable!

You could just as well post a link to the "crying Brittany dude".

 

[randycat99]

Perhaps you are a victim of friendly fire. If that is the case, then I apologize if my remarks came off too strongly. It's just very common that when someone speaks so confidently about the perils of "jitter", it's like holding a sign up that says, "I have no idea what I'm talking about, but I read about it someplace that looked really fancy." Upon further questioning and discussion, it usually does come out that this person really has no idea what "jitter" is about. The same scenario frequently plays out for the "fancy cable" crowd (hence, the style of remarks that ensued).

As for my shooting heroin remark, there really was an underlying message to it (albeit, broached in humor). Think endorphins vs. heroin and how the physiological impact can influence the perception/sensation of sound.

 

[Rainbow Man]

The page was an example

..And I'm not impressed by it at all. Especially not the "testing" section towards the end where they don't describe the testing methodology used, nor present any quantifiable evidence in their supposed audio tests. IE, they don't run an actual audio signal through their fancy oscilloscope to show any appreciable difference.

To show a digital pulse-train becomes fuzzy on a nanosecond timescale is one thing. The digital pulses aren't what we hear in the end. And to propose that these nanosecond variations turn into any kind of audible changes - much less big ones like the page you linked..is just prepostrous. Completely prepostrous.

if your interested you can use Google and spend a few months reading up on the subject.

The problem is that 'reading up' on this subject is most likely just going to be more of the same contrived hyperbole and overexaggerated unscientific nonsense as the page you linked to.

The conclusion back then was "yeah we have a problem".

That was in like nineteen-fricken-eighty.

I think a thing or two has happened since then yes?

They themselves admit you need a ten meter cable to coerce jitter on the millionths of a second timescale. How could anyone possibly hear that?

And who has to use a ten meter cable anyway, why not just switch to optical then?

And thanks for pointing me out as being the troll while ignoring the others.

It's not trolling to point out the nonsensical nature of much of the so-called audiophile market. It wasn't actually aimed at you personally.

Peace.

 

[-tkf-]

The problem is that 'reading up' on this subject is most likely just going to be more of the same contrived hyperbole and overexaggerated unscientific nonsense as the page you linked to.

Peace.

Or it could be that no matter what you read you would just dismiss it because your PC runs @ 3ghz

Here is more to dismiss:
http://www.tnt-audio.com/clinica/diginterf1_e.html