HQ 24/96 sound explained

igor_levicki · Aug 8, 2005, 02:17 PM

You say:

Quote:

Originally Posted by ursamtl

The war zone comes up when comments along the lines of "you seem to be the type of person..." come up. This is personalizing the debate and criticizing the individual instead of the idea.

And a bit before that you say:

Quote:

Originally Posted by ursamtl

As for some of your other statements, let's just say you obviously don't understand how stereophonic sound is produced and propagated if can't understand Bob Katz's point about bit depth and sampling rate affecting stereo separation. Are you saying that a widely known and respected audio professional is wrong and you're right?

Now isn't that "personalizing the debate and criticizing the individual" as you said before? Aren't you implying that I am stupid because I dare to question world renowned Bob Katz's statements? You see, I am the type of person with an inquiring mind and I don't take anything for granted. Bob Katz's article has some valid points (about mixing and about analog being better than digital) but then again has some gibberish like that one about stereo separation and IMO it also carries too much hype and exaggeration but I will write that off to his fame.

For your info, I have listened to the first quadrophonic records at my fathers house long, long time ago. He also happened to be the tone engineer at the state radio so I happen to know a lot about sound and recording and what is more important I know how to tell the myth from the truth.

Quote:

Originally Posted by ursamtl

If you check Elliot's site, you'll find he is trying to sell a product. So is digit-life. It's called advertising space!

Elliot offers free schematics and PCBs (and a few of them very good ones by the way). digit-life does a lot of critical and scientific based reviews. You should know what type of "selling" I had in mind when I mentioned it. You know, like writing a review that boasts something to the skies and has the advert on the same page so you can immediately rush to buy it while still under impression of an "unbiased review" which has product specs blatantly copied from the specification written on the box?

Quote:

Originally Posted by ursamtl

Yes, the ESI Juli is an interesting card, but it only exceeds the A2ZS specs by any significant degree when one uses the SPDIF outputs.

ESI Juli has measured 103.9 dB dynamic range in 24/96 mode, right? Isn't that better when compared to 97.2 dB measured in 24/96 for Audigy 2 ZS? And that is for analog, not SPDIF mode. Oh I forgot, you trust advertised 108 dB for A2ZS (you said it!), ESI Juli has less than that so it is actually poorer! Also, -90 dB channel separation of A2ZS in 24/96 (the one you can lose only by using lower bit depth because so says famous Bob Katz!) is actually better than ESI Juli -101 dB, right? Come on, give me a break! And don't forget to check THD and IMD figures too.

Yes, it does lack 5.1 and perhaps some other features of A2ZS but it has superior sound quality. Let me remind you what Maxim Liadov from digit-life wrote about it:

"Multimedia cards (for example, Audigy2/ZS) are usually infamous for timbre distortions, various tones and harmonics at high frequencies. So despite their rather high quality converters, such cards cannot be used for professional work. Juli@ is free from these drawbacks, it sounds true and clear. The only difference from Lynx is in a tad lesser detail, which will hardly influence its professional capability, except for the mastering."

Now I must admit that I haven't heard Lynx and Juli yet (although I have heard Creamware Luna recently which is pretty close if not equal to Lynx). Until you claim different I will assume that you haven't heard them either but yet you are stubborn in defending A2ZS and taking it as a reference when it clearly doesn't deserve that.

Don't get me wrong, when I bought A2ZS back in November 2003 it was superior to the majority of onboard sound "solutions" and pro and semi-pro cards like Juli were way out of reach because of their price. A lot happened since then. Soundstorm got ASIO drivers. Then Intel pushed for integrated 7.1 24-bit HD-Audio codecs. Now there is also ASIO4ALL and although not "the real thing" it is still very close. If I haven't already shelled out 120 EUR for A2ZS Retail back in 2003 I doubt that I would do it now -- I would either use this integrated ALC882 that I have on my Asus P5LD2 mainboard or get Juli for a bit more money.

adi111p · Aug 9, 2005, 08:37 AM

I see interesting debates here

, so let me clear up some things because I see many mistakes in your knowledge.

I saw some posts in which you think that if 2 cards have the same codecs both must have the same SNR (106vs 108 dB). Total mistake and not true. If you had some electronics knowledge you would see how absurd are these statements. The most critical thing in audio electronics (if you have the needed parts) is PCB routing, and then passive components thermal noise. MAudio Revo has the PCB with special routing to maximize SNR (I read that in their documentation somewhere). Also if you see cards with the same PCB routing that doesn't mean thay have the same PCB routing, they might have more than 2 layers (which are the visible layers), they might have up to 10 layers, from which 8 layers you can't see...

Also everybody wants better SNR, IMD, THD and dinamics, but what about your audio sistem? Isn't there some limitations? what if you have an amplifier with only 100SNR, would the difference between -106dB soundcard and a -116dB soundcard be observed? I don't think so ... But even if the amplifier is capable of allowing you to see the difference, but what about your speakers? Only high quality speakers are capable of delivering a large amount of dinamics, but would it be enough for you to sense the difference?

Also ASIO4ALL converts the signal to the best possible format (which can be 48khz/16bit unless you change that in your KX ASIO Contol panel) so have you first setup correctly kx ASIO before using ASIO4ALL? Or have you thought that ASIO4ALL will do that? well, it doesn't.

Apart from that, a lot of people were asking me how to get 96KHz output with creative's drivers/kx... Well, I allways check with a spectrum analizer to see if I really have that desired output, and not changed into DSP's format (by making a loopback and recording to 24/96)

Also nobody checked Audigy2 soundcards performance in 24/96 in ASIO mode (i mean THD/IMD/SNR performance). A good thing to know about ASIO is that it might really connect your audio stream to the output DACs without processing because of the low latency required, and I see a big difference in sound between ASIO 24/96 and DirectSound24/96, so I don't think anybody tested A2 at it's full quality potential

ursamtl · Aug 9, 2005, 01:59 PM

adi111p,

You make a valid point about the PCB routing. I doub this is something you'll see much in marketing documents because it's not something that's short, simple and easy to express in a couple of words. This being said, if the test conditions are equal then the effect of thermal noise should be picked up by the test and be reflected in the test results. In other words, if both cards are tested in the same computer system that has been allowed to run for a certain amount of time, then a comparison of test results should be valid. There is nothing absurd about that!

I completely agree with you about the effect of amplifiers and speakers on the total performance of a system. It is fine to take some theoretical mathematical absolute such as 24-bit sound should exhibit a dynamic range of 144.5dB. If the system cannot reproduce anything near that, it's a useless measurement that means virtually nothing in the real world.

Now I'm afraid you've completely lost me with your comment about ASIO4ALL. Do you mean that ASIOALL is not in fact a separate WDM driver that emulates ASIO? You seem to be saying that it sits on top of whatever driver is installed and only matches the bit depth and sampling rate available from the underlying driver.

Finally, you make an interesting point about no one testing the A2's ASIO output. It's well known on here (thanks to Eugene) that the A2 DSP chip is only 16 bit so any date passed through the DSP will come out with a bit depth of 16. To keep the bit depth at 24 the data must not be passed through the DSP. I'm only a begineer at C++ programming so I don't yet have the ability to dig into the ASIO SDK or any driver code to see what's going on. What you're saying does sound quite logical. I'd be interested in seeing some test results.

igor_levicki · Aug 9, 2005, 10:43 PM

@adi111p:
I was aware of PCB layout and component quality/thermal noise issues simply because I am into DIY myself. Anyway, good point, well worth of adding into this thread.

About ASIO4ALL, I believe it is standalone ASIO driver which provides ASIO functionality by interfacing to whatever WDM drivers for whatever sound card you may have installed.

Anyway, my point about ASIO4ALL was that using it with integrated HD-Audio enables you to do some pretty nice stuff and to almost match A2ZS in sound quality without losing a single cent over the cost of your mainboard.

Aug 9, 2005, 10:43 PM	#34
igor_levicki Intel Black Belt Join Date: Dec 2003 Location: Belgrade, Serbia Posts: 128 Rep Power: 0 System Specs	@adi111p: I was aware of PCB layout and component quality/thermal noise issues simply because I am into DIY myself. Anyway, good point, well worth of adding into this thread. About ASIO4ALL, I believe it is standalone ASIO driver which provides ASIO functionality by interfacing to whatever WDM drivers for whatever sound card you may have installed. Anyway, my point about ASIO4ALL was that using it with integrated HD-Audio enables you to do some pretty nice stuff and to almost match A2ZS in sound quality without losing a single cent over the cost of your mainboard. __________________ Regards, Igor Levicki -- Creative Audigy 2 ZS Retail (SB0350) kX 3538h

Aug 9, 2005, 08:37 AM	#32
adi111p Sound Quality Freak Join Date: Jun 2004 Posts: 51 Rep Power: 0	I see interesting debates here , so let me clear up some things because I see many mistakes in your knowledge. I saw some posts in which you think that if 2 cards have the same codecs both must have the same SNR (106vs 108 dB). Total mistake and not true. If you had some electronics knowledge you would see how absurd are these statements. The most critical thing in audio electronics (if you have the needed parts) is PCB routing, and then passive components thermal noise. MAudio Revo has the PCB with special routing to maximize SNR (I read that in their documentation somewhere). Also if you see cards with the same PCB routing that doesn't mean thay have the same PCB routing, they might have more than 2 layers (which are the visible layers), they might have up to 10 layers, from which 8 layers you can't see... Also everybody wants better SNR, IMD, THD and dinamics, but what about your audio sistem? Isn't there some limitations? what if you have an amplifier with only 100SNR, would the difference between -106dB soundcard and a -116dB soundcard be observed? I don't think so ... But even if the amplifier is capable of allowing you to see the difference, but what about your speakers? Only high quality speakers are capable of delivering a large amount of dinamics, but would it be enough for you to sense the difference? Also ASIO4ALL converts the signal to the best possible format (which can be 48khz/16bit unless you change that in your KX ASIO Contol panel) so have you first setup correctly kx ASIO before using ASIO4ALL? Or have you thought that ASIO4ALL will do that? well, it doesn't. Apart from that, a lot of people were asking me how to get 96KHz output with creative's drivers/kx... Well, I allways check with a spectrum analizer to see if I really have that desired output, and not changed into DSP's format (by making a loopback and recording to 24/96) Also nobody checked Audigy2 soundcards performance in 24/96 in ASIO mode (i mean THD/IMD/SNR performance). A good thing to know about ASIO is that it might really connect your audio stream to the output DACs without processing because of the low latency required, and I see a big difference in sound between ASIO 24/96 and DirectSound24/96, so I don't think anybody tested A2 at it's full quality potential

Aug 9, 2005, 01:59 PM	#33
ursamtl Montrealer Join Date: Apr 2004 Posts: 126 Rep Power: 0	adi111p, You make a valid point about the PCB routing. I doub this is something you'll see much in marketing documents because it's not something that's short, simple and easy to express in a couple of words. This being said, if the test conditions are equal then the effect of thermal noise should be picked up by the test and be reflected in the test results. In other words, if both cards are tested in the same computer system that has been allowed to run for a certain amount of time, then a comparison of test results should be valid. There is nothing absurd about that! I completely agree with you about the effect of amplifiers and speakers on the total performance of a system. It is fine to take some theoretical mathematical absolute such as 24-bit sound should exhibit a dynamic range of 144.5dB. If the system cannot reproduce anything near that, it's a useless measurement that means virtually nothing in the real world. Now I'm afraid you've completely lost me with your comment about ASIO4ALL. Do you mean that ASIOALL is not in fact a separate WDM driver that emulates ASIO? You seem to be saying that it sits on top of whatever driver is installed and only matches the bit depth and sampling rate available from the underlying driver. Finally, you make an interesting point about no one testing the A2's ASIO output. It's well known on here (thanks to Eugene) that the A2 DSP chip is only 16 bit so any date passed through the DSP will come out with a bit depth of 16. To keep the bit depth at 24 the data must not be passed through the DSP. I'm only a begineer at C++ programming so I don't yet have the ability to dig into the ASIO SDK or any driver code to see what's going on. What you're saying does sound quite logical. I'd be interested in seeing some test results.