Thread: audio over USB v. optical

PCM lacks clocking? Did he go into detail? Without digging to far into it, PCM or Pulse Code Modulation has to be clocked somehow. Without PCM every single CD would be useless.

http://en.wikipedia.org/wiki/Pulse-code_modulation

Found a few links discussing improvements to the CD. They may or not be relevant, but if you look at some of Werewolf's posts, he gives light to what maybe to look into.

http://forum.elitecaraudio.com/showt...hreadid=126052
http://forum.elitecaraudio.com/showt...hreadid=126082

I think one thing you have to remember is that people who spend thousands on DACs are going to have a bias against computers and less inexpensive DACs because they are under the impression cheap = poor quality and $$$ = good quality.

Links to reviews comparisons?

The past 2 issues in Stereophile magazine have had reviews of USB DACs.

Stereophile is generally considered unbiased as far as I've seen.

Nov. 2007 issue reviews Bel Canteo e.ONE DAC3
Jan. 2008 issue reviews Benchmark DAC1

Both by John Atkinson

I don't think its a fair assumption that those who are willing to spend a lot of money will automatically have a bias against lesser expensive products. I've read a number of reviews in the past that point to quite the opposite.

Some people claim that there is no discernible difference between running audio through the k-mixer in Windows and running bit-perfect. Do they claim this simply because they don't have the ear to be able to discern the difference? Do they claim this simply because they don't understand what's going on?

Well it's good that they give unbaised opinions sometimes.

Did they take measurements or was it purely subjective? What didn't they like about them specifically? Were they listening to them via headphones, speakers, etc.

Edit: http://forum.stereophile.com/forum/s...34&an=0&page=0

Originally Posted by durwood

PCM lacks clocking? Did he go into detail? Without digging to far into it, PCM or Pulse Code Modulation has to be clocked somehow. Without PCM every single CD would be useless.

http://en.wikipedia.org/wiki/Pulse-code_modulation

Originally Posted by linked Wiki article

In real systems, such a device is commonly implemented on a single integrated circuit that lacks only the clock necessary for sampling, and is generally referred to as an ADC (Analog-to-Digital converter)

From what I'm gathering PCM is the data its self.

Sitting here talking to Monte the best comparison I can come up with is this:

The clock is what actually makes PCM useful. Much like having lines of code that are uncompiled. Just because the lines of code are uncompiled doesn't mean that they are not lines of code, it just means that they don't have a useful meaning just yet. In other words, PCM data without a clock, is just PCM data...it's data that has not yet been utilized.

If the PCM protocol were to include clocking, then wouldn't it have to specify the clock timing as well? Then wouldn't that narrow down the purpose of PCM and it's protocol and make it entirely too specific for general use? Perhaps I'm getting way too ahead of myself...lol

Another comparison...Cat5 cable is designated Cat5 by a standard (protocol). A piece of Cat5 cable cannot be designated as Cat5 without being able to transfer at least X amount of data a second. So in order to specify clocking in with the PCM protocol, wouldn't you really need to specify the clock speed as well for it to be a legitimate protocol?

Originally Posted by durwood

Found a few links discussing improvements to the CD. They may or not be relevant, but if you look at some of Werewolf's posts, he gives light to what maybe to look into.

http://forum.elitecaraudio.com/showt...hreadid=126052
http://forum.elitecaraudio.com/showt...hreadid=126082

I'm reading through these now! Probably going to take me a bit though...hehe

btw - here's the thread I have going on Stereophile: http://forum.stereophile.com/forum/s...ge=0#Post32327

I like this stuff.. I have been wanting to get back into new stuff after a few month break..

I was always thinking that PCM was the type of data, as you say above, while SPDIF is the means of transferring the PCM data between devices. I guess something like the SPDIF is the pipeline and the PCM is the water going through the pipe.

The SPDIF has a clock embedded(Dunno if that is the right term) so that is where the DAC knows how to recover the clock the same as the device sending the data.


As for an actual spinning CD transport, If I Understand, much of the quality of the transport is in the circuit design and the quality of the oscillator that is doing the clocking prior to sending through the SPDIF.. I would think a computer cd drive has a pretty simple/cheap oscillator, making it not the best transport.

If you are not familiar with Computer Audio Asylum, here it is

a whole lot of that site is over my head, but if i read enough, i usually get the point..

Originally Posted by Felix509

I would think a computer cd drive has a pretty simple/cheap oscillator, making it not the best transport.

Actually, according to AlexO:

Originally Posted by AlexO

If you're connecting the drive to the computer, there isn't much to be gained from various drives or even interfaces. Some drives and interfaces are faster than others, but they're all way, way more than adequate for audio purposes. Furthermore, since the transfer from the drive will be sent as data encapsulation, at data speeds and there will be considerable buffering on the hard drive to address any error issues, the role of the CD ROM is greatly diminished.

So, unless you're connecting the CD ROM straight to the DAC, bypassing the computer, for our purposes here, the quality of the drive is irrelevant because all of them are beyond specs needed for audio streaming TO THE COMPUTER.

Hope this helps.

Originally Posted by Felix509

If you are not familiar with Computer Audio Asylum, here it is

a whole lot of that site is over my head, but if i read enough, i usually get the point..

I'll head over and check it out! Thanks for the link.

I'm like you, I can dive head first into this stuff, but I can only stay under water for so long before I must come back up for air.

I have found that each time I dive, I can stay under longer though...lol

Yep, that makes good sense if you are going into your computer from the CD drive.. I dunno why i was thinking you may be going directly out of the CD to the processor..

USB streaming also streams using PCM. One of the problems with USB audio is that it has no clock embedded with the PCM data. That also supports that PCM has no clock embedded.

I am searching now for the best card to output my SPDIF from, as that seems to make a big difference, but am relying on the uber experts on the Computer Audio Asylum to guide me.. <---Make sure you have Loads of time before undertaking this forum, it is full of Info and links to more info etc..... Some real experts on computer audio on here, the guy who makes Benchmark Audio, The guy who is Empirical Audio, among others..

I simply cannot stand the way that forum is set-up.

I've told it to display like vBulletin, but it's not doing so. No wonder they call it the asylum....:|

I know its been a while, but if anyone is interested there are a couple issues preventing this from happening..Not impossible, just more difficult than expected.

For multiple channels @ 24/ 96, the only chips capable of processing that amount of data are the TUSB3200A and the TAS1020B.

These chips, due to USB 1.1 being full speed at 12Mb/s, can theoretically handle 4 channels or 2 stereo pairs a piece. For the highest quality possible, asynchronous mode needs to be employed because the clock would then be on the DAC itself and timing would not depend on the computer. The DAC would tell the computer when to send the data. This is better than most other soundcards or DACs, which use adaptive mode for clocking. So that is where the problem comes in. Firmware needs to be written for these chips to work correctly, which is very complicated. The guys over at CEntrance have written firmware for the TAS1020B that has the potential to run 3 channels of 24/ 96 per chip. Therefore, unless we can find anyone to write firmware to support 4 channels per chip, we are stuck with three channels of bit perfect audio. Which is not all that bad because if two of the chips were used, we could have 6 channels with 4.5 Vrms Analogue out. However if we used two chips, we would need two usb ports available (not a usb hub, but two ports straight from the motherboard to handle the data). I am trying to get a license for the code to be able to use it in the product. Ill see if they cant squeeze one more channel (mono) for a sub. I also spoke with Jan at Thuneau- Frequency Allocator Software that processes real time audio (Time alignment, Xover, Eq, phase allocation). She said she would license the software and make a custom version for the soundcard so it will sit neatly in the background. But for now the big obstacle is getting the firmware either from CEntrance or writing our own to handle idealy 4 channels of 24/ 96 per chip.
Ill keep you updated about the status of the license. Would you still be interested if we can only get 6 channels possibly 7?

Thanks,

Ryan S.