[CDewey]

Will it make a difference which digital connection I send to my DAC from my computer? I was going to use USB and leave the two optical inputs available for other things. Is there any reason I should not do that?

Looking to go ALAC > Peachtree Nova > SENHD580. I plan on using the built in headphone amp for awhile on the Nova. Looking to upgrade to another amp in the next few months.

 

[FallenAngel]

Yes, completely different standards.

 

[CDewey]

So I should use optical out of my computer instead? I can do both from my PC.

 

[Currawong]

Try both, see if there is any difference. Whether there will be a difference or not depends on the implementation of them in each device.

 

[rosgr63]

With USB you might experience latency problems.
As Currawong said try and see.

 

[Zeroblade]

It seems USB usually works with bitrates only up to 16bit/44.1KHz, while digital (either coax or optical) goes up to 24bit/192KHz. CD-quality recordings only go up to 16bit/44.1KHz though, so I suppose it doesn't make much of a difference. Still if you can go with digital, why not?

 

[Pepsi]

Mostly depends on your source. If your media library mainly consists of mp3 to CDrip flacs, i'd just go wth usb. Go digital if you plan on going to 24bit.

 

[Zeroblade]

Is there really any commonly found media format that has 24bit? Apparently DVD-audio has it, but I pretty much never see it on the store shelves.

 

[freakydrew]

if I may expand on this thread...if optical is the better choice for the o/p, what about coax for the same purpose? any difference in overall sound between coax and optical? I would think optical would be the best because it travels on light whereas everything else travels via a cable of some sort?

 

[Prog Rock Man]

Whilst the connections and cables have various differences to them, their role is to transfer digital data. So long as that data is error free and the DAC chip is able to process that data without any errors or jitter (which is apparently not an issue with USB) then all three should sound the same.

 

[Zeroblade]

I've read elsewhere that coax carries not just the digital info, but timing information as well. No idea if this is true or otherwise, however.

 

[etiolate]

Quote:

Originally Posted by Zeroblade /img/forum/go_quote.gif I've read elsewhere that coax carries not just the digital info, but timing information as well. No idea if this is true or otherwise, however.

well, as far as i know with my knowledge of digital transfer, the timing is kind of built into the data transfer. as the data is transferred, it should arrive in a specific time-regimented order. it is part of being in the digital domain.

 

[linuxworks]

Quote:

Originally Posted by freakydrew /img/forum/go_quote.gif if I may expand on this thread...if optical is the better choice for the o/p, what about coax for the same purpose? any difference in overall sound between coax and optical? I would think optical would be the best because it travels on light whereas everything else travels via a cable of some sort?

short answer: there is no diff between physical transports of opto or coax.

both are exactly and completely the same once they reach the spdif receiver chip. I build spdif circuits and know this field well. there is NO DIFF between coax and opto. anyone who tells you otherwise does not know the actual theory or has never built anything using such components.

pick the one that is easy or compatible with your gear and relax about the so-called cable experts.

 

[linuxworks]

Quote:

Originally Posted by etiolate /img/forum/go_quote.gif well, as far as i know with my knowledge of digital transfer, the timing is kind of built into the data transfer. as the data is transferred, it should arrive in a specific time-regimented order. it is part of being in the digital domain.

what they mean is that the clock (timing) and data (real data) are intermixed but not in the way you might think. there's no 'time code' or absolute number sequence like smpte would have, for example.

the timing is *implied* by the clocking of bits (0s and 1s). each time a bit time is 'found' the 0 or 1 is sent out the serial connection and into a (conceptual, sometimes even real) shift register where the 16 bits are collected into a sample. after the right amount of delay after the last actual bit is received and clocked in, THEN a left or right channel value is 'ejected' to the analog section.

even with usb audio you still have to 'know' when to eject the samples and usb audio still does not include absolute time numbers. basically, as soon as the 16 bit samples are 'dropped' into the usb buffer, it still has to clock the 16 bits into an analog value and 'place' that on the analog wire at the right time. there is no extra help in usb-audio compared to spdif audio; both can have jitter and neither is any better than the other. the jitter depends on the local clock (if there is one; and there is, in asynch usb) and not so much on the source clock (inside the sender's end of things).

 

[linuxworks]

Quote:

Originally Posted by Zeroblade /img/forum/go_quote.gif It seems USB usually works with bitrates only up to 16bit/44.1KHz, while digital (either coax or optical) goes up to 24bit/192KHz.

usb IS digital. of course!

and usb audio can run in a few modes; a standard usb-audio mode (usb 1.x) which is very standard but also is limited, as you say, to 16bit/48k as the max speed of data you can send thru (usb 1.1 limit).

some devices DO have usb2 support (0404usb emu, for one; dac1 is another that comes to mind). those do not support 'standard' usb in the same way that consumer simple dongle devices do. I know my 0404usb box can run at 24/96k just fine over usb2.0 and even a low-end laptop.