[MaloS]

On the contrary, knowing that logic behavior is only logic on abstract level and shares alot of properties with analog behavior in reality, hence USB cables can make a difference. I am not going to claim that there is a drastic difference between $20 and $200 USB cable (most likely minimal), but the possibility of an overly resistive USB cable reducing volume is very very easily to explain. Cheers.

 

[Dept_of_Alchemy]

I thought USB data is buffered on both ends... either way, as far as I know, the understanding on Head-Fi has been that USB cables do not make a difference (unless if it's broken, obviously).

 

[error401]

Quote:

Originally Posted by MaloS /img/forum/go_quote.gif the possibility of an overly resistive USB cable reducing volume is very very easily to explain. Cheers.

Only if the device is powered by the USB bus, which should be obvious. For the volume to be reduced in a consistent manner, the effect on the data stream would have to be deliberate. Not affecting control frames, doing some binary math (don't forget that 'zero' is 0.5 2^n) to reduce the value of the sample proportionally. This is really non-trivial, and you couldn't do it without a lot of effort into capturing the signal, passing USB control frames directly, pulling the data out of the audio frames and modifying it, then restuffing and recalculating checksums. In short, you'd have a full-fledged USB transceiver.

I don't totally discount the possibility that a USB cable can have a real effect on the output of a connected device - and it's definitely possible if the device is powered by USB - but if effects in the digital domain were present, they would be timing related and likely so insignificant they'd not be audible or measurable or so drastic they would be obvious.

Digital communications is not as trivial as 'a bunch of zeroes and ones'. The real world is not discrete, and physics interferes. I'd be willing to bet that not one of the ultra fast, very low voltage signals flying around your motherboard right now look anything like a square wave on the scope. These analog effects can intrude on the operation of the circuit in ways that won't break it by affecting its timing.

Of course, I don't think it applies here at all. Buy a cable that's to spec and it will work as well as the $200 solid silver one will. In fact, it'll probably work better since that fancy cable is likely not to spec.

 

[mojo]

Quote:

Originally Posted by MaloS /img/forum/go_quote.gif but the possibility of an overly resistive USB cable reducing volume is very very easily to explain. Cheers.

It isn't. Transmitted signal levels are 0.0–0.3 volts for low and 2.8–3.6 volts for high. Resistance will not be enough to cause signal level errors in almost any cable.

Perhaps resistance could lower the USB power output slightly, but we are talking a few ohms here on a 500mA max draw.

Most likely, people are plugging two USB cables into different ports and then switching between the two. Due to the way USB works in Windows, this can result in two different sets of settings, one for each port/cable.

 

[Earlephant]

Quote:

Originally Posted by mojo /img/forum/go_quote.gif That isn't the important bit though, the interesting part is about jitter. No! I refuse to disagree! Seriously, I think we are actually in agreement on the main issue - that expensive gold plated USB cables are a waste of money.

Are we really talking about timing errors or dropped frames / packets that aren't re-sent? I presumed that timing information would have been in the encoded file?

 

[maarek99]

The volume can't be different with a different cable. It just can't. Like the earth just doesn't rotate the other way around suddenly. I call psychoacoustics on this one.

There's no difference.

 

[mojo]

Quote:

Originally Posted by Earlephant /img/forum/go_quote.gif Are we really talking about timing errors or dropped frames / packets that aren't re-sent? I presumed that timing information would have been in the encoded file?

What you have to remember is that USB is not buffered! It is more like SP/DIF - the audio is digially encoded but is decoded by the DAC in realtime. There is no buffering of samples, no timing information. The DAC decodes each sample as it arrives, with possibly a bit of re-clocking to try and reduce jitter.

There are no re-sends, as although there is error detection there is no method for error correction. All it does is allow the DAC to avoid outputting a bad sample, perhaps by repeating the last sample or just outputting silence. I bet most DACs don't even do that though, they just output the error.

The file is not sent in encoded form. The PC does all the encoding, all it sends is decoded samples (words).

This is why USB audio is not the holy grail it could be. If there was buffering, jitter would be very much reduced and easier to handle.

 

[Earlephant]

Quote:

Originally Posted by mojo /img/forum/go_quote.gif What you have to remember is that USB is not buffered! It is more like SP/DIF - the audio is digially encoded but is decoded by the DAC in realtime. There is no buffering of samples, no timing information. The DAC decodes each sample as it arrives, with possibly a bit of re-clocking to try and reduce jitter. There are no re-sends, as although there is error detection there is no method for error correction. All it does is allow the DAC to avoid outputting a bad sample, perhaps by repeating the last sample or just outputting silence. I bet most DACs don't even do that though, they just output the error. The file is not sent in encoded form. The PC does all the encoding, all it sends is decoded samples (words). This is why USB audio is not the holy grail it could be. If there was buffering, jitter would be very much reduced and easier to handle.

OK found this article interesting: http://www.positive-feedback.com/Issue22/nugent.htm

 

[Rambaud]

Anybody tried these at £33.00 for 1 metre?

http://tinyurl.com/yo8wo3