[DatHak512]

So i'm reading a lot and learning a lot about jitter. From what i understand so far, digital jitter is fairly prevelant in digital coax and digital optical transmission methods, but i don't know, or haven't read anything about jitter regarding USB transmission methods. So can you tell me all you know about jitter, and USB jitter?

The reason i'm asking is because i have a USB amp with a built-in DAC, and in light of all my recent research, i'm wondering how much of an effect digital jitter is having on my sound. Thanks!

 

[tschanrm]

Read this thread:

http://www4.head-fi.org/forums/showthread.php?t=134969

Should give you enough info to help you. Basically, USB can have lower jitter (as low as jitter related to the clock even) than coax or optical s/pdif, but you need the right chip. Hope this helps

 

[nopietns]

tschanrm, this -> http://www.head-fi.org/forums/f9/her...o-cool-134969/ ?


Oh, just read through http://www.head-fi.org/forums/f46/ba...-audio-247637/

But it's kind of hard to accept that jitter is *that* irrevelant.

 

[elrod-tom]

My thoughts on this are anectdotal...

I have a balanced HeadRoom Desktop with onboard DAC. I can notice a difference using a PC and either a Linksys wireless G music bridge or M-Audio transit as a souce for the optical feed between the USB and optical ins (haven't really tried the coax yet)...and the optical wins the day. I've heard similar comments from others with different DACs.

Is there anything to this? Who knows...

 

[DefectiveAudioComponent]

Quote:

Originally Posted by tschanrm /img/forum/go_quote.gif Basically, USB can have lower jitter

USB can be completely without jitter.

 

[gyrodec]

Nothing, including USB can be without jitter. This is impossible - it would violate the uncertency principle. It can have very, very low jitter, but only as low as the clock it is synchronized to. And, as all clocks have jitter, and must have by the laws of physics, it will not no zero. A couple of pico-seconds or so is possible, but no better.

 

[DefectiveAudioComponent]

If the device transfers the data in bulk mode (asynchronous, thus not synced...) and then re-clocks it, then the cable jitter does not matter. Right? How could it?

 

[wavoman]

Quote:

Originally Posted by DefectiveAudioComponent /img/forum/go_quote.gif If the device transfers the data in bulk mode (asynchronous, thus not synced...) and then re-clocks it, then the cable jitter does not matter. Right? How could it?

I think we have four separate threads all active on USB quality at the moment ... see links I show below.

Anyway, to answer the quoted question:

In theory, the USB cable could matter this way: the receiving firmware still has to operate in hard real-time, since we are playing music here, not transfering files. If the USB cable were badly enough made then the speed of transmission (even a perfect bit-for-bit stream) could be so slow that the receiver misses edge transitions and fails to re-construct the bitstream correctly. This has been discussed in other threads, and is explicitly mentioned as a possibility on usb.org, especially in the case when ferrite beads are used on a USB 2.0 cable. But I would call this "cable quality", not "cable jitter", since the cable has no clock. (Other posters in other threads have said the exact same thing).

If the transmitting USB firmware has horrible jitter (which it won't) then also in theory the receiving side could be foiled.

But with a decent PC audio computer, and a modern USB DAC, this ain't gonna happen ... anything half-way decent has jitter so low that the effect on audible SQ is nil.

The OP should not worry -- no jitter effect.

As another poster (nick charles) said in a different thread about this very same issue:

So many things is life to worry about, why add another one needlessly...

See especially his post with references:

http://www.head-fi.org/forums/f21/so...ml#post4428173

For more on all this follow the thread:

http://www.head-fi.org/forums/f46/ji...ection-343295/

I have argued in yet another thread that it is worth spending $15 to get a certified USB cable, especially if you are using the freebie they threw in the box, and it has ferrite beads. I would not suggest spending $$ on an audiophile USB cable. I have seen $50 USB cables at shows, with ferrite beads (which usb.org recommends against) ... I believe they are a waste of money.

So when all is said and done about cables making a difference or not... - Page 14 - Head-Fi: Covering Headphones, Earphones and Portable Audio

Other ideas and suggested cables are in the currently active thread:

http://www.head-fi.org/forums/f46/di...cables-237984/

Some other posters think (a) USB 2.0 certification is bull, or (b) all USB cables no matter how cheap pass certificaion. I believe both of these assertions are incorrect, and would point to the developer's faq at usb.org.

But jitter causing you a problem in your set-up? Very very doubtful.

 

[LnxPrgr3]

Quote:

Originally Posted by gyrodec /img/forum/go_quote.gif Nothing, including USB can be without jitter. This is impossible - it would violate the uncertency principle.

Please explain.

 

[gyrodec]

Defective - as a USB cable is useless without the stuff on either end, saying that the USB cable has no jitter is pointless, even if it may be true for bulk mode. The jitter on the clock/circuitry used to reclock it can't be avoided, and so are part of the USB system, hence the USB does have jitter - in the real world.

The reason that the laws of physiscs say you can never have no jitter is because quatum uncertainty, i.e. there is always a bit of randomness about things at the quantum level, means that it is impossible to build a clock without at least some jitter. Even the best clocks we have, cesium atomic clocks, have just alittle bit of litter in the "ticks", and this can never be removed. Hence as all clocks have jitter, and all digital audio is tied to a clock at some point, it is impossible to have anything jitter free in didgtal audio. This is a very esoteric point, but Defective made an absolute statement, and so I just had to make an absolute response.

 

[DefectiveAudioComponent]

Quote:

Originally Posted by gyrodec /img/forum/go_quote.gif Defective - as a USB cable is useless without the stuff on either end, saying that the USB cable has no jitter is pointless, even if it may be true for bulk mode.

It is quite important to know what components in a system that can affect the sound. So I think it is a very good point that the cable does not affect the sound with bulk mode transfers.

 

[LnxPrgr3]

Quote:

Originally Posted by gyrodec /img/forum/go_quote.gif Defective - as a USB cable is useless without the stuff on either end, saying that the USB cable has no jitter is pointless, even if it may be true for bulk mode. The jitter on the clock/circuitry used to reclock it can't be avoided, and so are part of the USB system, hence the USB does have jitter - in the real world.

Yes, but if you aren't using the USB to clock your DAC, the the USB cable can neither hurt nor help your jitter levels.

You know, with an internal sound card, your computer writes to it in blocks in response to interrupts from the card (well, there's other ways to do this as well, but this is the conventional method). However, your operating system's scheduler introduces timing variations in when this audio is actually written. So long as the delay is less than the length of audio left in the sound card's buffer, there is no audible consequence, even if your operating system's timing writing to the card is very jittery.

Put another way -- when you stream music online, packets carrying the sound don't arrive timed perfectly, or even vaguely close. Packets may even take completely different routes to get to you. So long as you don't have to wait so long you deplete your player's buffer, this is completely irrelevant to sound quality. The source, an online stream, is extremely jittery, and yet it doesn't matter.

Quote:

Originally Posted by gyrodec /img/forum/go_quote.gif Even the best clocks we have, cesium atomic clocks, have just alittle bit of litter in the "ticks", and this can never be removed. Hence as all clocks have jitter, and all digital audio is tied to a clock at some point, it is impossible to have anything jitter free in didgtal audio. This is a very esoteric point, but Defective made an absolute statement, and so I just had to make an absolute response.

So we can't hope for absolute perfection, because it will never come. I am glad my hearing doesn't have infinite resolution, and is thus fooled by lesser clocks

We can, however, make the USB cable irrelevant to the DAC's clock, as noted in the three threads covering this topic I can think of off of the top of my head, and probably others as well.