[arnyk]

  That BER spec sounds like it's for a standard USB link with error correction, rather than for USB audio. Are you sure that's typical for non-ECC usb audio links?
 
(Admittedly, I wouldn't expect the BER on a USB link to be bad enough to make any audible difference, but still, in the interest of pedantry and correctness...)

 
Interesting. Of course the performance of just one old junker DAC means nothing.
 
It appears that the in-use BER of an asynch USB DAC is zero because errors are detected and retried.
 
I ordered a USB BER tester, and we shall see what we shall see.

 

[castleofargh]

  Interesting. Of course the performance of just one old junker DAC means nothing.
 
It appears that the in-use BER of an asynch USB DAC is zero because errors are detected and retried.
 
I ordered a USB BER tester, and we shall see what we shall see.

 

[jcx]

RedBook has substantial error correction - but that's for the media - digital links can regularly achieve 10^-12 BER - GHz computers wouldn't work too well if they didn't
 
(The CIRC code used on audio CDs can correct burst errors of up to 3500 bits (2.4mm) and can interpolate error bursts of up to 12,000 bits http://www.usna.edu/Users/math/wdj/_files/documents/reed-sol.htm )
 
 
on jitter a point often missed is that it is a multiplicative/modulation error - if you have low signal amplitude (actually low slew rate for jitter) then you have lower modulation products - so tales jitter of spoiling bass or reverb tails means the poster doesn't know what they're talking about

 

[GreenBow]

 
The Meridian Explorer you are using is working in asynchronous mode, that means it has its own clock. So cable jitter can and does not apply here.

Yeah I know. I wrote that in post 226. However that only applies to the signal being given a more accurate clock to work on. Not cable jitter which apparently the cable itself can cause.
 
Plus there are apparently other faults in correct data that can come from poor quality cables, like interference. It depends entirely on the error rate of these, and some people on this thread are claiming this is very low. I keep searching online for it. However the main article that keeps popping up is from a website that is selling cables. A little biased maybe. 

 

[interpolate]

In a high-precision workstations often they will be a wordclock device to stabilise the drift in timing caused by jitter. At the same time, better circuit paths and improved DAC processors will alleviate the need for this in the future. Meaning only the need for a one quality device with built-in clock management handling all processes will make this cheaper and more effective to do.
 
It's a rather petty thing worrying about picosecond jitter errors when it gets to the end-user stage.

 

[castleofargh]

  In a high-precision workstations often they will be a wordclock device to stabilise the drift in timing caused by jitter. At the same time, better circuit paths and improved DAC processors will alleviate the need for this in the future. Meaning only the need for a one quality device with built-in clock management handling all processes will make this cheaper and more effective to do.
 
It's a rather petty thing worrying about picosecond jitter errors when it gets to the end-user stage.

the less they understand, the more people will worry. unknowns are always more worrisome than perfectly know problems(it's irrational but human). look how people are fine with a tube amp creating massive distortions, FR changes, and phase shifts. and how they get scared ****less about the very same thing in an EQ with mostly values several magnitude smaller. ^_^ 
jitter -110db down under the music is the same, it's irrational because they don't have a clue what they're looking at or how it actualy affects sound.

 

[Steve Eddy]

the less they understand, the more people will worry.

And the more garbage you can feed them that they'll swallow without question.

se

 

[interpolate]

I don't worry and just because I don't have Genelec MIdfields in a Swedish Custom Studio next to the Fjords doesn't mean anything I listen too will necessarily sound bad.

 

[arnyk]

  In a high-precision workstations often they will be a wordclock device to stabilise the drift in timing caused by jitter.
 
At the same time, better circuit paths and improved DAC processors will alleviate the need for this in the future. Meaning only the need for a one quality device with built-in clock management handling all processes will make this cheaper and more effective to do.
 
It's a rather petty thing worrying about picosecond jitter errors when it gets to the end-user stage.

 
Drift in timing and jitter are two different things. Problems with timing accuracy is due to a very long term frequency inaccuracy, and while jitter is due to a very short term frequency inaccuracy.
 
They are specified separately and controlled by different parts of the clock circuitry.
 
Timing accuracy is not a problem unless the workstation has a number of stand-alone peripherals that have independent clocks that need to be synchronized with each other.

 

[Bibo]

No, asynchronous USB is not only about the better clock, it also means that there is a data buffer in the dac. The data that is sent into dac chip is read from the buffer, not from the USB bus. So any jitter that is in the signal that feeds the buffer is eliminated.

Edit: I replied to Greenbow

 

[interpolate]

I was referring to multiple devices being used in tandem and all synchronised by a wordclock. Such as a protools and adat system.

This is the environment i want to get into.

 

[GreenBow]

No, asynchronous USB is not only about the better clock, it also means that there is a data buffer in the dac. The data that is sent into dac chip is read from the buffer, not from the USB bus. So any jitter that is in the signal that feeds the buffer is eliminated.

Edit: I replied to Greenbow

I don't if know the Meridian Explorer DAC re-clocks.
 
(It's probably the case that the ME clocks are pretty good though. Just thinking out loud there. It's timing does sound a little off occasionally though. Not as bad as the PC on-board DAC.) You never knew what was coming next with that DAC, even on music you had heard lots before. It was actually exciting.)
 
@Bibo, it might  be worth looking up HPET. One of the game developers on Steam suggested me use it when I was getting audio issues in game. I was slightly over-clocking the CPU using a software app. The game was not aware and causing audio dropout skipping. Enabling HPEY in the BIOS and doing a command at Crompt fixed it.  
 
This is out of my field though so I am only speculating for you. I am however interested in ProTools myself. In my undergraduate I used ProTools and loved it.
 
At the moment I am looking at buying Cubase or ProTools. Bot have a demo version. The Cubase one is online to download for thirty days. The ProTools one was on the way, last time I looked, and you sign up for it.
 
Getting a good ADC is a priority too for me. The M-Box that come with ProTools is probably where I will go. Or I might need to get an entirely independent one to work with Cubase. At least with Cubase having a current demo I can work out whether it will do what I need it to. The last copy of it I had would not, and I got frustrated and sold it eventually; was gutted.
 

 

[cjl]

  (It's probably the case that the ME clocks are pretty good though. Just thinking out loud there. It's timing does sound a little off occasionally though. Not as bad as the PC on-board DAC.) You never knew what was coming next with that DAC, even on music you had heard lots before. It was actually exciting.)

 

There's absolutely no chance that any DAC that functions at all could possibly change the timing of the actual music in a perceptible way.

 

[arnyk]

No, asynchronous USB is not only about the better clock, it also means that there is a data buffer in the dac. The data that is sent into dac chip is read from the buffer, not from the USB bus. So any jitter that is in the signal that feeds the buffer is eliminated.

Edit: I replied to Greenbow

 
There is only one pair of data wires which is used for both data and commands in both directions in USB-1 or USB-2 so that means that all data that is sent into the DAC chip comes from the one and only USB bus. Yes, the data goes from the bus into a buffer and then on to the DAC chip.
 
This is true for all three relevant USB modes. 
 
(1) In the Isosynchronous Adaptive or simply Isosynchronous mode , the audio clock is derived from the data clock in such a way as to avoid data over runs or under runs. While it is so close to being correct that there is no audible change in timbre or pitch, it is strictly speaking an approximation.
 
(2) In the Asynchronous Adaptive mode the audio clock(s) s for input and output are derived separately by a similar means as above.
 
(3) In the Asynchronous protocol, the audio clock(s) are local to the I/O device and can be as exact as is desired. Data flow over the bus is managed so that there are no overruns or underruns.

 

[GreenBow]

  There's absolutely no chance that any DAC that functions at all could possibly change the timing of the actual music in a perceptible way.

Sorry but that's incorrect.