Curious, have you ever heard differences in media players?
Ahhm, less "media players" as such, more like system settings and player settings, such as RAM Cache size.
In "Days of Future Past" AMR "made" in limited quantities a Media server/player based on J-River and an HP Touchscreen plus active 25m USB Cable, so it could be placed at the listening position. Of course, those were the daze of windoze XP IIRC.
These were meant as sales help for dealers who were shall we say less than computer literate. A lot of the setup was manipulating system settings to prioritise audio processes and adjust buffering from HDD from the standard settigs to produce an audio optimised system.
Roons owner told me usb reclockers are rubbish XD
First, "USB Reclocker" is a misnomer. You cannot as such "reclock" USB. And once of sufficient quality for reliable operation, improving the clock for the USB subsystem has no effect, as the USB clock is not in any way linked to the audio side clock. (Incidentally all said here about USB also largely applies to Ethernet Audio).
As USB Audio streaming is isochronous, you have guaranteed bandwidth but no retransmission of corrupt data. This is the opposite of "bulk mode" (which is used by HDD's) where data integrity is guaranteed but bandwidth is not.
Most people a little more involved in IT will have had cases where a normally fast HDD transfers data extremely slowly and changing the USB cable fixes the speed or where one USB port drops the HDD transfer speed compared to others. The issue here is signal integrity. If the device was not an HDD but a USB DAC the sound may be changed substantially, you might even get dropouts, with a HDD the frequent retransmission of data (and with mechanical HDD the need to go back to previous data in the reading of the data chain) causes slowdown instead, but Data is not corrupted.
How so? If there is corrupt data (more often than you think) the USB Audio subsystem will use the same error handling as CD, starting with error correction (if there is enough good data to reconstruct the corrupt data) in which case all is well, UNLESS the USB Bridge consumes extra power when handling errors, which may generated error dependent power supply modulation that affects clocks or audio.
If enough data is corrupt that errors cannot be corrected, they are concealed by interpolating the missing sample in a straight line, finally, if there is not enough good data for interpolation the system mutes the audio stream.
If a USB Audio device run's in USB Audio Class 2 (as opposed to USB Audio Class which runs on USB 2 full speed - 12Mbps), it always runs in "High Speed" (480Mbps) mode even if it only streams 16/44.1. The difficulty in reliable duplex data transmission via a single pair of wire at 480Mbps is orders of magnitude more challenging that doing it at 12Mbps.
In addition, the USB interface circuitry is often on chip with the relatively computing intensive USB Audio Bridge, meaning it will be compromised compared to a stand alone item or just USB Chip (e.g. Hub) and usually are more sensitive to poor signal integrity.
These are the mechanisms (other than ground loops and USB Power quality) that underpin the audible differences in all this USB stuff. Under ideal conditions there will be nothing to observe. If (say) the PCB layout at a USB socket on a laptop or PCB motherboard is slightly subideal and if then a cable is near tolerance limits, despite both passing USB 2 certification, combine them and our problem risk materialises.
The key issue as a result is signal integrity. We have roughly three kinds of USB "Gizmo's" in terms of how they handle USB data ((power, ground etc. is left aside here).
An Isolator is inherently also a a repeater, so while it will break ground/earth loops it will also create a new clean high signal integrity USB signal on the downstream port and needs to receive a USB signal on the Upstream port.
A Repeater (often advertised as "Reclocker", "Regenerator" etc.) does not isolate (but a ground loop breaker could be build in like with the iFi iUSB micro) but also addresses signal integrity issues. In terms of signal integrity, these are as effective as an Isolator.
Finally a filter (like iFi iPurifier V1 and it's clone the AQ Jitterbug or Pioneer BonnesNotes DRESSING) is purely passive and may use a number number of strategies to avoid interference causing signal integrity issues. The effect on signal integrity is usually less than repeaters but can nevertheless be notable.
So, USB audio is subject to issues that may be improved by, among other means (such as best practice design of components), "USB Gizmo's" of varying nature.
Non of these devices however is a "USB Reclocker" in the sense of (say) a reclocker of SPDIF or AES-EBU signals, as the audio clock in USB Audio is in no way dependent or linked to the USB System's clock and clock quality in such devices does not affect audio clock jitter in the DAC attached.
Of course, a poor quality and jittery clock in the USB Gizmo may cause data errors and corruption, so a minimum level of clock quality is essential to the function. Often such Gizmo's run on chip's with a designed in crystal oscillator that only needs a crystal attached (no other parts).
Perversely, feeding such a chip with a 50 USD "Femtoclock" oscillator instead of the crystal it is meant to use will make things WORSE rather than better, as it exposes the clock signal interface to so-called "Ground Bounce" in the IC causing jitter that is absent if the 50 cent generic crystal is used...
Thor
) and autocorrection protocol can not be used to fix it the data is still corrupted, am I right or are you referring to some other condition as signal integrity error?