“official”?   “Linux supports UAC2.”

Both are transparent.

    All the magic smoke got let out of that chip.   Good news: you get to buy yourself a christmas present, a shiny new sound card!

Yes; the STX II base card is exactly the same.

The are several different GX620 models; yours looks like "Desktop" or "Small Form Factor".   Anyway, you'd need a low-profile sound card, such as a Xonar DG/DS/D1.

  No; the card's DMA controller does not support moving data in a format suitable for DSD.

That "usb_set_interface failed (-71)" definitely indicates a hardware problem.

So "hanging" means that nothing is shown on the screen? This is normal (except for the missing sound).   Google says others have the same problem, but only with Intel USB controllers, and that the device forgets its own name (see "lsusb -v"). This points to a hardware problem.

What kernel version are you using? Those various players, do they hang too?   Are there any error message in the system log (see (/var/log/messages, or the output of dmesg)?

    For the effects of heat on the sound quality, see page 12 of the PCM1792A datasheet.   The card itself should work fine. None of the card's components are run near their limits. (Compare with a CPU or GPU: 192 kHz is 0.000192 GHz.)   If you want to improve cooling, remove the...

From what I've heard, the microphone input has not changed (i.e., designed for nothing more than VoIP or Karaoke).   Use your own mic preamp connected to the line input.

Galaxy Far East Corp makes the Tenor chips.   There is an official specification for high-speed USB audio devices (USB Audio Class 2.0, which is not the same as USB 2.0), but according to their website, the Tenor firmware does not support that and requires a hardware-specific driver. What you...

    The Xonar DG uses the CMI8786, which is a relabeled CMI8788. The Rosewill 702 uses the CMI8768, which is just a CMI8738 with two additional channels.

  The driver has been merged into the upcoming 3.11 kernel.   Quote:

This sounds as if the amplifier does not get enough power. Check the power connection.

There are indeed measurements where a signal transferred over a coaxial cable has better rise times than with a Toslink cable. (How much that affects the jitter is another question.)   However, both coaxial and Toslink have the problem that the clock must be derived from the signal. Using...

This is a well-known bug in the Xonar Audio Center (inherited from C-Media's similar tool); there is just no mechanism to choose between multiple instances of the same driver.   Using VMWare won't help you; an emulated sound device inside the virtual machine would require a working driver on...

Graphic cards don't get their power from the S/PDIF input, so it's possible that the Phoebus output does not have +5 V. (Can you measure this?)

At the moment, the USB audio driver has no mechanism to tell applications what kind of outputs it has, so what you see is just a guess.   There is no analog data inside the computer or on the USB bus; "Analog Output" means that the device will convert the data into an analog signal. With the...

  Why do you think this is possible?

No, USB is better. An S/PDIF receiver must reconstruct the sample clock from the signal, and the signal gets distorted due to the cable's impedance, thus introducing jitter. When using asynchronous USB, the receiving device uses its own sample clock.

The USB descriptors don't tell much.   The Windows driver, however, is from Thesycon, which means that the chip is not from C-Media or Archwave, who have their own drivers.   I don't know of any USB Audio Class 2.0 chips except those from these companies and XMOS.

Creative says:   If there were any impressive technical measurements, Creative would shout them from the rooftops.   There is a reason that all sound cards that set a great value on sound quality take care to separate the digital and analog processing stages.

S/PDIF signals use biphase mark code and contain two types of information: the PCM values, encoded in the number of signal flanks, and the word clock, encoded in the timing of the signal flanks.   If the clock signal were a simple square wave, any degradations during transmission would...

The DAC chip itself, the PCM1793, does support 24 bits at 192 kHz. However, the USB interface chip, the PCM2707, does not support more than 16 bits at 48 kHz, independent of what driver you're using.