High amounts of THD over USB and Coax...

[rastus]

robertzombie > All this interest, means we are interested brother... not trying to pick fault, just very interested.

[robertzombie]

You can see Archimago's measurements where, using ODROID-C2 with Volumio 2 and USB digital music streaming, he gets excellent results. So its perfectly possible to get good RMAA results from a SBC, I just need to figure out why my Pi is throwing up anomalies through two digital outputs

[robertzombie]

[rastus]

[robertzombie]

Test results with wifi and bluetooth disabled, and then also with a different power supply, coming soon.

[PeteB]

[robertzombie]

Same results with wifi/bt disabled and alternate power supply

[silvrr]

Did you try without the ethernet cable connected?

Also, do you have the software or hardware volume control enabled? I can hear issues with the software volume control so that may be another item to eliminate.

[robertzombie]

[robertzombie]

Aha! I've figured out what the problem is, but I can't figure out why this is happening. The error is limited to the Pi 3's digital outputs and it appears to be some form of compression. Here is an image of the RMAA test file:



And here it is, captured from the Digi+



Notice the peaks are compressed? For the record, the recording of the Pi's analogue out is a faithful looking reconstruction of the test waveform so my playback settings are correct.

Why on earth is the Pi's digital audio output doing this? (yes, USB output gives the same results).