In tests conducted when Hugo was being developed, there was no significant difference in measurements from optical, coaxial or USB.
But recently in order to be able to test Dave, I purchased a Audio Precision APX555. This sets new standards in test equipment, being able to measure distortion at -150 dB with 2.5v RMS input.
So I thought I ought to re run the measurements:
Dynamic range (-60dB, ref full OP, Awt)
Optical: -120.7 dB (l) -120.7 dB (r)
Coaxial: -120.7 dB (l) -120.7 dB (r)
USB: -120.7 dB (l) -120.7 dB (r)
All identical.
Using 3v RMS output THD and noise ref full OP A wt:
Optical: -114.6 dB (l) -114.2 dB (r)
Coaxial: -115.5 dB (l) -115.0 dB (r)
USB: -115.5 dB (l) -115.4 dB (r)
Surprise in that the clearly better sounding optical (smoother much deeper sound stage) is slightly worse; this is possibly an AP ground issue.
On jitter using the J-test, I can resolve 500 fS of data related jitter - and I measured absolutely no jitter artifacts at all with all of the inputs.
Here is a stereo plot from the USB input using 24b 48k J-test, the -160dB noise floor equates to jitter of 500 femto S.
So there is no significant measurement difference from the USB to the optical or coaxial inputs. And, Hugo has no measurable (its below 500 fS) source jitter artifacts on USB, optical or coaxial inputs.
Rob