I guess what I’m reacting to is the subjective audibility of jitter.
Jitter is normally measured in pico, nano or microseconds. Anything over 100 microseconds results in data loss, gaps pops and clicks so can be heard by anyone. Computer component manufactures intentionally insert jitter into the master clock of the computer, and CPUs in order to limit the RF emissions from them, and thereby pass FCC testing. This is desirable if you want to sell any computers. And more than compensated for by the way a CPU processes, caches and buffers data.
Audiophiles claim to be able to hear jitter at the 10’s of nanoseconds range. Most USB to spdif convertters claim to be able to reduce jitter to the 10’s of picoseconds range. However measuring jitter is pretty hard normally. Expensive equipment is needed, but with the Cubox-i data we can make an approximation.
At 44100 it is misampling at a rate of 17/44100 per minute or 6.4 microseconds per second. This is well in line with internal data tolerances but impossible for any internal algorithm or synchronous DAC to reclock accurately by a factor of about a million. If my maths is right.
The only way to de-jitter is to bypass the audio circuits completely via usb and allow the clocking to be done asynchronously by an external async DAC, or async USB converter.
I can assure you that the cubox-i toslink output sound quality, at 16/44.1 is a lot worse than the reclocked USB/spdif/Async Dac output and is most certainly audible.
If you are worried about aesthetics you can get async USB/spdif converters in USB stick form with a single coax spdif at the other end.
Yes yes I know! I know! And knew.
I said it was another matter altogether meaning it’s off-topic :mrgreen:
Again: know and knew. BTW anyone can claim, but some of those claims don’t seem to hold up under scrutiny. Some claim to hear differnces in lossless files played from different disks. Audiophools is a term sometimes used. We need to remain critical. This is the paper Ethan Winer refers to in his post over here.
And this some of his other input on Artifact Audibility: http://ethanwiner.com/audibility.html
Yes and there are online jitter calculators too.
This level of correction ‘accurately’ is to the best of my knowledge not claimed anywhere by either Adrian (Triode) or myself. He does however claim to have done some improvement via SW on the Wandboard, thinks the issue on Cubox-i can be addressed and I don’t know why that might be untrue? He most decidedly never claimed full resolve!
If you want to have it done to the level you suggest yes. Not disputed before by me, nor now.
To the first part: lot worse, yes. Of course. Obviously.
No assurance required here (that’s an authority argument BTW so invalid, but I get your point). The audibility supposedly being most certain however will not stand as it depends on the listener’s ears -as I stated previously. Seen no argument against that so far.
Know that of course. But even these are way too wieldy keeping in mind the svelte Cubox-i 2x2x2".
Just not right aesthetically or otherwise to buy a tiny all-in-one and then add extra stuff to the supposedly all-in-one, some ‘solutions’ being larger than the device itself.
Not worried myself BTW, have plenty other (Linn DS, DACs, HTPC) devices available, using Cubox-i for tinkering. Don’t want and can’t use electrical S/PDIF, must have Toslink for this particular application.
The better solution IMO would be to use the older Cubox which can be found quite cheaply and which does an OK job clockwise while maintaining the fantastic form factor. Think I have one lying around somewhere.
Ontopic: the question was why it downsampled, the answer was given correctly by volumio’s developer :mrgreen:
I guess what I’m reacting to is the subjective audibility of jitter.