[LAU] [SOLVED] Example captured: Crackles in audio, drifting intermittent noise etc.

Will Godfrey willgodfrey at musically.me.uk
Tue Nov 1 07:02:30 UTC 2016


On Tue, 01 Nov 2016 00:24:16 -0400
termtech <termtech at rogers.com> wrote:

> Yeah nice one but I doubt heat related - more like voltage drop spike 
>  related, more activity on the power rails causing more spikes causing
>  the audio card to glitch. 
> 
> In the original thread we talked about weak capacitors on the sound card.
> 
> I was seriously contemplating that possibility once again today, 
> until I managed to capture the sound and analyze it :
> 
> https://dl.dropboxusercontent.com/u/53315356/Test_tone_passthru_noise.ogg
> 
> It is a recording of a hardware audio sine generator (my keyboard), 
>  input to the Delta1010, recorded as it is being passed through to an 
>  audio output via Jack. What you hear is what I hear out of that output.
> 
> Hey all you digital audio experts!
> Listen to the recording, be patient and watch how the distortion slowly
>  drifts in then out. Open it in Audacity or something, and notice the spikes.
> 
> Seems to me, that you can tell by the distortion's overtones which slowly 
>  rise in spectrum then disappear, that this /resembles/ a textbook case of 
>  some read pointer 'meeting up with' and passing some write pointer and 
>  when they meet there is distortion. 
> In other words they are at slightly different /rates/.
> 
> Do you hear what I mean about being easily recognizable as a sync problem?
> 
> But how? And why core related?
> I mean... I'll have to check chip docs but the card may have separate 
>  read and write sample rates... 
> But no, look closely at the spikes in the wave in Audacity. It appears some 
>  buffer is 'starting' too early or too late - like the buffer is suddenly 
>  being switched in the middle of a wave and it is that switching time 
>  that is varying.
> 
> So maybe not a /soundcard/ clock rate problem but something is not filling 
>  the buffers at the right time? Again - core synchronization?
> 
> Also, aside from the drifting noise, notice the slight glitches in the sound 
>  that are pervasive throughout. 
> /That/ is what I hear even with simple playback of test tones on websites!
> You can hear the clock rates jumping around slightly. Yes, I know this is 
>  /usually/ caused by Speed Step, Hyper Threading and so on, but the
>  fscking thing does it with all such setting turned off. On the other PC too.
> 
> Only choosing TWO (or one) cores instead of FOUR stops all these noises.
> 
> Thanks for listening. Whaddya think?
> Tim.

As a matter of interest, have you got access to a USB sound unit? If so does
that exhibit the same problem?

The reason I ask is that I have a Novatech nspire laptop with a quad core i5,
and have no issues at all using it with a KA 6 - indeed I used it on last
year's LAC where it was running for an hour and a half. At that time it wasn't
even running an RT kernel.

-- 
Will J Godfrey
http://www.musically.me.uk
Say you have a poem and I have a tune.
Exchange them and we can both have a poem, a tune, and a song.


More information about the Linux-audio-user mailing list