Some background on repeating soft clicks

[PeterSt]

I know that when something is wrong in the software, repeating (very) soft clicks occur. Something like prrt prrt prrt prrt (say these 4 times are per second);

Yesterday I had this one time with the 0.9pd version myself, and it looks similar to what others have reported; once it is there it won't go away (but it will be reset when a next track starts), and you'll get rid of it at pressing pause-start or at changing one of the parameters which change settings in real time (like Invert, Q1 position (not priority !)).

It is my guess, what happens is this (derived from the knowledge of the software being wrong, which is not the case here) :

In one occasion the processor was not able to cope with the refill of the buffer towards the DAC (better : soundcard). Say, it went behind.
Because of this the synchronization between the knowledge of the amount the buffer is filled (the buffer of the 48 samples which is talked about more often), and how it actually is, is out of order. Now each and other time the buffer runs empty the DACs keep on repeating the last fed samples (or similar, but that's not important).
For the general sound picture you may not even notice it (but it is dead wrong obviously), and at more wild music you won't hear the clicks.

People report this, I think, since the Processor Core Appointment.
Also, there would be logic in this to happen, because while the sound engine is appointed to its own core, when it itself needs more (within the given time) than the one core can cope with ... you're out of luck, whereas without the Processor Core Appointment, there was the other core jumping in.

If you have these kind of clicks, please report it;
The fact that I had it once in xx days does not worry me. But if you have it several times during a few albums, it is not the way to go.
Please note that you should judge this per 0.9q (and up) were it for comparing with Appoint Processor Core shut off (the older versions were not completely correct in that). When you don't compare, you can use 0.9m-1 and up (BUT when you have 4 cores you'd need 0.9q anyway).


For techies reading this : No, I can't recover (nor signal) such a situation because of the way I deal with the buffers.


My judgement is based on one situation only so it may be premature. Otoh, I combine it with other people reporting similar (which is only similar (ehh, the same) when it is about the soft clicks as described).

Thank you,
Peter