uploaded control v133 and play v44, optimised play and control. More susceptible to glitches if other things happening on pc, but best sq.
Worked out a better way to communicate between control and play.
This play version has the minimum instructions and is all register moves. The previous version I said was the minimum code had 2 extra moves, but have removed those. So this version is proof of the theory that minimum code in the render loop is best.
after play it tends to lock up the pc for a while, so stop it before the end if that's a problem.
uploaded control v45 which doesn't have the tendency towards glitches, don't know how much it affects sq doing it this way.
133 + 44 (or 45) doesnt work on waveio with 2.29 thesycon. On 44 only heavy distortions. On 45 i hear music in the background through distortions.
131 and 43 plays fine.
No problem here
Preferable combo v133 and v44
Only problem at the end of the track Mqnplay resources and memory keep going up
Which eventually will lock pc
Very detailed version
It seams i can not use tasker with 133/44or45. Though all previous versions worked fine with assigned different cores for mqnplay, mqncontrol and rest processes.
My i5 'T' haswell is downclocked to 800mhz.. is it too weak for newest mqn versions?
What you think?
minionas wrote:It seams i can not use tasker with 133/44or45. Though all previous versions worked fine with assigned different cores for mqnplay, mqncontrol and rest processes.
My i5 'T' haswell is downclocked to 800mhz.. is it too weak for newest mqn versions?
What you think?
I run i7T but at its native speed. Also run tasker without problem. Setting mqnplay on realtime and a dedicated core.
I think current MQn (or maybe all KS MQn) are running on the sharp of properly timed data transfer.
I would try with native CPU frequency and see if it performs more stable and acurate
