Interesting results so far - in that there's no clear benefit/degrade of specific changes. I though I'd share what the modification is in test #2 to encourage more people to try if they are interested...
Test #2 contains 4 test case these are: A: runs a process which does nothing (sleeps for a second and then restarts) B: runs a process which uses as much cpu as possible, but runs as a "IDLE" priority process, i.e. uses all cpu when no other processes are running C: as B but also reads and writes from random locations in memory (memory block larger than the cache to create constant activity on the memory bus) D: does nothing The basis of this is to examine the hypothersis that cpu load and memory access have an impact on sound quality. Its possible to argue that the load on the cpu and or the frequency and duration of tasks running on the cpu could have an impact on sound. Also that turning priorties and buffer lengths influence the durations different processes spend running and the frequency they do. I also note that the linux kernel puts the cpu into a lower power mode whenever in the kernel's idle process so spending more time in idle will mean lower power load and also changing the time/frequency of the kernel doing real work vs stitting in the idle process is likely to change the load the cpu places on the cpu power rails. Now my test case was to see whether cases B and C have noticable and repeatable impact on audio quality in comparison to the two control cases of A and D. I would say from the existing data they do not. The inference from this would be that cpu load does not have a significant bearing on the audio quality, but I wonder whether now you know the test cases whether people are able to hear any difference.... -- Triode ------------------------------------------------------------------------ Triode's Profile: http://forums.slimdevices.com/member.php?userid=17 View this thread: http://forums.slimdevices.com/showthread.php?t=93583 _______________________________________________ Touch mailing list Touch@lists.slimdevices.com http://lists.slimdevices.com/mailman/listinfo/touch
