soylent.green wrote:I've still problems with asio. Nobody can tell what's wrong with it. And there's no more vital sign of sonic core in the past tow weeks. what's going on? Is 64bits win7 to big a hurdle to take.…
Today I've installed the new 64Bit RC which should bring more improvements including enhanced ASIO-Drivers. Full of excitement I've checked it with my simple Cubase project (1 instrument track Halion Player...). It's still crackels and noises ;-(
But I've found out that noise artefacts are almost gone when setting processor affinity for cubae.exe and scope.exe to exactly the same *one* processor core. So I run now Cubase and Scope on one core and leave the other 3 for the idle process to have some fun...But though this setup improves the situation on the crackle side, yet they are not completly gone. Compared to a maximum of 2 -3 crackles per second before I now have 1 crackle in 30 seconds.
It seems there's some multicore problem with the asio drivers. Can anyone else confirm that (I mean that setting processor affinity reduces crackles)!
--- update ---
It's even true for lower ULLI settings. Switching to 7ms@44.1KHz shows some crackles but by fare less than without having setting processor affinity.
Hm, this is another very very interesting point:
1.) When switching scope.exe to run on core 1 and cubase to run on core 1 - By far the4 best setup, as described above
2.) Scope.exe on core 1, cubse on core 4 - Completly full of crackles, it's more than 5 or 6 per second
3.) Cubase on core 3 - Better than setting in 2.) but still lots of crackles
4.) Cubase on core 2 - It's like setup 1, probably even better, I'm not quite sure.
5.) Cubase on core 1 and 2 - Results like for setup 1
I'm puzzled...Does it depend on the core distance, distance in term of core number. Distance 3 from core 1 to core 4 gives the most terrible results...Any processor architecture gurus that are able to shed some light on this????