joeyjojo
100+ Head-Fier
- Joined
- Mar 14, 2011
- Posts
- 172
- Likes
- 19
I discovered some useful things yesterday which might be useful to others.
I've been using foobar WASAPI in windows vista -> xonar DX -> toslink -> dacmagic -> amp etc. There are three important details in this tale. First, for years I've been running BOINC, which is a distributed computing platform which loads the CPU at 100% day in day out for scientific calculations. Second, like many computers my motherboard has EIST etc. to clock down the CPU when not heavily loaded. Third, with this system I sometimes noticed very slight clicks occasionally in the music, but they were so very practically-inaudible I ignored them.
Recently I bought an M-Audio 2496 to test its ADC capabilities (quite poor by the way) and coaxial out. Using the same chain as above, listening to WASAPI foobar over coaxial there were occasional, very noticable dropouts in the music. Annoyed, I wondered why the DX would be fine but the M-audio not? So with a bit of research I found this thread and this program called LatencyMon.
Sure enough, running latencymon told me in big red letters that my computer was incapable of real-time audio and would suffer from dropouts, and suggested this might be due to the CPU power saving options. With these turned off in the BIOS but still running BOINC it didn't do much better. With BOINC off but keeping EIST, same problem, dropouts whenever the CPU clocked down. Finally with power saving off and BOINC off, it measured the latency as sufficiently low to avoid problems.
And sure enough, clicks are gone from both cards running them this way! (As a compromise I have BOINC operate only when the computer has been idle for a while.)
Not sure how many others run BOINC and foobar at the same time but maybe they'll find this useful.
I've been using foobar WASAPI in windows vista -> xonar DX -> toslink -> dacmagic -> amp etc. There are three important details in this tale. First, for years I've been running BOINC, which is a distributed computing platform which loads the CPU at 100% day in day out for scientific calculations. Second, like many computers my motherboard has EIST etc. to clock down the CPU when not heavily loaded. Third, with this system I sometimes noticed very slight clicks occasionally in the music, but they were so very practically-inaudible I ignored them.
Recently I bought an M-Audio 2496 to test its ADC capabilities (quite poor by the way) and coaxial out. Using the same chain as above, listening to WASAPI foobar over coaxial there were occasional, very noticable dropouts in the music. Annoyed, I wondered why the DX would be fine but the M-audio not? So with a bit of research I found this thread and this program called LatencyMon.
Sure enough, running latencymon told me in big red letters that my computer was incapable of real-time audio and would suffer from dropouts, and suggested this might be due to the CPU power saving options. With these turned off in the BIOS but still running BOINC it didn't do much better. With BOINC off but keeping EIST, same problem, dropouts whenever the CPU clocked down. Finally with power saving off and BOINC off, it measured the latency as sufficiently low to avoid problems.
And sure enough, clicks are gone from both cards running them this way! (As a compromise I have BOINC operate only when the computer has been idle for a while.)
Not sure how many others run BOINC and foobar at the same time but maybe they'll find this useful.
