ouch, live 7 is glitchy!

GUY SMILEY · Post by **GUY SMILEY** » Mon Jan 21, 2008 9:19 am

Cache wrote:I discovered it was my Remote SL assigned as a control surface that was causing my glitches+pops.
<from post to bugs>
I have a remote SL with the latest drivers from novation.
when i have the Remote assigned as a control surface in my preferences, i get pops and glitches (Cpu spikes up to 50%) when i move the volume faders of a playing audio track.
If i have remote assigned as multiple control surfaces (eg port 1, port 2....) then the pops and glitches get more extreme (Cpu spikes up to 200+ %)
If i remove the Remote from the control surface assignments, then the pops and glitches disappear.
this is with Beta3
I'm on windows and using a DMX6Fire sound card.
the latency or whether im using 96 or 44.1 khz doesn't make any difference.

I am also suffering from this Glitchorama with Motu Traveller (and nothing else) connected. MME or Direct Music both

Timur · Post by **Timur** » Mon Jan 21, 2008 9:43 am

Guy Smiley, do you still suffer from the glitches with 7.02 B3 and using MME? I can still produce glitches with that Beta, it's much better than 7.01 and nearly on par with 6.10 (actually even better than 6.10 for some settings). I'm closing in on the problem though, there is a definitive connection with the audio-driver/interface being used (I own 3 different ones).

Cache · Post by **Cache** » Mon Jan 21, 2008 11:44 am

holy sH8t. its an ASIO/LIVE problem

yeah i changed my audio driver type to mme/direct and the cpu spikes are gone.

wow.

this is like a 1990s problem. hahah

Timur · Post by **Timur** » Mon Jan 21, 2008 12:23 pm

I'm still investigating, but there is a fundamental difference between my three audio-interfaces, one would glitch at 440 buffers while with the other I can add half a dozent more Midi output tracks at only 256 buffers without glitches.

GUY SMILEY · Post by **GUY SMILEY** » Mon Jan 21, 2008 1:47 pm

Timur wrote:Guy Smiley, do you still suffer from the glitches with 7.02 B3 and using MME? I can still produce glitches with that Beta, it's much better than 7.01 and nearly on par with 6.10 (actually even better than 6.10 for some settings). I'm closing in on the problem though, there is a definitive connection with the audio-driver/interface being used (I own 3 different ones).

Hi Timur,
yes still experiencing same problems with the new beta release.

Timur · Post by **Timur** » Mon Jan 21, 2008 1:50 pm

I'm working to create a video that demonstrates how Live 6.10 and Live 7.02 are affected by Midi output with different audio-interfaces being used. Just need to find a good freeware screen capture software and the best way to record the sound.

GUY SMILEY · Post by **GUY SMILEY** » Mon Jan 21, 2008 3:37 pm

i'll watch it

Timur · Post by **Timur** » Mon Jan 21, 2008 4:06 pm

All in all I have solved the riddle of how Midi and ASIO output are connected to each other and how Midi slows down ASIO. Whenever you double the number of audio-buffers (in samples) you can output double as many Midi events. Strangely there are huge differences between my different audio-interfaces, with the X-Fi performing worst, then the Kore 1 and the M-Audio performing incredibly better than the other two.

I can crash Live 7.02 regulary by changing audio-buffers during playback or changing interfaces, the X-Fi is most prone to crashing here. While I write this Live has crashed but stays in memory with the X-Fi driver locked and an ongoing loud tone coming out of my speakers.

Interestingly turning off multi-processor support does affect the performance with these tests in that the sound gets more affected and Live's CPU meter doubles, but the real CPU load does not change according to task-manager is stays well below 50% even when Live is already running havoc.