FR theme: Performance/Graphics issues

[disturb]

I don't get the 20% cpu on a blank project either, my cpu spikes and glitches only occure when interacting with the gui.

A simple test :
I dragged the same audio loop 5 times in the seq (so i have 5 different tracks)
Expand the tracks to see the waveforms, then shift select the clips (so that the accurate waveforms are displayed for all tracks)
After that, zooming in and out i get :
-60% cpu when zooming with ctrl + mousewheel, and the gui get more and more sluggish as i zoom in (to the point where it's not responsive at all)
-same cpu usage but the gui stays responsive when zooming with the gui control (dragging the dot)

Really weird.

[plastique]

i dont have any graphic issues.
but ext2 standalone uses up to 20% of cpu IDLE>(1 empty audio+1 empty midi)
if i load ext2 vst inside xt 1.4.....the whole thing dont use more than 8%.
conclusion.....the standalone is flawed with some seroius performance issue.no other sequencer i ever tried uses more than 5% idle.

Did you turn off this frequency display thingy already ? I believe it's somewhere in the setup options ...
I can remember, at the time before I noticed it and could turn it off, xt also required a huge cpu consumption (up to 80% on my low-spec test machine, up to 20% on the others), for basically doing nothing.

[disturb]

Anyone noticed an increase in load time recently ?

[abi]

Samplitude, just as Logic, are highly optimized - they perform buch better than Sonar.

In Samplitude you can go to 95% without having crackles and the gui stays as fast as with 5% - and with Logic it's very similar.

i just spent another hour on comparing the performance of sonar and samplitude (both le versions thoug i think they share the audio-engines of the full versions). you're right jens, when the first crackles appear in samplitude the cpu meter shows between 85 and 95 percent. it outperforms sonar, although sonar can manage (up to) 85 percent on my machine.

i mean it's not that you can't produce music with XT2. thanks to freeze i managed to finish a quite heavy project lately. but if XT2's performance was closer to that of the so-called big players it would certainly become more sort of an option also for people who own the "big players".

[jens]

i just spent another hour on comparing the performance of sonar and samplitude (both le versions thoug i think they share the audio-engines of the full versions). you're right jens, when the first crackles appear in samplitude the cpu meter shows between 85 and 95 percent. it outperforms sonar, although sonar can manage (up to) 85 percent on my machine.

the biggest difference between the two you'll experience when you zoom and scroll and do other screen-operations while they're almost maxing out...

[ik_ik_ik]

Samplitude, just as Logic, are highly optimized - they perform buch better than Sonar.

In Samplitude you can go to 95% without having crackles and the gui stays as fast as with 5% - and with Logic it's very similar.

i just spent another hour on comparing the performance of sonar and samplitude (both le versions thoug i think they share the audio-engines of the full versions). you're right jens, when the first crackles appear in samplitude the cpu meter shows between 85 and 95 percent. it outperforms sonar, although sonar can manage (up to) 85 percent on my machine.

i mean it's not that you can't produce music with XT2. thanks to freeze i managed to finish a quite heavy project lately. but if XT2's performance was closer to that of the so-called big players it would certainly become more sort of an option also for people who own the "big players".

Whilst I'd agree you can, thankfully, produce music by often having to resort to freezing things one way or another (and holding your breath whilst recording) you simply can't perform it. The downside (as a long time sufferer of XT performance problems) is it so quickly becomes a process of pre-meditatated preparation and construction. That's not ideal at all - especially when the headroom is available but it's crippled by the GUI for some reason.

On the subject of CPU meters - different applications measure different metrics so comparisons are hard to make. I seem to recall XT's meter is tied to buffer processing so CPU meter might be a bit of a misnomer? Other applications might be measuring things differently for example showing absolute system CPU load so the numbers will never stack up against each other or even indicate the same bottleneck. I've certainly seen XT's meter maxed out but Task Manager showing 50%.

On a bit of a tangent I really liked Art Teknika Console when I came across it years ago as it neatly showed you the CPU hit of each plugin and is it Reaper I read about recently that has really detailed CPU monitoring? It would be great to get a real handle on this in XT too. Would be even better to be able to throttle individual plugins somehow - how much safer it would be to bring a plugin to its knees rather than let a plugin bring everythng else to its knees...

Inigo

[jens]

On the subject of CPU meters - different applications measure different metrics so comparisons are hard to make. I seem to recall XT's meter is tied to buffer processing so CPU meter might be a bit of a misnomer? Other applications might be measuring things differently for example showing absolute system CPU load so the numbers will never stack up against each other or even indicate the same bottleneck. I've certainly seen XT's meter maxed out but Task Manager showing 50%.

yeah, of course you need to use the task-manager to ensure objectivity (which hence is what I always do)

[1barEnuff]

tonight i gave xt2 a run round the block and didn't experience
any
serious problems straight away, although i do have a fairly decent computer compared to my old p3 933. i did however notice the scroll thing, the cpu usage does climb bigtime but until i compare it with another sequencer i wouldnt know how bad it is.anyway i did notice one thing and that is: when i add a reverb the sound goes crazy. first i added sir as an insert on the main output (3+4 in my case)
and it crackled real bad also i noticed that triplets was enabled in the object bar i could hear along with the interference/crackling sound a distinct triplet clicking noise and when i turned triplets off,it stopped but the general crackling/digital distortion never, also the sound was a kind of gurgling sound like underwater or something and the cpu was only about 55-60 percent.i also tried ambience and the fx comp reverb and they all messed the sound up though sir was the only one that clicked in time. the least noisey was classic reverb but it still made the sound crackle. this was a fairly big project in side ext1 vst in xt2 with midi tracks loads of vst effects and a good few vsti's and audio tracks but without a teverb in xt2 it was sound. man this an essay for me!

cheers, john

[_starcraft_]

i dont have any graphic issues.
but ext2 standalone uses up to 20% of cpu IDLE>(1 empty audio+1 empty midi)
if i load ext2 vst inside xt 1.4.....the whole thing dont use more than 8%.
conclusion.....the standalone is flawed with some seroius performance issue.no other sequencer i ever tried uses more than 5% idle.

Did you turn off this frequency display thingy already ? I believe it's somewhere in the setup options ...
I can remember, at the time before I noticed it and could turn it off, xt also required a huge cpu consumption (up to 80% on my low-spec test machine, up to 20% on the others), for basically doing nothing.

yes that frequncy thing dont really make that much difference to me.
i have a 1.86 centrino.....i know its not powerful.....but no other sequencer i ever tried consumes more than 5% idle..... infact xt1.41 doesnt consume hardly any cpu at all idle.
so far i'm just forced to use xt 1.41

[abi]

On the subject of CPU meters - different applications measure different metrics so comparisons are hard to make. I seem to recall XT's meter is tied to buffer processing so CPU meter might be a bit of a misnomer? Other applications might be measuring things differently for example showing absolute system CPU load so the numbers will never stack up against each other or even indicate the same bottleneck. I've certainly seen XT's meter maxed out but Task Manager showing 50%.

yeah, of course you need to use the task-manager to ensure objectivity (which hence is what I always do)

for the record: i use the task manager's cpu meter, too.

[horn]

Ok,
I just started testing my ssytem...
I am not getting the 20% cpu hit for a blank project that you all are.
I have 18 audio tracks playing and have only reached 16% for the energyXT2.exe thread in my task manager. No clicks, pops, or stutters or graphics issues.

Dell Latitude D620 laptop: T2400 @ 1.83GHz x2, 1GB RAM, integrated simtel audio w/ ASIO4All, NVIDIA Quatro NVS 110M supporting a second monitor.

This is true. The 10.06. beta seems to be already really optimized. I did some recordings with an electric guitar yesterday (Standalone and VST versions) and I must find out that the latency behaviour is already much better than within 1.4. Furthermore i realized that if you just play around with loops the CPU usage doesn't increase any more to the amounts of the first versions.

[ik_ik_ik]

On the subject of CPU meters - different applications measure different metrics so comparisons are hard to make. I seem to recall XT's meter is tied to buffer processing so CPU meter might be a bit of a misnomer? Other applications might be measuring things differently for example showing absolute system CPU load so the numbers will never stack up against each other or even indicate the same bottleneck. I've certainly seen XT's meter maxed out but Task Manager showing 50%.

yeah, of course you need to use the task-manager to ensure objectivity (which hence is what I always do)

for the record: i use the task manager's cpu meter, too.

Cool. Just wanted to point it out, in general, as these XT performance discussions more often than not descend into lots of anecdotal comparisons and then fizzle out whilst there's a fundamental and repeatedly confirmed problem that needs resolving.

[_starcraft_]

This is true. The 10.06. beta seems to be already really optimized. .

where do i find it? i only see the 8.06 on the site.
also would be useful to know how much cpu xt2 is usin idle on each computer.
ideally a sequencer should NOT use any cpu at all when idle.
even if its 5%-6% on a new system.....its still cpu wasted.

what if i wanted to install xt2 on those UMPCS which have just come out....their cpu is limited to 1.1 or 1.4 ghz in most cases.
if this werid cpu usage is not solved xt would become not suitable for those systems....which is a shame cos its the only host i know which would work nicely on small screens.

[1barEnuff]

hi, on my internet computer which is a pentium 3 933 with a voodoo3 video card and 512ram and windows xp pro(the minimum recommended specs are1.4ghz i think) xt2 uses between 37 and 44 percent cpu when i open it up empty on the other hand on the computer i use for music which is an amd 3800x2 with built in graphics nvidia (6200 i think) 1gb ram which the graphics share 128mb and windows xp pro x64 xt2 uses two percent cpu when i open it up empty. so it looks like different specs make a massive difference to me

cheers, john

[_starcraft_]

hi, on my internet computer which is a pentium 3 933 with a voodoo3 video card and 512ram and windows xp pro(the minimum recommended specs are1.4ghz i think) xt2 uses between 37 and 44 percent cpu when i open it up empty on the other hand on the computer i use for music which is an amd 3800x2 with built in graphics nvidia (6200 i think) 1gb ram which the graphics share 128mb and windows xp pro x64 xt2 uses two percent cpu when i open it up empty. so it looks like different specs make a massive difference to me

cheers, john

here we go....this seems to give a better picture. i hope it helps nail down the prob.
how much does xt1.41 use on your internet machine?
i remember on my old pIII 1.0ghz Logic 5.5 would not use any cpu all when idle....it would be the plugs or processing which consumed cycles.
so why now a sequencer need to pick up cpu...and so much of it?.
even if its only 2% on new machines...if that 2% adds up each thing u do no wonder the performance fails.

btw i i tried the new beta etc...
i get 9% cpu specifically from XT2 idle.(task manager process information)
thats my final and precise answer. (i'll corect where it says 20%)
1.86 centrino dothan processor.
x300 ati graphic card.
1gb of moemory
9% gone for opening xt2....compared to the 0-2% hit when i open XT1.41.
xt2 uses 4 times more cpu than xt1 it seems.