Audio Glitches at CPU meter 20%
Audio Glitches at CPU meter 20%
I've no doubt this has been discussed before but if you try and search "CPU", "spike" and "RAM" you get about 5 million results so I'm gonna ask it again.
Right... I have a problem
I'm working with a pretty large live set. When I'm playing my CPU meter hovers consistently at around 20-23%. Occasionally I will get a horrible glitchy noise. Sometimes this is accompanied by a clearly visible spike on the CPU meter. I imagine when I don't see one it is a very quick, transient spike. These glitches happen when using effects such as delays, but sometimes simply triggerring a clip will cause one.
Why do these spikes happen and push me over the edge when my CPU meter is so low (when not spiking)? Is this something that more RAM would solve? I.e. CPU can cope but bottleneck in in the memory available. I'm not playing any clips from RAM so I'm not sure how more RAM would help here.
Does having a large set (lots of rows) use up more memory/cpu even if you only have the same number of clips playing simultaneously as you may in a more conservative set?
I really need to sort this out as I'm playing out with this software relatively regularly and though I've not had any particularly bad experiences in from of an audience, I have occasionally had severe problems in the studio which would be very embarrasing out live. Also, I'm somewhate reluctant to splash out on more RAM if something else (bug? CPU? ....) is likely to be the culprit.
Right... I have a problem
I'm working with a pretty large live set. When I'm playing my CPU meter hovers consistently at around 20-23%. Occasionally I will get a horrible glitchy noise. Sometimes this is accompanied by a clearly visible spike on the CPU meter. I imagine when I don't see one it is a very quick, transient spike. These glitches happen when using effects such as delays, but sometimes simply triggerring a clip will cause one.
Why do these spikes happen and push me over the edge when my CPU meter is so low (when not spiking)? Is this something that more RAM would solve? I.e. CPU can cope but bottleneck in in the memory available. I'm not playing any clips from RAM so I'm not sure how more RAM would help here.
Does having a large set (lots of rows) use up more memory/cpu even if you only have the same number of clips playing simultaneously as you may in a more conservative set?
I really need to sort this out as I'm playing out with this software relatively regularly and though I've not had any particularly bad experiences in from of an audience, I have occasionally had severe problems in the studio which would be very embarrasing out live. Also, I'm somewhate reluctant to splash out on more RAM if something else (bug? CPU? ....) is likely to be the culprit.
Hi, thanks for your help.
There seems to be no rhyme or reason. Sometimes I will get a glitch when playing a sole track of audio (all stereo) when I click on a new clip on that track. But it definitely happens more when there are more tracks playing.
If i want to induce a glitch i will have 9 tracks of audio playing with loads of delays and reverbs and go nuts with lots of parameters on my Control Freak. Sometimes I'll get glitches, sometimes I won't. That's why it's so damn frustrating.
The one thing I haven't tried is loading clips into RAM. Thing is I have loads of clips so I wouln't have a clue where to start. I wouldn't know which to load into RAM (some of these are whole tracks too - my set consists of a mixture of loops and whole tracks). Obviously I couldn't load all the tracks into RAM so I don't see how I could manage it so that some of the tracks playing simultaneously would be streaming from disk and some from RAM at any one time.
Thing is, I'd never suffered this problem till 2 weeks ago. The only thing that's changed is that I've developed my set and it's got bigger. Not THAT much bigger though and there's still no more going on at any one time than there was before.
The most annoying thing is that it's not even like I'm sailing close to the mark with the CPU meter.
There seems to be no rhyme or reason. Sometimes I will get a glitch when playing a sole track of audio (all stereo) when I click on a new clip on that track. But it definitely happens more when there are more tracks playing.
If i want to induce a glitch i will have 9 tracks of audio playing with loads of delays and reverbs and go nuts with lots of parameters on my Control Freak. Sometimes I'll get glitches, sometimes I won't. That's why it's so damn frustrating.
The one thing I haven't tried is loading clips into RAM. Thing is I have loads of clips so I wouln't have a clue where to start. I wouldn't know which to load into RAM (some of these are whole tracks too - my set consists of a mixture of loops and whole tracks). Obviously I couldn't load all the tracks into RAM so I don't see how I could manage it so that some of the tracks playing simultaneously would be streaming from disk and some from RAM at any one time.
Thing is, I'd never suffered this problem till 2 weeks ago. The only thing that's changed is that I've developed my set and it's got bigger. Not THAT much bigger though and there's still no more going on at any one time than there was before.
The most annoying thing is that it's not even like I'm sailing close to the mark with the CPU meter.
I had the same problem with my Acer Aspire laptop and i was on the verge of thowing it through the window. It would gitch regardless of load, soundcard or latency.
Then i stumbled upon: http://forums.amd.com/index.php?showtopic=17630&st=350
It turns out there's a problem with the battery monitoring function of some laptops - you can figure out if you're affected by moving hoving your mouse over the battery icon and waiting for the tooltip to appear (whilst you're playing audio). If it gitches then, then you've got the same problem. Read the thread for the reasons why...
Anyway, i fixed it by going into device manager and disabling "Microsoft ACPI-Compiant Control Method Battery" - it now works flawlessly, but the battery monitor doesn't work so i keep it plugged in most of the time.
Might help, might not.
FatCat.
Then i stumbled upon: http://forums.amd.com/index.php?showtopic=17630&st=350
It turns out there's a problem with the battery monitoring function of some laptops - you can figure out if you're affected by moving hoving your mouse over the battery icon and waiting for the tooltip to appear (whilst you're playing audio). If it gitches then, then you've got the same problem. Read the thread for the reasons why...
Anyway, i fixed it by going into device manager and disabling "Microsoft ACPI-Compiant Control Method Battery" - it now works flawlessly, but the battery monitor doesn't work so i keep it plugged in most of the time.
Might help, might not.
FatCat.
Don't want to speak too soon but tentatively... "you fekin legend!! "
It used to make the exact same glitch noise with a CPU spike on scren when I jacked the power cable out and it switched to batteries. I turned off the checking mode and now I can pull the power cable in and out to my heart's content without any dropouts. The fact that it's exactly the same noise suggests to me that you may have cracked it.
It adds up with the fact that it seems largely unrealated to track count etc if it's periodic battery checking.
Was annoying me so much cos it would be so unprofessional to play out with a glitching laptop.
Cheers mate.
It used to make the exact same glitch noise with a CPU spike on scren when I jacked the power cable out and it switched to batteries. I turned off the checking mode and now I can pull the power cable in and out to my heart's content without any dropouts. The fact that it's exactly the same noise suggests to me that you may have cracked it.
It adds up with the fact that it seems largely unrealated to track count etc if it's periodic battery checking.
Was annoying me so much cos it would be so unprofessional to play out with a glitching laptop.
Cheers mate.
Hi,
I know you guys are PC guys but does this solution apply to Mac's as well?
I have an iBook that glitches and it's slowly getting out of hand. Definitely happens the most when I use the reverb plugin that's in Live. CPU goes up to 42% at most (I'm sending multiple tracks to this main reverb effect)
And I agree, it don't get much more unprofessional than a glitchy laptop!
It happens on my ibook:
640 MB ram
900mhz g3 PowerPC iBook
M-audio Duo
2 mics
Thoughts anyone? Anyone else have this issue?
I know you guys are PC guys but does this solution apply to Mac's as well?
I have an iBook that glitches and it's slowly getting out of hand. Definitely happens the most when I use the reverb plugin that's in Live. CPU goes up to 42% at most (I'm sending multiple tracks to this main reverb effect)
And I agree, it don't get much more unprofessional than a glitchy laptop!
It happens on my ibook:
640 MB ram
900mhz g3 PowerPC iBook
M-audio Duo
2 mics
Thoughts anyone? Anyone else have this issue?
Live 6.0.3
1.25 GHZ Powermac G4
Mac OSX [10.2.8]
M-Audio Duo
Oxygen 8 Midi Keyboard
Digitech Vocal 300
Fetal Doppler
1.25 GHZ Powermac G4
Mac OSX [10.2.8]
M-Audio Duo
Oxygen 8 Midi Keyboard
Digitech Vocal 300
Fetal Doppler
Really wish I could be more useful seeing as my problem has (hopefully) been solved.
All I can think of (and these don't apply quite as much to Macs I don't think) is to make sure that any background tasks etc are turned off. It seems that most probs with CPU spikes on PC is when your machine decides monentarily to do something else rather than process audio - ie virus checking, checkin battery power, updating the clock, all kinds of silly housekeeping things that aren't really necessary.
All I can think of (and these don't apply quite as much to Macs I don't think) is to make sure that any background tasks etc are turned off. It seems that most probs with CPU spikes on PC is when your machine decides monentarily to do something else rather than process audio - ie virus checking, checkin battery power, updating the clock, all kinds of silly housekeeping things that aren't really necessary.
I'd echo that. From the initial post my immediate reaction was that neither Live 4 nor the CPU are to blame, but rather all the other crap that Windows gets up to all on its own
As far as I understand it, Live's CPU meter does not show the additional load that some of this stuff places - it shows the load that Live itself is using.
By the way, my specs are almost the same. I don't get audio glitches until around 70% in the CPU meter, which is not unexpected.
As far as I understand it, Live's CPU meter does not show the additional load that some of this stuff places - it shows the load that Live itself is using.
By the way, my specs are almost the same. I don't get audio glitches until around 70% in the CPU meter, which is not unexpected.
headquest, when you said your specs are the same, were you talking to me, the mac guy?
I'll look into what you guys are saying. However, it seems I can MAKE the glitches happen by starting and stopping my audio quickly, or adding more layers, or singing into the micropone as it runs through a heavy reverb.
I'll check it out...
I'll look into what you guys are saying. However, it seems I can MAKE the glitches happen by starting and stopping my audio quickly, or adding more layers, or singing into the micropone as it runs through a heavy reverb.
I'll check it out...
Live 6.0.3
1.25 GHZ Powermac G4
Mac OSX [10.2.8]
M-Audio Duo
Oxygen 8 Midi Keyboard
Digitech Vocal 300
Fetal Doppler
1.25 GHZ Powermac G4
Mac OSX [10.2.8]
M-Audio Duo
Oxygen 8 Midi Keyboard
Digitech Vocal 300
Fetal Doppler
Hi all,
I have disabled the "Windows ACPI-Compliant Control Method Battery" and "Microsoft AC Adapter" but I still have the problem when playing audio tracks in Ableton.
Its strange as I can play tracks in WMP no problem and even when the battery is unplugged...
Is anyone aware of a bios update for this issue?
I have an Acer Aspire 1804 laptop, 3.2ghz - 512mb Ram.
Thanks all,
Duffer
I have disabled the "Windows ACPI-Compliant Control Method Battery" and "Microsoft AC Adapter" but I still have the problem when playing audio tracks in Ableton.
Its strange as I can play tracks in WMP no problem and even when the battery is unplugged...
Is anyone aware of a bios update for this issue?
I have an Acer Aspire 1804 laptop, 3.2ghz - 512mb Ram.
Thanks all,
Duffer
Just like to add...
If you have a pc try to disable ACPI anyway if it is going to be used for music as this has long been known to cause problems for audio.If you can, install windows as standard pc mode without acpi for better audio performance.
But I am having more glitchy clicks on my mac than usual and have only noticed this since upgrading to live 5.
I am sure it is just an optimisation/teething issue that I hope gets sorted as ableton update and bugfix this version.
The strange thing is I know live 5 is more processor hungry than 4 but little clicks although not a major problem should not happen at such low processor use
I hope this has been mentioned on the bugs forum and ableton are aware of it. If it hasnt a new topic should be started
If you have a pc try to disable ACPI anyway if it is going to be used for music as this has long been known to cause problems for audio.If you can, install windows as standard pc mode without acpi for better audio performance.
But I am having more glitchy clicks on my mac than usual and have only noticed this since upgrading to live 5.
I am sure it is just an optimisation/teething issue that I hope gets sorted as ableton update and bugfix this version.
The strange thing is I know live 5 is more processor hungry than 4 but little clicks although not a major problem should not happen at such low processor use
I hope this has been mentioned on the bugs forum and ableton are aware of it. If it hasnt a new topic should be started
-
- Posts: 3603
- Joined: Tue Aug 09, 2005 8:26 pm