why does Ableton only using one thread core? CPU overloading.
- KVRian
- 773 posts since 23 Apr, 2002 from audio/hamburg/germany/earth/space/unkown!
Just a friendly reminder that not every (audio) routing can be processed in parallel!
Though normally audio happens a lot in parallel, (e.g. different tracks, etc.).
But for example, a huge mastering chain can not be processed in parallel, so even distributing the workload on different cpus will not speed up the processing, it will just free up more resources for other things.
Anytime one processor has to wait for another processors output, multithreading will be mostly useless in terms of performance gain.
Though normally audio happens a lot in parallel, (e.g. different tracks, etc.).
But for example, a huge mastering chain can not be processed in parallel, so even distributing the workload on different cpus will not speed up the processing, it will just free up more resources for other things.
Anytime one processor has to wait for another processors output, multithreading will be mostly useless in terms of performance gain.
-
- KVRist
- Topic Starter
- 438 posts since 21 Feb, 2006
i use m-audio delta 1010lt PCI, i had to set the sample rate at 88200 and buffer at 1026 so it wouldn't make that awful cpu static processing sound. the 256 buffer and 44.1 Khz wasnt able to run without clipping. Its so hard to pinpoint problem. Im troubleshooting other DAWS at the moment.EnGee wrote:I have a similar CPU Rayzen 5 1600X (it is a little bit faster than the 1600 but otherwise is mostly the same). I have Live 9 Intro but I didn't notice any problem. Anyway I don't have any izotope product (they are CPU hungry products, aren't they?)
Meanwhile try to play with latency till you can no more see/hear any maxing to your CPU. Also what is your audio interface and what ASIO you are using? Sample rate? Those are additional info helps. For me, I use Presonus iTwo Asio driver with 256 buffer and 44.1 Khz sample rate.
Anyway, it is not a miracle CPU! It is better than the i5 -4570 I used, but not like a day and night! So, maybe you are expecting too much. I can max this cpu easily with just one instance of Korg Arp Odyssey (I keep repeating this because this is the almost the only high cpu demanding plug I have for now), if I choose a very heavy preset (with 16 unison ..etc). Can you try another plug that behaves similar between the DAWs?
There is also differences between the DAWs in relation of CPU usage. Also, some plugs have 'bugs' with a specific DAWs. I suggest to write the steps clearly to produce the cpu maxing, so we can try to duplicate and give you faster/more accurate feedback.
my other computer was a i-7 3770 which is about 5 years old and in some ways runs better, which is very infuriating.
-
- KVRist
- Topic Starter
- 438 posts since 21 Feb, 2006
yes i saw this before i posted this thread.fluffy_little_something wrote:https://help.ableton.com/hc/en-us/artic ... e-CPU-load
thanks
-
- KVRist
- Topic Starter
- 438 posts since 21 Feb, 2006
sounds like i made the wrong choice in processors? what do you suggest?dasdeck wrote:Just a friendly reminder that not every (audio) routing can be processed in parallel!
Though normally audio happens a lot in parallel, (e.g. different tracks, etc.).
But for example, a huge mastering chain can not be processed in parallel, so even distributing the workload on different cpus will not speed up the processing, it will just free up more resources for other things.
Anytime one processor has to wait for another processors output, multithreading will be mostly useless in terms of performance gain.
- KVRian
- 773 posts since 23 Apr, 2002 from audio/hamburg/germany/earth/space/unkown!
I think you misunderstand. Not the CPU, the audioprocessor e.g. the plugin.zoogoo wrote:sounds like i made the wrong choice in processors? what do you suggest?dasdeck wrote:Just a friendly reminder that not every (audio) routing can be processed in parallel!
Though normally audio happens a lot in parallel, (e.g. different tracks, etc.).
But for example, a huge mastering chain can not be processed in parallel, so even distributing the workload on different cpus will not speed up the processing, it will just free up more resources for other things.
Anytime one processor has to wait for another processors output, multithreading will be mostly useless in terms of performance gain.
-
- KVRist
- Topic Starter
- 438 posts since 21 Feb, 2006
im not sure what this means, but it says that the DAW isnt the problem, its the computer.Reefius wrote:You could try running LatencyMon to check how well computer can handle realtime processing.
its a new build.
You do not have the required permissions to view the files attached to this post.
- KVRAF
- 24452 posts since 7 Jan, 2009 from Croatia
Using 88200 sample rate will put more strain on the CPU than 44100 (which is why you also had to increase the buffer size).
I'm also thinking M-Audio's ASIO drivers could be a bit of a culprit here... they're not the best.
Also looking at LatencyMon results, your CPU is being throttled, so yes, set power options to High Performance.
I'm also thinking M-Audio's ASIO drivers could be a bit of a culprit here... they're not the best.
Also looking at LatencyMon results, your CPU is being throttled, so yes, set power options to High Performance.
-
- KVRist
- Topic Starter
- 438 posts since 21 Feb, 2006
ok so i set it to high performance which i had to do manually through the gpedit.msc which was slightly annoying. But success. One thing i noticed right away was the cpu bar dropped dramatically in ableton, also as far as i can tell the clipping is gone.EvilDragon wrote:Using 88200 sample rate will put more strain on the CPU than 44100 (which is why you also had to increase the buffer size).
I'm also thinking M-Audio's ASIO drivers could be a bit of a culprit here... they're not the best.
Also looking at LatencyMon results, your CPU is being throttled, so yes, set power options to High Performance.
yes i was ableton to set the audio back to 44100 and that clipping wasnt as bad but still there.
As far as latencymon, looks like the the DPC levels are not as high, but still get the same warning.
the problem is said with the m-audio driver. Is there something else i should be using instead? Ive tried ASIO but had problems with that as well.
im definitely getting closer to fixing it. it is better, but still not there.
any more suggestions would be greatly appreciated.
- KVRAF
- 24452 posts since 7 Jan, 2009 from Croatia
No, you didn't need to set it via gpedit at all... There are Power Options in Control Panel...
ASIO should be the audio driver to use. If you still have problems with that, it's time for a better audio interface!
ASIO should be the audio driver to use. If you still have problems with that, it's time for a better audio interface!
-
- KVRAF
- 9150 posts since 7 Oct, 2005
This is a good progress. On the other hand that sound card is very old! I remember it! Maybe you can test Asio4All driver for the internal sound card and see if there is a difference.
i7-3770 is a very good cpu still! If I were you, I would have chosen Ryzen 7 1700 or higher to feel the jump in performance. Anyway, the advantage of Ryzen 5 1600 is that it has 6 cores with 12 threads, so it is a good 'average' CPU.
Cubase has its own power options profile that uses it while running Cubase. The system returns to its original profile (in my case Balanced) after quitting Cubase (it is still an option). However, for other DAWs it is better to set the power options to high performance manually.
i7-3770 is a very good cpu still! If I were you, I would have chosen Ryzen 7 1700 or higher to feel the jump in performance. Anyway, the advantage of Ryzen 5 1600 is that it has 6 cores with 12 threads, so it is a good 'average' CPU.
Cubase has its own power options profile that uses it while running Cubase. The system returns to its original profile (in my case Balanced) after quitting Cubase (it is still an option). However, for other DAWs it is better to set the power options to high performance manually.
Using: Cubase Pro 15, Bitwig 5, Tascam US-4x4HR, MODX6, DM12D, LaunchKey 49, Yamaha guitar(Pacifica 612v) and bass (BB234) and some virtual instruments and synths.
- KVRAF
- 24452 posts since 7 Jan, 2009 from Croatia
A4A is rarely a better choice than manufacturer's own ASIO drivers. But with M-Audio, anything is possible 
For a desktop computer, especially a DAW, it's best to have power options set to High Performance ALYWAYS.
For a desktop computer, especially a DAW, it's best to have power options set to High Performance ALYWAYS.
-
- KVRist
- Topic Starter
- 438 posts since 21 Feb, 2006
yes i have used ASIOall , in fact i made another post asking for an alternative, since it was having issues. its much worse then the m-audio driverEvilDragon wrote:A4A is rarely a better choice than manufacturer's own ASIO drivers. But with M-Audio, anything is possible
For a desktop computer, especially a DAW, it's best to have power options set to High Performance ALYWAYS.
the problem is the communication with the card and computer. as you can see in the latencymon The DPC is to high.
my guess i need to change some driver on the PCI itself.
idk
- KVRAF
- 24452 posts since 7 Jan, 2009 from Croatia
No, you just need a better audio interface methinks. PCI drivers won't change things here...
By the way... That is a PCI audio interface, but I'm not sure if there are any mobos for recent CPUs out there that have PCI slots? It's all PCIe these days... If you're using some sort of PCI-to-PCIe adapter, that could explain the bad behavior you're getting here.
Time to move on, I would say, and get a proper USB or Thunderbolt external audio interface.
By the way... That is a PCI audio interface, but I'm not sure if there are any mobos for recent CPUs out there that have PCI slots? It's all PCIe these days... If you're using some sort of PCI-to-PCIe adapter, that could explain the bad behavior you're getting here.
Time to move on, I would say, and get a proper USB or Thunderbolt external audio interface.
-
- KVRist
- Topic Starter
- 438 posts since 21 Feb, 2006
nope its an MSI am4 with 2 standard pci slots. no adapter.EvilDragon wrote:No, you just need a better audio interface methinks. PCI drivers won't change things here...
By the way... That is a PCI audio interface, but I'm not sure if there are any mobos for recent CPUs out there that have PCI slots? It's all PCIe these days... If you're using some sort of PCI-to-PCIe adapter, that could explain the bad behavior you're getting here.
Time to move on, I would say, and get a proper USB or Thunderbolt external audio interface.