You are not logged in. Please login or register.
RME User Forum → FireWire & USB series → Pops and Crackles (SOLVED)
I 've been experiencing pops and crackles with my Babyface Pro FS (Win 10) whenever I Record-Arm a vst instrument in Reaper. The buffer size is set to 256. The crackling happens even on light projects, where the CPU and R(eal) T(ime) CPU are both hanging around 10%.
i7 11700Κ, 3,6 Ghz
Gigabyte B560 HD3 (rev. 1.0)
KINGSTON HYPERX FURY RGB DDR4 32GB 170€
SAMSUNG 980 1TB 127€
Windows 10 21Η2 (ΟS Build 19044.2364)
Did you optimize your Windows PC for audio?
How long is your USB2 cable? Did you check in the RME driver settings, whether you have CRC errors?
Did you try every USB3 and USB2 port already, some might work better than others?
Which Windows energy profile is in use? High or Ultimate Performance?
Did you disable energy saving in the BIOS? You have to either disable C-States or to set it to C0/C1 and to disable T-/P-states.
Did you disable energy saving for USB and PCIe in the extended energy saving settings?
If this all doesn't work. You should measure DPC latencies whether you have a bad driver blocking CPU cores for audio related processes for too long.
Audio performance related threads:
https://forum.rme-audio.de/viewtopic.ph … 36#p196936
Collection of threads:
https://forum.rme-audio.de/viewtopic.ph … 04#p186404
I 've been experiencing pops and crackles with my Babyface Pro FS (Win 10) whenever I Record-Arm a vst instrument in Reaper. The buffer size is set to 256. The crackling happens even on light projects, where the CPU and R(eal) T(ime) CPU are both hanging around 10%.
i7 11700Κ, 3,6 Ghz
Gigabyte B560 HD3 (rev. 1.0)
KINGSTON HYPERX FURY RGB DDR4 32GB 170€
SAMSUNG 980 1TB 127€
Windows 10 21Η2 (ΟS Build 19044.2364)
Does it crackle only in this specific case?
Does it stop when you disarm?
Did you try to record the vst instrument and listen to the result?
Could it be a sample rate mismatch somewhere? Or a mismatch between Midi and audio? What sample rate do you have set for the project? And what are the clock settings for the Babyface?
Does it crackle only in this specific case?
It also often crackles when I stop the playback or when I move the playhead around without stopping. This happens even with just one audio track, without any processing.
Does it stop when you disarm?
Yes it does and the RT CPU meter goes back to 1%
Did you try to record the vst instrument and listen to the result?
It sounds fine when played back with the Record-Arm disabled.
Could it be a sample rate mismatch somewhere? Or a mismatch between Midi and audio? What sample rate do you have set for the project? And what are the clock settings for the Babyface?
No sample rate mismatch, Reaper sets its sample rate automatically the same as Babyface (I disabled the option to set its own sample rate within the project). The sample rate is 44.1KHz. Clock Source is set to Internal in the the Fireface application.
What do you mean by "mismatch between Midi and audio"? All I can say to this, is that the crackling happens only when record-arming tracks that have vst istruments on them, not when record-arming audio tracks.
Did you optimize your Windows PC for audio?
Yes I did.
How long is your USB2 cable?
It's 2 meters
Did you check in the RME driver settings, whether you have CRC errors?
It shows Errors 0/0
Did you try every USB3 and USB2 port already, some might work better than others?
I tried, it made no difference. I only have one USB Root Hub in my mainboard. The RME shares it with a wired keyboard and mouse.
Which Windows energy profile is in use? High or Ultimate Performance?
High Performance is used. I see no option for "Ultimate Performance"
Did you disable energy saving in the BIOS? You have to either disable C-States or to set it to C0/C1 and to disable T-/P-states.
I just did it.
Did you disable energy saving for USB and PCIe in the extended energy saving settings?
I just disabled the "USB Selective suspend" setting. The "Link State Power Management" was aready off.
In some rare situations I had alerts about a synchronisation problem between midi and audio on my system. I don’t know what exactly the issue was. Anyway, it’s not really comparable because it was with Logic on Mac…
Another question: does it appear related to the same vst instrument always?
Disabling can already be a game changer.
The power profile ultimate performance is hidden, google to get the command for the dos box to enable.
Its an alternative to High Performance, but I didn't investigate whats really the difference, both should do.
If I remember correctly it is a power profile from a Windows for Workstation product.
Performoning Latencymon measurements you should see a difference before / after disabling of C-States.
What also helps: to enable TURBO and disable C1N
EIST you can keep enabled to be able to control the clock by Windows energy profiles.
CLOCK_SPREADSPECTRUM you may disable to get a fix clock without deviations.
If you have a nVidia Card you can try to disable energy saving on the nVidia card with the powermizer tool.
Check the links that I gave to you, it has been written already a lot to this topic, there is nothing to add and its a lot of work to re-type everything again and again.
If all fails try to decouple recording interface from rest of USB infrastructure, the Sonnet PCIe cards with FL1100 chips work nicely. Mainly for USB3 but they are also well in USB2 compat mode.
In some rare situations I had alerts about a synchronisation problem between midi and audio on my system. I don’t know what exactly the issue was. Anyway, it’s not really comparable because it was with Logic on Mac…
Another question: does it appear related to the same vst instrument always?
No. it could do it with any vst instrument
Disabling can already be a game changer.
The power profile ultimate performance is hidden, google to get the command for the dos box to enable.
Its an alternative to High Performance, but I didn't investigate whats really the difference, both should do.
If I remember correctly it is a power profile from a Windows for Workstation product.Performoning Latencymon measurements you should see a difference before / after disabling of C-States.
What also helps: to enable TURBO and disable C1N
EIST you can keep enabled to be able to control the clock by Windows energy profiles.
CLOCK_SPREADSPECTRUM you may disable to get a fix clock without deviations.
If you have a nVidia Card you can try to disable energy saving on the nVidia card with the powermizer tool.
Check the links that I gave to you, it has been written already a lot to this topic, there is nothing to add and its a lot of work to re-type everything again and again.
If all fails try to decouple recording interface from rest of USB infrastructure, the Sonnet PCIe cards with FL1100 chips work nicely. Mainly for USB3 but they are also well in USB2 compat mode.
Thank you so much!
Your PC is surely more powerful from CPU perspective compared to mine from 2014.
Compare PassMark Mixed Benchmark numbers and single thread performance:
Your CPU
https://www.cpubenchmark.net/cpu.php?cp … mp;id=3896 Passmark: 24667 SThr: 3422
My CPU of that time
https://www.cpubenchmark.net/cpu.php?cp … mp;id=2838 Passmark: 11479 SThr: 2394
So your CPU is definitively more powerful, so if you have configured all properly, have no blocking drivers etc, then you should be able to get the same performance like me
https://www.tonstudio-forum.de/blog/Ent … cks-de-en/
And with a BBF Pro you do not have that many audio channels to transfer over USB compared to the UFX+ / UFX III or RayDAT...
oli77sch wrote:In some rare situations I had alerts about a synchronisation problem between midi and audio on my system. I don’t know what exactly the issue was. Anyway, it’s not really comparable because it was with Logic on Mac…
Another question: does it appear related to the same vst instrument always?No. it could do it with any vst instrument
Further questions:
- do you have another DAW / VST host application? If yes, does the issue appear also there?
- does the issue appear also when you use another audio hardware (internal soundcard)?
- how do you play the software instrument? Midi keyboard connected to the Babyface Midi input? Or does it have its own Midi-over-USB connection?
To me it looks more like an 'in-application' issue. Only for my understanding: what kind of recording are we talking about? Internal audio from the VST instrument output or Midi recording? I don’t know what possibilities you have with Reaper. Maybe it’s different from Logic where I can create an audio track and select the VST instrument as input. So I have an audio recording (WAV) at the end. But I also can simply record the Midi data when I arm the channel with the virtual instrument and hit 'rec'.
- do you have another DAW / VST host application? If yes, does the issue appear also there?
.
No, Studio One projects don't crackle UNLESS a Reaper project with a vst instrument record-armed is open in the background.
- does the issue appear also when you use another audio hardware (internal soundcard)?
.
I have a Focusrite soundcard and the issue does not appear there
- how do you play the software instrument? Midi keyboard connected to the Babyface Midi input? Or does it have its own Midi-over-USB connection?
.
Midi keyboard connected to the Babyface Midi input
To me it looks more like an 'in-application' issue.
It appears like an "in-application" problem (Studio One doesn't crackle) but at the same time like an "interface-related" (Focusrite doesn't crackle).
Only for my understanding: what kind of recording are we talking about? Internal audio from the VST instrument output or Midi recording? I don’t know what possibilities you have with Reaper. Maybe it’s different from Logic where I can create an audio track and select the VST instrument as input. So I have an audio recording (WAV) at the end. But I also can simply record the Midi data when I arm the channel with the virtual instrument and hit 'rec'.
We are talking about Internal audio from the VST instrument. Reaper handles VST instruments as insert plug-ins. I don't think there's a workaround to this. And even if there was, it wouldn't be very practical to use workarounds for something so basic that I do so often.
Your PC is surely more powerful from CPU perspective compared to mine from 2014.
......
I have a Focusrite interface and it doesn't exhibit the same problem. If the crackling was caused by a blocking driver, wouldn't that same blocking driver cause crackles with the Focusrite?
I can’t imagine what is going on at the moment you arm the track. If I understand it correctly, Reaper does record the Midi data only, not the audio itself from the virtual instrument. So I can’t see why it makes such a difference between armed and unarmed track. It’s a mystery to me…
Maybe you can try other sample rates and other buffer sizes once, just to try?
Differences between RME and Focusrite
- interfaces have a different number of channels, all channels will be transferred over USB, no matter whether used or not
so we talk about a higher Bandwidth of data that needs to be transferred instantly
Focusrite might use
- higher safety buffers in their drivers but then also creating higher RTL (round trip latency)
. other USB transfer modes that might work better with your USB infrastructure (HW/driver)
I am uncertain whether I understood your description of the error well enough.
I am not sure, whether you have the typical "audio loss" if there are problems with Windows settings / driver / DPC latencies
or whether its a problem with application(s).
Can you make a video to illustrate the issue better?
Does reaper have a function like asio guard in Cubase? When this is enabled in Cubase then arming a vst track does make the cpu jump higher as it drops the latency down for that track. I’ve no idea if reaper has a similar feature though and it still sounds quite an excessive jump in your case.
Does reaper have a function like asio guard in Cubase? When this is enabled in Cubase then arming a vst track does make the cpu jump higher as it drops the latency down for that track. I’ve no idea if reaper has a similar feature though and it still sounds quite an excessive jump in your case.
Yes, I think Reaper has the "Render-Ahead" option in the "Anticipative FX Processing" menu. I will try to fiddle around with that.
Noone would have thought (and noone did) that a bad plugin could cause such symptoms.
The problem was caused by Convology XT. I liked it so much that I always used at least a couple of instances on every project. That's why the problem was always there, leading me to conclude that it was systemic.
Yesterday I started fiddling around with things just to see what would happen.
Disabling the channels with Convology on them allowed for a smooth recording of MIDI and audio with very low buffer sizes.
So there you go.
I live and I learn.
Hope this helps someone someday....
I remember Ozone is similar, required me to use ASIO buffer sizes of 256 or even 512.
Therefore, I stopped using/updating it. Good that you found it.
I 've been experiencing pops and crackles with my Babyface Pro FS (Win 10) whenever I Record-Arm a vst instrument in Reaper. The buffer size is set to 256. The crackling happens even on light projects, where the CPU and R(eal) T(ime) CPU are both hanging around 10%.
Well, you wrote "light projects". Multiple use of heavy reverbs is not light project.
Anyway good that you have found the source of your problems.
The problem was caused by Convology XT. I liked it so much that I always used at least a couple of instances on every project.
Try to change your work.
Use Convology XT only one time in the project as send-effect.
One empty track with only Convology loaded as insert effect.
Send certain amounts of the other tracks to this send-effect-track.
Try to change your work.
Use Convology XT only one time in the project as send-effect.
One empty track with only Convology loaded as insert effect.
Send certain amounts of the other tracks to this send-effect-track.
Even with this setup Convology creates a problem. I will be replacing it with Valhalla or something...
RME User Forum → FireWire & USB series → Pops and Crackles (SOLVED)
Powered by PunBB, supported by Informer Technologies, Inc.