I’ve had this problem since I installed Cubase 12, and it has continued in Cubase 13. While playing back my project, the sound will briefly crackle and then drop out altogether. Also all the meters in the mixer stop showing any activity, even though the playhead/cursor continues to move.
The problem persists until I bring up Studio Setup->Audio System and press the reset button or unplug/replug the interface. I’ll hear another crackle, maybe a brief moment of whatever was still in the buffer, and then sound will be normal until it happens again. It might happen 5 times in an afternoon of work, or not at all. Project complexity seems irrelevant - it can happen on a project with one track. It also might happen during recording or just playback while I’m doing nothing with the mouse or keyboard. I can continue to edit the project to my heart’s content, just no sound!
When I installed Cubase I did the OS optimizations recommended, and I avoid use of other applications while running Cubase. Cubase almost never crashes on this machine, and is otherwise well-behaved. I’ve checked the OS logs and looked for Cubase dump files. Nada.
Here are all the details you’ll undoubtedly ask for:
Cubase 13.0.41 Pro Build 256
Lenovo Ideapad AMD Ryzen 5 6600 HS
Windows 11 Home 23H2 22361.4037
Focusrite Scarlet 2i2 (3rd Gen) Firmware 1605 build 4.119.13
My Buffer size has been set at 192, which results in a round-trip latency of 22.0 mS.
When I significantly increase it to 512, the latency increases to 50.6 mS, enough to be annoying while recording audio.
But since I’m in mastering mode right now, I’ll leave it at 512 and see if that helps the no sound problem, then maybe decrease it only while recording audio.
Sounds like the interface losing connection…make sure you have disabled all usb power saving (device manager, open each usb hub properties and untick allow this device to turn off to save power) also disable usb selective suspend.
The interface losing connection was my first thought, until I noticed that the meters on the Cubase mixer also stop registering any signal, even though the playhead keeps moving over plenty of audio data.
But yes I’d already also disabled any and all power-saving options from the usb hubs, controllers, etc I could find. There are so many “devices” virtual and real on a Windows system that I won’t pretend I’ve not missed something along the way.
Pretty sure I have seen similar behaviour before when the interface lost connection. Check carefully those power management settings and also consider switching any internal audio to another device to make sure it’s not receiving any sample rate change requests…or at least disable all system sounds and any startup programs that are non essential.
Great idea! Just pointed Windows audio at the internal speakers - it had been going to my Focusrite interface - and turned off all Windows system sounds. We’ll see how that works. Thanks for the suggestions.
For what it is worth - I have Focusrite Scarlet 2i2 (4th Gen) The audio output also goes dead for no apparent reason several times a day - maybe after 5mins or maybe 10-15 hours or more. I have spent an inordinate amount of time with Focusrite (and also Steinberg and even LatencyMon developers) trying to solve this problem and have given up. Other audio interfaces I have don’t do this…
There have been several threads on the internet - e.g. on Reddit etc - that also experience this with Focusrite. I have pointed this out to Focusrite but they consistently ignore this fact.
As with everyone else - they all expect LatencyMon to point to the problem. However in this case LatencyMon is unable to help - provided you have already got your latency running at acceptable levels - e.g under 100 microseconds. Their expectation is that when the audio drops there should be an obvious culprit shown in LatencyMon at that instant - but there is not. It is the wrong tool for this particular job.
I have given up trying to fix it since it is only a minor inconvenience that is quickly fixed by resetting the interface in Focusrite Control 2…
Here are some threads I found in May when I was researching this… There may be a solution in one of them now - who knows?
