I named a few in my reply. Google helps: But here you are some links:
https://process.audio/de/products/decibel
https://www.sonnox.com/news/new-sonnox-toolbox-listenhub
https://www.globaldelight.com/boom/
I cannot believe this situation, too. But it being a software problem I think that a switch to another processor architecture can bring up such bugs. Anyway what I do not understand is that this should have occurred in beta testing and should be a known issue. But there is silence from apple about this. Also, when it is realtively easy to mostly eliminate the problem by using a software driver "between" the software playback stream and the hardware outputs then why does apple not build that ability into core-audio or rosetta or wherever in the stream a tiny buffer seems to be neccessary. It would still be a workaround I suppose, but it could be communicated and used until one's OS works without any software that relies on rosetta. I also wrote to all audio product manufacturers that are important to me to please not sleep on the issue of translating their products to the apple native architecture, because some manufacturers consider to not upgrade "older" stuff like hardware drivers, interface software or VST Plugins. We need to get off rosetta as quickly as possible, but I feel some manufacturers struggle with providing updates as quickly as we would wish for.
I still hope apple can and will do something about it. In the dev-forums the talk goes that they know about this (how would they not) and might be able to provide a fix. But in the meantime we are at 11.5 so 5 updates did not fix the issue.
So my observations again:
The cause of the issue is rosetta. As long as ANY (even just ONE) process is active using rosetta it occurs.
The problem worsens immensely when using aggregated devices.
The wrappers (third party software core-audio-drivers) help a lot, but might not be perfect and might need to be reset from time to time (meaning that it might be necessary to manually reset the buffer size like once every 3-4 hours) depending on the task and load (throughput) that the driver has to handle.
Another Observation:
Those crackles occur especially on Youtube (or any other "thing" that comes out of a browser. It is ridiculously present in Spotify, but it. doesn't occur with iTunes/Music. So there seems to be some software that gets affected by this more and other software that isn't affected at all.
I never had any problem with crackling once I work in my DAW's (Cubase, mainly). Those are Applications that introduce a buffer themselves because they have to to be a able to handle the processing they apply to the audiostream) So this way rosetta also seems to have enough time to do it's translation for the processors to calculate and the audio stream passes through all soft and hardware stages without loosing samples. If crackling occurs (which also happens and always has, when the sample buffer size is set to small for the system to handle the processing) you just set a higher sample rate buffer size and things are fine again.
So I hope this helps y'all to understand and workaround. Pity we have to.
cheers,
Stefan