I’ve got an unusual problem with the audio engine freezing up in BFD 3.4.5 build 28 a64. I’ve narrowed it down but haven’t quite solved it. Here’s what’s happening – specifically load Andy Johns AFJ Kit 5. Play an existing or new pattern (internally or MIDI) that triggers the Mid Tom of AFJ Kit 5. At some arbitrary time (maybe after a few repeated measures or more), the BFD meters will momentarily pin and the audio engine will quit. Silent. Sometimes the audio engine will revive by hitting play again, sometimes it requires a BFD (app or plugin) restart. I experience the same crashing in BFD standalone as well as BFD audiounit in Logic 10.7.8., both on M1 Max Macbook Pro running 13.4 (running regular or rosetta). A few times it completely crashed Logic. Whenever I play the AFKJ 5 Mid Tom manually, in an internal pattern, or by midi, it will play okay for a few times (or more) but inevitable crashes BFDs audio engine. Take away or substitute Mid Tom of AFJ Kit 5 and it runs fine. And haven’t noticed this behavior with any other kits or drums.
I tried uninstalling, redownloading and reinstalling the Andy Johns kit. I also reinstalled BFD 3.4.5.28. Fiddled various BFD preferences. All without any luck. Any thoughts what’s wrong and how to fix? Could the AFJ Kit 5 Mid Tom be corrupted in the master installer?
(on edit: sorry I didn’t see there was a separate discussion section for extension pack bugs. Hope this will still find helpful peeps)
At some point in BFD’s history (loooooonnngggg history!) the loudness calculation that the software generates, generated false -inf values. This causes the BFD engine to explode miraculously.
Quick solution for now - delete the artictweaks.xml file from the drum/artic folder, and also delete the mirror of it from:
Then re-load the drum. You should see the browser quickly perform a scan and upgrade of the kitpiece. It will recalculate the values, and the crash will stop happening.
Quick day-later update. Oddly, the -inf entry came back on the artictweaks.xml file. Not sure why or how. I went in and manually edited the entry value to be somewhere between the adjacent ones. And that seems to be holding for now.
I tried this method and it stops the meters pinning, but the articulation no longer plays through the direct mic at the offending velocity level. The adjacent velocities are fine. Looks like this (skip to a minute in):
JDrums… Interesting. I don’t have BFD 1 drums installed so I couldn’t duplicate your situation here. My problematic drum (AFJ Kit 5 Mid Tom) seems to be working fine now with the manual hack to the BFD2VelLayerTweak loudness value (that had been “-inf”). That particular drum only has one articulation so I guess troubleshooting was a little easier (?) Kudos to BFD_Drew.
