ThumbJam Forum

(I looked quickly through the list here, and didn't see any other bug reports of this -- sorry if I missed one and should be filing this notice there.)

Every once in a while (it's happened 2 or 3 times since the upgrade, in several hundred sessions I've recorded since the upgrade, so not very often) ... I record a session that comes out at hyper-speed and hyper-pitch. It looks to me like it's an octave above what it should be, with the concomitant speed increase as well.

I've attached a brief sample of a session that this has happened with, in the version that TJ gave me and then in a version that I slowed down in my sound processing program on my Mac -- I reduced the pitch by 12 semitones and the speed by 50%, which puts the pitch at exactly what it should be, and the speed at probably what it should be (there was no metronome running, so I don't know for certain).

I didn't have any other apps running in the background when I did this, except maybe the iOS Settings ... but I don't know if some other app that I had launched while TJ was running did something, and don't remember which apps were running earlier. So if it's a conflict with another app, I'm afraid I don't know what app. This weirdness happens so seldom that it'll be difficult for me to troubleshoot and identify the culprit, but I can try.

Has anyone else seen/reported this? Any ideas? It's really weird.

PS Here's a loop file recorded today also, with the hyper-pitch/speed thing happening (in case the information from the XML files helps). It seems that everything I've recorded after a certain point today is affected, loops and sessions. I'm thinking the next time I actually quit TJ and restart it, things will be back to normal.

Yes, this has been reported before. What it appears happens is another application switches the sample rate, and when TJ comes back it doesn't reset it back to the normal 44.1kHz (which is one bug), and then it incorrectly marks the recorded loop files as 44.1khz no matter what. The other problem is that even if the files were marked correctly at another sample rate, the code that plays back the loops automatically assumes they match the current samplerate (it doesn't resample them). All of these things together result in serious confusion when other apps hijack the samplerate, either before or after TJ is running.

Okay, interesting, thanks for the explanation.

So it sounds like I need to make sure TJ isn't running in the background, then quit other apps, then launch TJ. Is that right? (I'm definitely NOT complaining, I just want to get my procedures down. Otherwise, whenever this problem does occur I don't know it's occurring until after I've spend a bit of time recording something.)

Also, is it always going to be 12 semitones off? Or will it depend on the other app? Is there a mathematical formula I can check? (i.e., it'll always be a multiple of 12 semitones off, or something like that) So I can easily correct the files if I get stung?

Thanks!!

It depends on what the other app set the samplerate to relative to TJ's 44.1kHz. In your case, someone set it to 22.05 kHz, so it ended up twice as fast (an octave).

I was hoping to release the bugfix TJ this month, but it's looking like it might be early Feb.

Thanks, that helps a lot (considering the common sample rates).