dhjdhjdhj

For the last two years or so, I have been using the following abstraction to convert MIDI note numbers and velocities (these arrive as a single list) into the format required by the [vst~] object. See the first patcher below.

Recently, after becoming aware that sprintf was not the most efficient object and also having learned some time ago about the [trigger] object, I decided it was time to remove [sprintf] as well as the use of the [bang] object, replacing the latter with [t b]. See the second patcher.

While rehearsing yesterday, my system occasionally (maybe once every few minutes or so) generated a random amount of crackle/noise, lasting about a second. It didn't matter whether I was actually playing a keyboard (sending events to the VST), it also happened when an audio player object (IRCAM SuperVP) was just playing audio (default pitch/speed etc).

Watching the CPU utilization % on the DSP status, I did not see any sudden jumps or anything that correlated with the noise happening.

After putting up with this for a couple of hours and recognizing that the ONLY change I had made was to the patcher for converting notes to midievent messages, I switched back to the old version of that patcher. After that, the noise/crackling did not happen again, at all.

I'm quite baffled by this and am hoping someone has some insights. I'm running Max 5.1.9 on a Macbook Pro with Mavericks and audio interface is two MOTU 828Mk3s, connected as an aggregate device. I have the latest MOTU drivers installed, etc.

Recently, after becoming aware that sprintf was not the most efficient object and also having learned some time ago about the [trigger] object, I decided it was time to remove [sprintf] as well as the use of the [bang] object, replacing the latter with [t b].  See the second patcher.

Watching the CPU utilization %  on the DSP status, I did not see any sudden jumps or anything that correlated with the noise happening.

How did my patcher change introduce random crackling?