tedium

[eirias]

Something it is not possible to overemphasize about research: it is goddamn tedious. I set up an experiment over the weekend in remarkable time for me, just a few hours, including stimulus processing and script tweaking. There were some timing errors, but I figured it'd boil down to a memory problem or some other one-line fix. No: as it happens, it's the result of a MIDI->WAV conversion error that I've previously encountered, but had forgotten about. Now I have to go change the tempo of 168 MIDI files and reconvert them to WAV, all by hand; then I have to rack my brain about how I remove that variable silence using one of the editors at my command, preferably in some batchy way. At least the file transfer from the edit machine to the data collection machine no longer takes hours.

Comments

[kirin]

Ick. I don't suppose much of this is in a scriptable environment? I suppose the trick probably is having a MIDI editor with robust script support. I wouldn't be surprised if one existed, but I guess I wouldn't be surprised if it didn't either.

[eirias.livejournal.com]

It's not AFAIK, that's the main problem. I'm sure that there is such a thing, and it might even be free given that MIDI people are often nerdy, but if that were true it'd be pretty unintuitive, not at all appropriate for this project. Alternatively, it's just possible that there's one of the shiny polished studios out there that does something like this, but chances are we'd pay through the nose, and getting my advisor to cough up money for software is ... well ... let's not talk about that.

[harleybitch.livejournal.com]

Even with my student working on what i thought was tedious (pushing a button any time the bug does something we like), I ended up with what seems like more tedious stuff-- burning backups, staring at slides, writing, histology. Its weird when you look around and everything is tedious, however, if the experiment works, despite tedium, it is going to be freakin awesome!