![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
music software geeking
Mar. 21st, 2002 11:28 pmI'm getting to that disheartening point in learning this software whereat you start to suspect that the annoying roadblocks you hit are the program's fault, not yours. For example, all audio routing seems to be done through buses. So if you want to run part of a track through a downward pitch shift into a reverb, you need to aux-send the track to a bus, and insert the shifter and reverb in that bus. But there are only eight buses. If you buy the "platinum" version, then you get sixteen.
Fine, so the version I've got is limited for product-line marketing reasons. What offends me deeply is that the top-of-the-line version doesn't have arbitrarily many buses. That's so dumb. This is a computer, you dolts. Stop dragging your knuckles and get with the dynamic allocation.
I also wish sequencer writers wouldn't put MIDI into the core of their design. Work with your own fully-capable temporal data format, and translate it down to MIDI as necessary. Then you wouldn't have this braindeadness where mix automation -- stuff that stays purely inside the sequencer, never goes over a MIDI wire -- is represented as MIDI, one MIDI channel per mixer track, so you can only automate audio tracks 1-16 on one sequencer track. 17-24 go on a separate track.
And I'd like to see a sequencer oriented more towards chunks and less towards tracks, but I know I'm a freak on that point. Logic's "track of folders of sequences" model is actually pretty good as far as it goes.
Anyway, the learning example I'm doing is a jazzed-up reading of a Kathleen Raine poem, "There shall be no more sea". Twelve takes total, close-miked to cut down fan noise. Used snippets from four of the twelve, level-matched, filtered at points where I huffed into the mic, bass-shelved down to cancel proximity effect, compressed gently on general principles, with a little small-room reverb to add back ambience -- and now it just sounds like I recorded a single good take in the first place. Audio production is a funny business.
Fine, so the version I've got is limited for product-line marketing reasons. What offends me deeply is that the top-of-the-line version doesn't have arbitrarily many buses. That's so dumb. This is a computer, you dolts. Stop dragging your knuckles and get with the dynamic allocation.
I also wish sequencer writers wouldn't put MIDI into the core of their design. Work with your own fully-capable temporal data format, and translate it down to MIDI as necessary. Then you wouldn't have this braindeadness where mix automation -- stuff that stays purely inside the sequencer, never goes over a MIDI wire -- is represented as MIDI, one MIDI channel per mixer track, so you can only automate audio tracks 1-16 on one sequencer track. 17-24 go on a separate track.
And I'd like to see a sequencer oriented more towards chunks and less towards tracks, but I know I'm a freak on that point. Logic's "track of folders of sequences" model is actually pretty good as far as it goes.
Anyway, the learning example I'm doing is a jazzed-up reading of a Kathleen Raine poem, "There shall be no more sea". Twelve takes total, close-miked to cut down fan noise. Used snippets from four of the twelve, level-matched, filtered at points where I huffed into the mic, bass-shelved down to cancel proximity effect, compressed gently on general principles, with a little small-room reverb to add back ambience -- and now it just sounds like I recorded a single good take in the first place. Audio production is a funny business.
