sounds like the end of the movie. voice over setting forth the epilogue, or something, to me. our protagonist stuck in the bowels of the machine, cursed to wander it's halls for all eternity or somesuch. they had a good run of it before succumbing to their fate. wistful. bittersweet.
@flavigula @noisevember it's a nice thought. I feel like that means crossing that chasm of mixing and mastering that I've avoided thus far. I'd be interested if you have any suggestions. Anywhere between 'this ffmpeg invocation gets you a good bit of the way there' to 'this person over here would do a good job and has reasonable rates'
I will in any case be collecting them here https://stubbornerr.com/ at some point. Sans mixing and/or mastering.
I first get all of the tracks sounding as good as they can possibly sound alone, then with one another. The second step involves using an eq on each track to make sure their frequencies do not interfere with each other. This sometimes requires a bit of automation.
Then I stick a limiter on the master bus. I make sure the hard limit is -0.3db. I fool with the controls until everything sounds a bit more "present" - ie, compressed, but not squashed.
I realise that much of that is subjective. Using one's ears is pretty fundamental for judging at that point, however. Good monitors or headphones also help, but that's another discussion.
Oh, I should have mentioned that before the limiter, and before mixing stuff together, make sure all the tracks have headroom. Don't let them peak over -6db. -9 is better. Most of mine are way lower.
Export everything to a wav. Then comes the fun part.
I've been using ffmpeg and its wonderful loudnorm apparatus.
Read about it here: http://k.ylo.ph/2016/04/04/loudnorm.html
I usually set the LRA higher than as he details in the page. My pieces usually have lots of dynamics. On the Noisevember stuff, I used between 12 & 16 for the LRA.
The command for the last one I normalised was this (according to zsh history):
ffmpeg -i ./hupu-li-txifela-hupol_session_20201202_1935.wav -af loudnorm=I=-16:dual_mono=true:TP=-1.5:LRA=13:measured_I=-15.08:measured_TP=-4.39:measured_LRA=17.30:measured_thresh=-26.36:offset=-0.42:linear=true:print_format=summary -ar 44.1k -sample_fmt s16 10-hupu-li-tzifela-hupol.wav
As also described in the page, I used the dual-pass mode to know whereabouts to set the measured options.
Thus I have the final 'product'.
SoNoMu (Sound Noise Music) is a mastodon instance for musicians, sound-artists, producers of any kind of aural noise, songwriters, bedroom producers, sonic manglers and algorave livecoders. -> more...