(I started writing this post the other week, got distracted, and forgot about it…)
I noticed SWAM Solo Brass is on the way, which naturally caught my attention, especially since I’m (hopefully) getting an Osmose next summer…
However, much like my impression with modeled instruments so far, it’s all… very inconsistent. The SWAM Star Wars teaser is honestly terrible. The horns and tubas demo sounds ok, the tone of the trombones is not very nice at all (needs big hall reverb?), but the trumpet demo is pretty nice. SWAM strings sound terrible to me; the body resonances sound like cheap student instruments, and the vibrato is awful.
However, the main issue is that while these instruments rely very heavily on expressive playing, most of the demos are just someone jamming on a traditional keyboard + wheel setup, and by the sound of it, most orchestral demos and the like are recorded like that as well. That’s never going to sound anything like the real thing, or even a decent sample library based mockup.
Another issue is that modeled instruments put extreme demands on mix and effects, even more so than dry/studio sample libraries, which still usually have some “room color” in them.
So… I guess my conclusion is that while this is very interesting stuff, there are some issues preventing it from taking off:
- Demos need to be properly arranged, with carefully recorded and edited automation of all available parameters.
- Demos need to be properly mixed, to match the sound of instruments recorded in great sounding halls and the like.
- More seriously, for these instruments to ever become more than an odd exception that only a few random composers/musicians can use effectively, we need to actually figure out how to play them properly. Controllers, technique, methods, theory - just like for the original instruments. Either that, and/or AI based players that “know” how to play expressively, shifting the composer/arranger’s job from playing towards directing.