[Csnd] learning sound design
| Date | 2026-10-20 11:47 |
| From | Michael Mossey |
| Subject | [Csnd] learning sound design |
Hi, I'm inspired by Wendy Carlos's synthesis (such as Switched On Bach and her later stuff) and I've dabbled in electronic realizations of Bach, but I feel totally lost. I have some composition experience for acoustic instruments, mostly piano, and I'm largely self-taught in everything I know about music (I know a lot of 18th century theory and some 20th century theory), so I thought I would just dabble in synthesizers. I've used CSound and others. I'm also a programmer so I'm not afraid of CSound. But generally whatever I try to create is either simplistic or muddy or just not interesting like so many of the CSound compositions that I can find online or on this list or Carlos's music. Dabbling isn't working. I feel like I need some basic course or reading material to know how to approach sound design (by that I mean combining electronic sounds in interesting ways as part of realizing compositions like Bach or otherwise originating from acoustic music). Maybe I need a guided project. It could be CSound I use to learn, but I also like some VSTi synthesizer plugins, so I'm interested in learning sound design in a general sense. Any help would be appreciated. Mike |
| Date | 2026-10-24 18:49 |
| From | "Jeanette C." <000015cdd0ffa6cd-dmarc-request@LISTSERV.HEANET.IE> |
| Subject | Re: [Csnd] learning sound design |
Hello Mike,
not sure, if I'm the best person to reply, but I'm the first. :)
Wendy Carlos, in particular, was often talking about timbre and brightness. So
she would layer sounds. You could do that with expression. Csound understands
MIDI, so you can drive it live by MIDI input or first create a MIDI
performance and then run that through Csound as a MIDI file. Typical things
that made sense to me - and to Wendy: using filters, envelope parameters and
layering. Trigger an instrument only above a certain velocity to bring in
extra brightness. Program a filter cutoff to follow velocity. Or try the same
with attack values for amp and filter envelopes. Low velocity, longer attack.
If you want really expressive performances use other MIDI CCs like the
modwheel or the expression pedal or if you have a keyboard with MPE use that.
Tempo curves are also a general way to improve the expressiveness of a
performance. Or you can play one voice and then pick out only certain notes
and double them on another instrument.
So basically, treat your synthesizers and compositions like classic
instruments. Give them articulation expression, tempo and timing and, of
course space. A good reverb and placement in a room also help to liven things
up.
There might be actual reading material, but I have rarely had easy access to
those. You question is one of those big ones and the answer so oftren is: it
depends. It depends on your exact style, ideas and needs.
As for sound design itself: Csound is good and allows you a lot of complexity.
Though for your first steps, maybe a monolythic synth app or plugin might be
more helpful. Would you like to stick to (virtual) analogue synthesis or
include more modern styles too? There are many good videos and articles about
analogue sound design, increasing amounts of material on FM. Sound on Sound
magazine had a long running series with Gordon Reid. If you want to go more
hardcore and also include physical modelling, Julius O Smith has one or two
books online, which go from maths to practical building blocks. Here is one
example of my own. The instruments were partly inspired by that physical
modelling book:
https://youtu.be/LoTuh63B1GI
I hope anything of that is helpful!
Best wishes,
Jeanette
--
* PeerTube: https://makertube.net/@jeanette_c
* GitHub: https://github.com/jeanette-c
* Youtube: https://www.youtube.com/channel/UCMS4rfGrTwz8W7jhC1Jnv7g
Csound mailing list
Csound@listserv.heanet.ie
https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
Send bugs reports to
https://github.com/csound/csound/issues
Discussions of bugs and features can be posted here |
| Date | 2026-10-26 16:11 |
| From | Aaron Krister Johnson |
| Subject | Re: [Csnd] learning sound design |
Briefly, I'd say as a rule-of-thumb, something sounding "alive" hash to do with dynamic motion: something about the signal should evolve, either at the macro-level of the music, or in the micro-level of the timbre itself. Some thoughts on making things sound better at the timbral level:
"Muddiness" is usually a symptom of a bad mix, or an instrument needing EQ over its range. As this is case-by-case, I'd have to see what you're doing in particular. But in general, sweep an instrument over its range of notes and if certain notes pop "too loud" in a certain range, cut back the amplitude/volume on that note in that range. Think of it as a function curve of EQ over an instrument. Also, "muddiness" is usually a symptom of too much low-freq or to mid-freq volume. This is an EQ/mixing issue. N.B. none of this is particular to CSound...it applies to any mixing and music-making software.Another trick: mix in mono on mid-quality speakers or flat-response studio monitors. Avoid headphones initially. Phasing issues between channels and mistakes in your color/orchestration reveal themselves best in mono. Once you have a good mono mix, and you've tweaked the mix and EQ, stereo as a rule will sound that much better. On Tue, Oct 20, 2026 at 3:47 AM Michael Mossey <michaelmossey@gmail.com> wrote:
|