+1 for LFO in Ableton- it’s that moment when you realize Ableton itself is a modular rack. Add in the CV tools and you can have some amazing generative stuff going on inside the DAW itself.
I recently have been playing with the Buchla Easel V from Arturia and their Mellotron V as well, very nice characterful older-type sounds
iPlug2 is also a cool C++ framework (and free) for targetting multiple plugin formats. Probably gonna be more frustrating than JUCE because it’s in “pre-release” and there’s much less documentation, but hey, no splash screen! It’s predecessor, wdl-ol, is pretty similar and has more documentation but they’ve deprecated it. The original iPlug was actually made by the people behind Reaper!
In terms of like general advice, the musicdsp website has a lot of cool algorithms that can be helpful. “The Computer Music Tutorial” by Curtis Roads helped me a lot in undergrad but a lot of it is pretty complex/dated and more high-level/conceptual. It’s worth brushing up on or learning advanced math, calculus/trig/algebraic things come up a lot in dsp.
This severely needs a necro. I also need resources on DSP theory for n00bs
I’m thinking about getting srs here, with a few tools in mind:
I can get internal tools and libraries up and running just fine, but I’ve always sucked at math. I can make cool little noise boxes, LFOs, VCAs, DC sources, bitcrushers and basic shit from scratch but I want to be able to actually do cooler things. Preferably without relying extensively on libraries, because that’s lame.
How did everyone else get started?
Pirkle’s Designing Audio Effects Plugins in C++ would be my suggestion. I think there’s even a chapter called something to the effect of “DSP Filters without Math”.
The good stuff is often in white papers which are math and jargon heavy but you really just need to extract the basic idea and general equations. You learn pretty quickly which parts to skip after a couple of rounds of profuse eye bleeding.
I think to make anything really interesting you have to understand the math a bit - not derive the Fourier Transform from first principles or anything, but understand what the equations are doing to a waveform. Visualization in something like Desmos is super handy for getting a feel. And having a concrete thing you’re interested in is a great motivation to learn the mathematics enough to play with them. Just bashing things together in Reaktor or Max is a great way to see how things work, then you can reverse engineer that into programmatic algorithms with all the math.
That said, DSP wades into deep, deep waters very quickly. The little bit of DSP the music world uses it the tip of the iceberg for audio DSP (think radar and ultrasound) which is the tip of another iceberg for general DSP (everything from seismology to microprocessors to solid state material analysis). I mention this because it’s easy to go down rabbit holes that yield very little substantive material and distract from actually making sounds.
Another bonus resource is anything with Sean Costello of Valhalla DSP. Dude’s low-key brilliant but also sort of an everyman for programming effects. He’s got some good interviews on youtube, and his blog posts on Valhalla’s site are chock full of great info.