commercial viability something I know even less about (like, zero). I have no idea what's actually selling out there, or how much money they're making.
I had in mind stuff like audio plugins or standalone audio apps. there are enough books out there, and enough plugin-building frameworks (that abstract away all the platform-specific VST/AU/etc stuff), that I think it's reasonable a motivated person could make a handful of not-trivial projects to share and keep building upon.
again, we live in a time with the best learning resources that have ever existed for this stuff. sample projects galore.
and I think that's why any notion of commercial viability is difficult, because so many other people have access to the same resources. that's why there is an almost uncountable number of audio plugins and software in existence, because nearly anybody with some free time can create their own apps/plugins and share them.
Camden Watson
Thanks user.
Josiah Cruz
On a spectrum between "anyone can learn to do it in a few years of experience" and "black magic that only a select few can even try", how difficult is it to make a proper linear-phase filter/EQ that cancels out if you sum its output with the inverted dry signal (leaving only what you changed)? And how hard is it to have it run reasonably well?
Ethan King
That seems like an oddly specific project. Do you have any motivation behind that particular thing, or was it just an example?
Austin Rivera
I was just wondering, since a few months ago I was looking for linear phase EQs to do a /prod/ thing with, and found that they're not that common, so I thought it might be a very hard thing to do. If it's closer to "anyone can" than "black magic", I'd definitely like to try and build my own in the future, but this question isn't about a specific idea or anything.
Jackson Harris
Literally any time-dependent measurable quantity is a meaningful signal you fucking idiot
Jayden Cox
couldn't tell you, I'm not a DSP guy. I'm a generalist programmer who just happens to have an interest in audio/music software. I've only dealt with the outside of plugins, from an audio host perspective - never the innards.
well, a little bit of the innards. I'm currently modifying a cross platform GUI library to support being used to make GUIs for VST/AU ... but nothing touching the actual audio processing algos themselves.
off the top of my head I'd say, if you can understand the math behind what you're talking about, then the programming aspect is the easy part. you could probably find some simple C programs that do what you're talking about, operating on an input audio file (WAV or whatever). you could start from there (if you understand the math) and keep elaborating on the design until you have something you could plop into an audio plugin framework (ie, the stuff that generates all the VST/AU boilerplate for you, and you just focus on the algorithm)
Blake Adams
Ah, so most of the difficulty in implementing even the craziest ideas is in the math and not the programming, correct? Even for all those unparalleled incredible programs like Melodyne's polyphonic mode or spectral processors (like iZotope RX, the Zynaptiq plugins, etc)?
Chase Jenkins
>difficulty Eh it's not the math nor the programming though both can definitely present challenges. It's the creative design and arguably the design of your architecture's number of levels of abstraction.
