Airwindows again
-
So, me and @d-healey had a look at the Airwindows repo on Github and noticed that there was something weird in the license: while the repo's license was MIT, the source code read "All rights reserved". I e-mailed Chris from Airwindows about this and, incredibly, he replaced all instances of "All rights reserved" with "Airwindows uses the MIT license". Check the latest commit here (warning: when you click on the commit, it may take a long time for the page to load):
Commits · airwindows/airwindows
This contains all the Airwindows audio open source code - Commits · airwindows/airwindows
GitHub (github.com)
Chris also told me: "You’re welcome to adapt the plugins in whatever way you like". However, as I'm not good at C++, I have no idea how to implement Airwindows in HISE, so I need your help. If you people don't have time to do it, it's okay, you can give me some directions on how to do it myself. Airwindows and Faust inside HISE would make for a killer combo!!!
-
I've looked at his stuff in the past. A lot of his DSP is really eccentric, and I struggle to find a use for it.
Anyway... you can use SNEX for this purpose.
Grab your variables from the .h file, default values are in .cpp, and the function is in proc.cppDan Korneff - Producer / Mixer / Audio Nerd
-
@Dan-Korneff Same here, I'm also struggling to find the gems and there is a lot of noise and weird edge case tools (most of them related to the boring topic of dithering which I couldn't care less about).
I actually wrote a VST2Node wrapper at one point where you could just drop vanilla VST2 codes in the third party folder (I've reverse-engineered the VST2 API calls into scriptnode callbacks) and it worked with most of the AirWindows effects, but then I lost this file because it was deleted when I exported the DLL again. Just if anyone is wondering why it keeps references to the ThirdParty folder now instead of copying the files into the Binaries folde :)
-
@Christoph-Hart said in Airwindows again:
I actually wrote a VST2Node wrapper at one point where you could just drop vanilla VST2 codes in the third party folder (I've reverse-engineered the VST2 API calls into scriptnode callbacks) and it worked with most of the AirWindows effects, but then I lost this file because it was deleted when I exported the DLL again.
Oh, bummer. I would have loved to have this one.
As for why do I need Airwindows and believe they're useful, well... once I manage to implement them myself, you will know. ;)
-
@Dan-Korneff Could you expand on this? I have very limited coding knowledge.
I'd like to utilise this transformer simulation AW Coils (https://github.com/airwindows/airwindows/blob/master/plugins/WinVST/Coils/CoilsProc.cpp)
And I think if I'm successful once I would know how to do it by myself in the future, if I wanted to use other DSP. A lot of his stuff is quite weird, but I think a few of his hardware emulations are quite useful used subtly as part of a signal chain.
-
@Christoph-Hart While there is an excellent, complex and successful MIT licensed library written by Airwindows, why not add them to HISE, right?
Please add Airwindows library to HISE, I beg.
-
That's quite interesting.
In a month or so I'm planning to release a simulation here myself, for the multiphysics simulation of nonlinear circuits. -
@Christoph-Hart how does the oversampling node operate in Scriptnode? The reason I ask here is because I am working on a project with some code taken from the AirWindows library and I notice a lot of aliasing and artefacts. It seems to me that a lot of this code is designed to be super lightweight on CPU. Example here
I don't imagine dropping the node into an oversampling container in scriptnode would simply do the trick here. Am i mistaken? Also where do I find the oversampling container in the source code?
-
@HISEnberg sure, slap that airwindows node into a oversample container and compile the network again. You could also wrap around the oversampling template on the C++ level, but in the end it won't matter on a performance scale as it should compile down to more or less the same machine code.
