Moderator: Moderators (Main)
by pieroni; Wed Dec 05, 2012 5:04 am
For a project with University I am trying to create a VST. The starting point is to extend the VstXSynth to have more wave forms (which I have done, whether its correct is a completely different thing haha) as well as an ADSR envelope and simple Delay effect.
In terms of the delay, I have been told I can implement the delay sample code into the vstxsynth, however for whatever reason and whatever I try I can't for the life of me get it to work.
As for the envelope I don't even know where to start. So my question is simple, does anyone have any decent points of reference for a complete beginner to do these kinds of things, or any useful samples explaining the basics?
by arakula; Wed Dec 05, 2012 10:18 am
diggler wrote:Make it in synthedit export the modules and look at the code. This should give you an idea of how the code works in c+.
From earlier experience, my impression was that SynthEdit doesn't emit C++ source code. Am I mistaken? Since when?
by MadBrain; Wed Dec 05, 2012 11:50 pm
Delay: Use a "circular buffer" (essentially an array with the read/write pointers skipping back to the start once the pass the end). Make sure you zero your memory in the constructor.
Envelope: Just code a class that outputs a value that increases/decreases over time (depending on which shape you want).
If you're having a hard time still, you might want to improve your knowledge of C++ too.
by nix808; Thu Dec 06, 2012 12:21 am
you just want to multiply the oscillators stream with how loud u want it to be.
Release is problematic,
I dunno how it's done really.
But essentially just make an amplitude curve that is MIDI dependant and multiply with stream.
With delay, write to memory and subtract your number of samples which is length
when u read it.
There was some other thing that might be neccessary to avoid clicks. It involved the counting at the end of the buffer.
If your delay is clicking, let me know and I will look what I did to mine.
edit-also with the delay you need to feedback, for echo
by pieroni; Fri Dec 07, 2012 5:10 am
So far, I've implemented a lot of the code for Delay, taking into account the frequency amount (as mentioned above). I'm struggling a bit in terms of what needs to go in the process function. The ADelay sample appears to implement the process within the While Loop, am I correct in thinking this?
by David Carpenter Wind Core; Sun Dec 16, 2012 1:20 pm
To export a module into c++ source code right click and select build code skeleton.
This of course is written for synthedit but you could see how to write it for a different app by looking at the source to get an idea.
by random_id; Sun Dec 16, 2012 4:42 pm
pieroni wrote:The ADelay sample appears to implement the process within the While Loop, am I correct in thinking this?
Yes, the audio host passes a chunk of samples to the plugin to process. The While Loop in the example is where the sound is actually processed. The rest of the code is basically used to setup the controls (e.g., what happens when you click on a slider to increase/decrease the value). For the most part, you want to keep things as efficient as possible within the loop.
by daz.diamond; Mon Dec 17, 2012 1:27 am
diggler wrote:To export a module into c++ source code right click and select build code skeleton.
all this does is give the code for a blank module with the pins as per the module you selected - no code is generated for the actual functionality. it is meant as a tool to help 'port' sems from sdk2 to sdk3 ...