Apologies to Luis Garrido, who will get this email twice. Again, I fell afoul of the "Reply" button not sending back to the list.
----- Wow, there's so much swirling around my head, I don't know where to begin. You know, I definitely think that once synthesis and sampling hit software, the division between the two became somewhat artificial. I really don't see why you can't have a great modular synthesis engine that includes powerful sampling features. After all, the sampler engine would be just another module, right? That being said, I have something to say: If we're going to make a DSP language, then let's make a real freaking DSP language! Fill it up with some nice ready-to-use constructs, like delay lines (fixed and variable), basic oscillators and filters, FFT/iFFT functions, and so on. Reaper's JS is nice, but if you need a delay line (a commonly used construct), you need to make one yourself and manage local memory manually, which is a pain. CSound is a great language when it works, but the VST version is completely broken under Reaper, and the regular version has all sorts of issues (weird distorted outputs and the like) I've never been able to sort out. The use of PortAudio has also caused me numerous problems. I belong to the mailing list, and have emailed about these issues, and I still have broken CSound on my machine. I want to code DSP, not debug a software installation without any help. That out of the way, here are some of the many features I'd like a sampling engine to have: * Positional cross-fading: cross-fading between samples across the keyboard, e.g., you have samples at C and F# across the keyboard's range, playing notes in-between will be crossfaded between the two samples. Granted, this isn't too terribly useful for modern "every note sampled" type of libraries, but for samples spread out, it eliminates sample switching artifacts. * Loop-point modulation: being able to control the start and end loop points of any particular sample independently of each other. There are so many cool things you can do with a capability like this. * Cross-faded loops: Nothing much new here, except... * Multiple cross-faded loops: Being able to define more than one loop within a sample, and having the engine select the next loop to play, either at random or according to some setting, then cross-fading to the next loop. This would allow complex timbres to be animated in a more natural-sounding way that'll be different every time. I wish I could claim credit for this idea, but it was brought up on the CSound list recently and it stuck with me as a great idea. After that, things like envelopes, filters, etc., would be separate modules, I assume. A truly modular system with scripting would allow me to realize some interesting ideas I have involving sample-based synthesis and resynthesis, so I'm behind it. Looking forward to more on this. -- Darren ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ _______________________________________________ Linuxsampler-devel mailing list Linuxsampler-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/linuxsampler-devel
