Hello! For the past year or so, I have been working on a project that I have always found interesting. Basically, I want to make an electronic trombone. I’m limited to the power of a Raspberry Pi 3B, or an Atom board, because its going to be portable. I’ve been testing with a Raspberry Pi 3B, and it doesn’t seem to be powerful enough for such a simple task, I would think.
However, I’ve had quite some time writing the software efficiently. I need to make a brass/trombone synthesizer that is decently efficient. So far, I’ve been re-sampling my trombone playing, to create small loopable waves. Then within my software, I just change the pitch when I need to. It doesn’t sound like it would be that CPU intensive, but this alone uses up ALL of the CPU on my Pi 3B. I haven’t even added oscillation.
That all being said, I am fine with it using all of the CPU, but re sampling sounds pretty bad. I don’t want to record samples for every note on my trombone, and try to fuse them all so I can change pitch when I need to. When I change the pitch of a looped wave too much, It sounds really bad.
I’ve got the whole physical design finished, and sensors, I just need to write decent software. I’m currently using NAudio in VB.Net. I’m open to any ideas. The reason I came to Audacity forums is because I use Audacity to make the short wave samples for my program. I’ve already done some things to reduce speed:
•Multi-threading
•Reduced sample rate
•Reduced real-time editing buffer (Less Latency)
And other things over the past year. Its about as low quality as it will get, and the pitch bending sounds awful. Should I stick with resampling? Or is there some kind of real time synthesis algorithm I can use? I’m really not sure. All help is appreciated, thank you!