I’m probably biting off more than I can chew, but I wanted to get some advice before I dive in. My goal is to write a program that generates environmental noise: 1 – 4 channels loop (with some variation in volume and panning) and provide background ambience, and then multiple sounds are played randomly (also varying in volume and pan). Think sounds of the rainforest or soothing city sounds.

My base sound library is about 20 megs as wav’s. I’m sure they could be stored as mp3s or oggs in much smaller files, but do you think most modern pocket pc devices can decompress them on the fly without interrupting playback of multiple channels?

Suppose (for example) I’m dynamically mixxing 8 channels. Some of the base loops are longer, 30 seconds or >2 megs @ 44.1khz. Some are shorter, more like 100k. On average they’re about 200k uncompressed and @ 44.1khz. So if I only keep what’s needed in ram at the moment we’d need 3megs (about) at 8 simultanious channels. On a pocket pc, that’s kinda a lot.

What I’m getting at is this: since I only need to have a second (maybe less) in ram at any time, (1 second * 32khz * 2 [stereo] * 8 layers = 512kb) how can I best use FMOD to do buffered streaming from a variety of files at once, and best yet, if they’re ogg’ed or mp3’d would it perform acceptably on a 206mz pocket pc device? I suppose I could require a storage card, so the files could be stored uncompressed — could it work then?

I know I’m asking for wild speculation, but anyone who has done something remotely like this will have a more educated guess than I will having never done anything like it.

  • You must to post comments
Showing 0 results
Your Answer

Please first to submit.