Optimizing Performance |
Since the resources of a computer are limited it is wise to use them well. The two main resources we are concerned with are the memory usage and the amount of processortime consumed. You could aim for a low-memory kind of song if you want your music to be able to be downloaded from the web. If you use loads of big 'samples' of all kinds of exotic instruments you'll soon notice that your songs inflate to epic proportions. If you want people to be able to download them with a 14K4 modem to would have to wait forever. On the other hand, if you use only fancy synthetic sounds you can make your songs really small. Especially if you archive your song with a Zip-utility the endresult can be extemely small, where sizes of 5kb are no exception. This is ideal for Web based applications. The downside of using the synthetic sounds is that it consumes more processortime since the instruments are calculated realtime. Therefore if you are planning to use the music in a game for instance you could decide to only use samples (since memory doesn't matter here) and here and there some synthetic sounds and this way have a lot of processortime left to do something else as well. If you use Jaytrax just as a medium to make music and you don't have a special purpose for the endresult other than listening to it you can do any combination as you please. Keep in mind that with the synthetic sounds the rule of thumb is that: The more special effects you use on the more channels, the more processor time it will eat. So If you start composing, don't start immediatly with 16 channels open and play those fancy heavy special effects sounds, because chances are that the computer won't be able to handle it all A preferred method is to start off with say 4 channels and then start designing the instruments you will use. With the nr of channels set to 4 you can't overload the computer with more than rendering your sound more than 4 times. Don't forget that everytime you hit the keyboard and produce a sound the computer has to internally render the same sound again. With 16 channels this can obviously be too much. Then again, if you have a Dual-Pentium II 400Mhz forget what I have just said Typically the processor usage is less when you play your finished song then when you are still edting and creating it. This is because the more time-consuming synthetic sounds will usually be played in 1 or 2 channels in your finished song, but they will be played on every available channel when you are playing them 'live' on the keyboard.
Previous page | Next page |