AlChuck said:
Rock Star, I think you should hold off on the replies until you know a bit more about what you are talking about.
You could always use 128 instrument sounds, or more properly, there are 128 possible program change messages for a single MIDI port - but there are only 16 MIDI channels and only one sound can be assigned to a channel. So there are really only 16 separate notes you can have at any given time with a single MIDI port -- and at most it's possible to get one different sound per channel.
You were doing great up until this bit...
You can most certainly access more than 128 programs through MIDI. What you say was true until CC00 (Program Bank) and CC32 (LSB for Bank Select)were added (or more properly, defined). These CC (control change) messages allow synths with many more than 128 programs to work together and with computer applications.
A multitimbral synth can play more than one sound on a given MIDI channel at the same time. All sounds on that channel will follow the same MIDI data, but you can play more than one sound.
You can play as many different notes at one time as your synth has polyphony. The maximum polyphony on any single channel is going to be divided among the different sounds assigned to that channel and among all the channels that synth has been assigned to. In other words, if your synth is capable of 128 note polyphony and you have 8 different sounds (timbres) assigned to 8 different channels then you would have 16 notes of polyphony available on EACH channel. However, if you only assign one sound to one channel then you would have all 128 notes available on that channel.
I know this is confusing and I gave a VERY simplified example of how polyphony can be divided between different sounds on a given module. Some manufacturers use different schemes for drum programs and keep them separate from the instrument sounds when defining the polyphony of their units. Also, in a lot of instruments, doubling of notes to "thicken" a sound can reduce polyphony. Adding to the confusion, each manufacturer uses their own "language" to describe their products so that words like "programs", "voices", "multis", "performances", "timbres", "parts", "partials", "patches", "sounds", etc. take on different meanings. What Korg calls a "multi", Novation calls a "performance". They both are talking about a situation where you have several "programs" (a configuration of oscillators or samples, filters, amplifiers, etc.) organized into an arrangement of different channels, zones and layers across your controller.
As you can see, MIDI can be very complicated, but it can also simplify the process of creating music and you can manipulate MIDI in ways that would be difficult or impossible with the same audio data. Wide transpositions of pitch are very simple with MIDI whereas they are almost impossible with audio. Do you want to change every C to a C#? Nothing to it!
Also, MIDI files are MUCH smaller than the audio file of the same performance would be since they don't actually contain the audio data itself, only the instructions on how to perform it. The problem with this is that you don't have any control over how your MIDI files will sound on a different system. This is why the GM (General MIDI) spec was introduced- to minimize the variations between different systems. Now, a GM compatible file will sound more or less the same on all GM compatible instruments.
Manufacturers continue to advance the MIDI spec though there haven't been that many official enhancements over the 20 years MIDI has been around.
If you want to learn more about MIDI, go
HERE
Good luck!
Ted