I’ve been thinking about this thread all week and remembering the glory days when I used 3 keyboards + vocals to put out a full orchestra sound.
In front of me, a DX7 (as my master) MIDI’d to a Roland Pro-E arranger on top and a Roland U-20 to my right. On the floor was a Roland U-20 module also MIDI’d to the DX7. I used a Yamaha MIDI processor (routing, cut-off, bypass, transpose, etc) to mix my MIDI signals to all units.
It worked like this. My left hand on the DX7 gave me sounds when I wanted (based on velocity sensitivity) from the U-20 module. One of three volume pedals controlled that volume. The DX7 also spoke to the Pro-E arranger to dictate what chords I’m playing. The U-20 on the right was a stand-alone, just used for sounds. A second volume pedal controlled that volume.
I had instant access to any sound I wanted and could change patches on one keyboard while I played the other one.
But...the many sounds I could produce was SECONDARY to my main effect. That was to give the listener the impression it actually was a whole band. I did this a number of ways. Example: playing lead on one keyboard, I could move my hand to the other for fill-in’s. Most effective was to take over the melody note from my right hand with my left hand (near the end of a phrase) and while still playing the melody line with my left hand, bring in another instrument behind it with my now free right hand.
Of course, in those days, there were not only “players” but listeners too. Audiences have become an endangered species lately. Not a lot of incentive anymore for putting that kind of a setup together again!
Lucky