These synths are limited to 64 tones at a time.
So if you you're using a patch that uses 4 tones, only 16 notes can be played at the same time.
Each note in a rhythm set is also a tone, so this limit can easily be reached by trying to play lots of sounds at the same time.
There are settings that govern what happens when additional notes are requested.
For many compositions, this limit is not a problem.
Is one of the ways to control what happens when the synth runs out of voices in performance mode.
Some care needs to be taken to ensure that important parts of a composition have priority.
You can set a number of voices that will be kept for each part in a performance.
This is only used when a demand for more notes cannot be met.
For example, if you set 4 voices to be held for each of the 16 parts, it does not mean that unused voices will be kept for each part in case they're needed.
It means that notes currently sounding (up to the number of voices reserved) will not be stopped if a new request cannot be met.
Normally, you would set important solo instruments with a higher reserve than less important instruments.
Doing the opposite could cause notes to be missed from important parts when they could be better dropped from background parts.
This is the ability to play different sounds at the same time.
Performances (including GM mode) can simultaneously play up to 15 different patches plus a single rhythm set on part 10.
Rhythm sets have different percussion sounds on different notes.
You can use a similar approach to get more multi-timbrality from patches by using different sounds for each of the tones, and using key ranges and/or velocity ranges to split the tones.
Rhythm sets can only be played on part 10 of a performance.
That means that only one rhythm set can be played at a time.
By default, part 10 is set to receive on MIDI channel 10.
Some expansion cards provide basic drum kits as a patch menu that can be used on other MIDI channels.
Controllers can be used to modify patches and effects while you play.
In addition to the many inbuilt controllers (LFOs, ADSR envelopes, etc), you can use MIDI messages as well.
The keyboards models have physical controllers such as the pitch-bend lever, data sliders and pedals, but for the sound modules, you will need to insert MIDI controller messages in your sequences to get the same effects.
Controllers have a source and destination.
For example, pitch-bend (lever or MIDI message) is a source, and it controls the tones pitch, which is the destination.
You have a good degree of flexibility - for example you can set the pitch-bend lever to control filter frequency for a wah effect if you wish.
These synths allow up to 3 controllers to be used within patches.
The first controller uses modulation as the source, while controllers 2 and 3 can be set to use any of several different sources, including either of the two system controllers.
Each of the 3 patch controllers can modify up to 4 different settings in each of the 4 tones in a patch, and you can select from a wide variety of destination settings to modify.
System controllers are just a way to make it easy to use the same default controller sources in many patches.
Also, should you need to change a controller for all patches, the change only needs to be made in one place.
An additional 2 controllers can be used to modify effect parameters, however, you cannot select destination effect parameters; they are predetermined for each of the 40 effect algorithms.
Here's a diagram showing how controllers are assigned:
MIDI can only send one piece of information at a time, and each piece takes a small amount of time to send, and to be interpreted.
With busy compositions, a delay will sometimes be noticeable.
Here are a few suggestions if this is a problem:
Look at all of the messages being sent, and remove unnecessary messages (eg muted tracks and duplicated notes)
Thin control messages, such as note bend, aftertouch and modulation changes.
Often these are recorded in high detail, and leaving only one in every few messages will not affect the sound.
Delay non-percussive sounds by a fraction of a beat.
It is important that percussive tones (drums, piano, etc) sounds when you want them, and that other sounds (pads, strings, etc) are not competing all at the same time.
Send system exclusive messages before the start of the song, with a small delay between messages.
Changing patches take a small amount of time.
You can set the "Patch Remain" switch in the System settings to keep an old sound playing when a new patch is selected.
However, the effects changes are immediate, and that may change the sound of the old patch anyway.
The JV/XPs all have a reverb and chorus effect plus a fixed number of additional effects processors (from 1 to 3).
Each effect module can be se to any of 40 different effect types, but only one at a time.
So for a patch on its own, any or all effects can be used.
But for a performance, made up of up to 16 parts, often parts that use a patches with a particular effect (for example a Leslie effect on an organ patch), will lose their effect as a part in a performance.
Different modules have their own combinations of built in expansion cards, and empty slots.
Although you can freely swap expansion cards around, they are not really designed to be swapped on a regular basis unfortunately.