Also - off the posted topic - is there any way to store a voice that's been edited in voice mode to a mix voice location? I edited the mandolin voice in voice mode and was wanting to save it to a mix voice location rather than a user voice location. Doesn't seem to be a way to do

We'll answer this second question first... Any Normal Voice (non-Drum Kit) can be stored within the currently selected SONG/PATTERN MIXING location as a MIXING VOICE. Each Song/Pattern can store a maximum of 16, one per Part. Although you'd hardly ever use all 16. A total maximum of 256 Mix Voices can exist in any one MOXF internal Library.

Place the Voice in a [MIXING] Part of the Song or Pattern you want it associated.
From the [MIXING] screen
Press [1]-[16] to select the PART
Press [F5] VCE ED (Voice Edit, this drops you into full Voice Edit on the Voice in this Part)
Press [STORE] you will be initially offered the MV location (Mix Voice) for this Part.

This is how a Mix Voice is always stored, anytime you are using a normal Voice in a Mix you can fully edit it within the context of that ensemble of sounds, you can change the Insertion Effects, and fully customize the Voice for use in *this* composition... Thlen STORE it locally to a location within this Song or Pattern. This allows this specially edited version of the Voice will always LOAD when you load the the song or pattern, even if you load it individually.

I want to use to use a midi foot controller to control the volume of elements 3 and 4 - so I can play without the note release sounds when I want to.

?
As to the bigger question... I'm not sure how you intend to make this mandolin Voice play. What is "XA note off" do you perhaps mean KEY OFF?
It sounds like you want to trigger the Elements 3 and 4 when you release the keys... Is that what you are trying to do?

Not sure what the controller assignment you attempted is trying to do. Please give us a better idea of your concept. The XA CONTROL can be used to trigger a sound on key off, but this maybe a difficult thing to control. I'm also at a disadvantage as I don't know how the mandolin Voice you speak of is set to play now... Please describe this and how you hope to change it. Thanks.

On the mandolin sound - I want to turn on and off (or volume control at 0 and 127) the key-off triggered voice elements while performing via a midi control so while performing I can mute or sound the key-off triggered elements. In this song I play elec piano in my right hand and the mandolin in my left hand - I've got both hands busy - and I've got a midi foot controller I want to use to either play the key-off sound or not. It seems when an element is assigned to key-off that element no longer responds to volume or expression control commands at the element level. In the voice elements 1 and 2 have no XA, elements 3 and 4 have key-off XA. I can program the elements 1 and 2 to respond to midi volume - I can use my controller to turn their volume all the way down (where only the note-off elements sound). When I program elements 3 and 4 to respond to volume - they don't.

On the mix voices - the takeaway is to create a voice for a live mix don't do the editing in voice mode because the mix banks aren't accessible to store to in voice mode. I saved the edited voice as a user voice, assigned it in mix mode, and then did a voice edit in mix mode to save it as a mix voice.

Thanks for the help!

In the voice elements 1 and 2 have no XA, elements 3 and 4 have key-off XA. I can program the elements 1 and 2 to respond to midi volume - I can use my controller to turn their volume all the way down (where only the note-off elements sound). When I program elements 3 and 4 to respond to volume - they don't.

We see this as the heart of the matter. Every Element has an XA CONTROL setting, perhaps you mean Elements 1 and 2 are set to "normal" which means they will always sound normally in response to a qualifying Note-On command. If Elements 3 and 4 are set to follow XA CONTROL = Key Off this means they will begin their envelope when you RELEASE the Key. You can still control the volume of such Elements in the same fashion as a normally triggered Element.

The same rules will apply. You want to address them individually via the "Element Level" parameter.
With the Voice's CONTROL SET (EDIT > COMMON > [F4] CONTROL SET) you want to activate a Controller to act as a switch for Elements 3 and 4:
SOURCE = the controller you wish to use to activate and deactivate Elements 3 and 4
DESTINATION = Elm-Lvl (Element Level) - use the Element Switches to select just Elements 3 and 4
DEPTH = the amount and direction of change. This will depend on the controller you select to apply the change. If you select a SWITCH then one position will be active the other inactive. In the case of Element Level, that is one position will allow sound the other will prevent sound. You want to control it with a Foot Controller. So setting the DEPTH is going to require you apply a NEGATIVE DEPTH... here's why: the Foot Controller is only capable of sending positive values 0 through 127. You want to have your Controller apply ZERO output, so you need to set the DEPTH to a Negative value so that as you move the controller you move between one position for active and the other for inactive.

The amount of negative DEPTH will determine how far you have to move the pedal to get the response you desire.

If you only apply a Positive DEPTH, you will not hear a change in the Element Level because applying a positive value to a positive does not reduce the level. You need to apply a negative value to reduce the output of Elements 3 and 4... which will be sounding as soon as you release a key... if you manipulate your controller you control the output of Elements 3 and 4

Hope that helps.