How are you determining A.SW1 and A.SW2 are both on? Both of these switches will normally be reset to OFF - and there isn't a way to force them to ON when a Performance is loaded. If you're using the LEDs (lights) to tell you if the switches are "ON" or "OFF" - then you may be interpreting the results incorrectly.

First, make sure your half-glow is not defeated. Just for now. [UTILITY] "Settings" -> "Advanced" and make sure the "Half Glow" parameter is NOT set to "Off". If it's set to "Off" - you won't be able to tell the difference between "OFF" and "ON".

Next, recall your Performance you think the A.SW1/2 are defaulting to "ON". Press either switch A.SW1 or A.SW2. Is it now brighter than it was before? Brightly lit means ON. Dimly lit means OFF. If the LED is not illuminated at all (LED is OFF) this means A.SW does not do anything. Not that it is on or off. Rather that touching it will not cause anything to happen (locally to MODX) because it is not assigned as either a modulation source ("Mod/Control" -> "Control Assign" area) or as XA Control under AWM2 elements.

There is only one way for both A.SW1 and A.SW2 to be "ON" when you recall your Performance - and it depends on A.SW1 and A.SW2 already being ON from the last Performance before you switch. There's a setting to NOT reset controllers. A.SW1 and A.SW2 are both considered to be controllers - along with others (assignable knobs, etc). There's a setting where you can elect to retain the position of the controllers across Performance switching (recall). So only IF you have controller reset turned off AND A.SW1 and A.SW2 are both already on - then they can come up as "ON" when you switch to your Performance. Controller Reset is, by default, turned ON - so this would not be a "normal" setting for most.

Now you can go back and turn off half glow if that was set to off before.

But no - the state of A.SW1 and A.SW2 are not options you have in a PART/Performance.

You have to deal with knowing they will be reset to OFF and adjust your logic accordingly. For the most part - there are options to interpret the switches in opposite ways - so your programming can have positive logic or negative logic and, by way of programming, default to an inverse relationship so "off" of the switch (the default) means "on" for your parameter - or "off" of the switch (the default) means "off" for your parameter. XA control doesn't have all cases - so there's somewhat of a hole there. There's a setting for both switches off, A.SW1 ON, A.SW2 ON - but not both switches ON. So there's one case you cannot "reverse". This is somewhat of a small hole to plug - you can work around it although it does limit some choices.

Many thanks for your detailed reply, much appreciated as it will save me a lot more messing about trying to achieve something that isn't possible 😀

I will look at the performance and try to work out what those 2 switches are doing and then set it up like that for the first scene.

As a matter of interest, the performance in question is 'Soft and Dirty FM' and I like the sound when on the first scene with both Assign1 and Assign2 switched on 🙂

All Controllers have a normal Reset status. Much like the MW cannot be stored in any condition other than 0, the PB Wheel cannot be stored in any condition other than +0000, the Sustain pedal resets to +0, the Foot Controller resets to 127.

But this does not mean that you can’t have the parameter setting start as you desire. You simply need to reverse the Control Set that controls their application.

Here’s how to discover what you need to do:
Hold [SHIFT] + [HOME] to go to the “Overview” screen
(or from the HOME screen, touch “Motion Control” > “Overview” )

On this screen you can see what your panel Controllers, Assign Knobs, Sliders and Super Knob are doing.
Press [AsSw1] to focus the view ... you see it is assigned to Part 1
Press [AsSw2] to focus the view ... you see it is assigned to Part 1
Change the Part parameter from “Common” to “Part 1”
Touch the box that is the shortcut... “Edit Part 1 Control Settings”

As a matter of interest, the performance in question is 'Soft and Dirty FM' and I like the sound when on the first scene with both Assign1 and Assign2 switched on

If you would like the Rev Send and the Var Send for PART 1 to be initially recalled sounding as when those Switches are active... simply recognize that the AsSw always default to minimum (+0) — you must reverse the application of the Control Set parameters so that both Effect Sends are turned up initially, and the act of pressing the buttons will lower the Send Levels.

On the Part 1 “Mod/Control” > “Control Assign” screen
Make sure the “Auto Select” box is active (green)
Press the [AsSw1] to recall its assignment to control REV SEND
Press the [AsSw2] to recall its assignment to control VAR SEND

Knowing that the Switches will default to Off when the program is recalled, we must make the “stored” value reflects the Send Levels in the active setting condition.
We can do this by changing the Control Sets for the Rev Send and the Var Send. You can see the “Curve” as describing the start condition on the left side of the graphic and when engaged the value increases right side. The Curve starts on the center line and increases as it goes right.
We need the Ratio Value to be the start condition and as this is a On/Off Type Switch we simply need to lower the output back to the center line at the halfway point. (All values 0-63 are Off, all values 64-127 are On. As you move left to right half way is is 64).

Change the “Polarity” from Uni to “Bi”
Change the “Ratio” from +32 to “-32”
Do this for both...

This will allow the default status for the Rev Send and Var Sends to be recalled with their active state initially and when the AsSw1 or AsSw2 are engaged to now lower the Send value.

Extra Credit
Please recognize that the initial Reverb and initial Variation Send Levels do not have to be 0. They represent whatever setting is made on the mixer (Send), The graph of the Curve simply indicates the recalled or stored value as the starting minimum and the ramp going to the right indicates the value when engaged (maximum). Since Switches are either minimum or maximum, all values up to the halfway point are Minimum and all values to the right of halfway, are maximum.

If, however, the Curve is illustrating a Knob, for example, because it will send continuous values, you would hear the gradual increase as the ramp heads upward. The +32 setting can be understood as the increase from the “stored value” to maximum.

Left side = stored value
Right side = applied amount offset

Excellent, many thanks. I can see a few hours experimentation coming up later today 🙂 This synth is incredibly deep isn’t it !!

This is very interesting. I was about to post that I couldn't figure out why my AS2 switch on my Montage seemed to be working in inverse after I changed some of the settings, but Bad MIster's post explains that.

An important thing here seems to be that while the switch is not lit it is unassigned, but dimly lit doesn't really mean off, as when it is dimly lit it will still be controlling the sound based on where the curve starts. So dimly lit and brightly lit are maybe best thought as A and B (or controller stored and controller active) settings, and that when the switch is dimly lit it applies the A setting (the value at the start of the curve) and when brightly lit it proceeds through the curve.

Does this sound right?

From many places in the Manual: ”The currently selected button fully lights, the button stored information lights dimly, and the button without stored information is turned off.”

The “1/2 Glow” and “1/4 Glow” UTILITY settings cause any button that is available to offer a parameter in the current context to glow... those that are currently *selected* glow at full brightness. This applies to Assign Switches, Scene buttons, right front panel functions, etc.,

This way you know, immediately, when pressing a button will do something versus when it is active versus when it will not do anything at all.
For example, you cannot always just push the [EDIT] button... there are times when you cannot drop into Edit... it will appear dark. If it is glowing, you can enter EDIT from your current context.

But it sounds like in this case even though the AS1 and AS2 buttons are dimly lit, they are still doing something (in the solution above you gave where one configures the start of the curve to not be at zero), unlike the Edit button that is not doing anything when dimly lit other than letting you know you can press it. But by saying the dimly lit Assign buttons "are still doing something" I mean that they are using the "stored value" in the curve to offset the original value that is stored in the Performance. So it seems like all dimly lit buttons don't act the exact same way, except that it mainly means the button is available to do something more, if I am understanding this correctly.

The Assignable Switches only have two states. They are ON or OFF. The LED that is attached to the Assignable Switch has 3 possible states. Assuming you have half-glow setup so you see the 3 LED states (off, half-glow, full-glow) - then the following applies:

LED OFF = Button not assigned to an active function. Therefore, no matter if the button is "ON" or "OFF" - it doesn't matter since there is nothing at the other end consuming the state of the button. More commonly, we would internalize this as: "this button doesn't do anything".

LED = Half Glow (dimly lit) means the assignable switch is OFF

LED = Full Glow (brightly lit) means the assignable switch is ON.

... this doesn't really work any differently than the other buttons (except there are some exceptions that do not light up the same). But the example you gave about the curve does not speak to the button state. It speaks to what the button does in its default state. Let me see if I can tease this out by quoting you.

But it sounds like in this case even though the AS1 and AS2 buttons are dimly lit, they are still doing something (in the solution above you gave where one configures the start of the curve to not be at zero), unlike the Edit button that is not doing anything when dimly lit other than letting you know you can press it.

Dimly lit doesn't mean "not doing something". Dimly lit for buttons that hold a state, like assignable switches, means both that the button is routed to some destination/XA-control AND that the switch state is OFF. Brightly lit means that the button is assigned to some destination/XA-control AND that the switch is ON.

You're not wrong in noticing that not all buttons work exactly the same way. The general rule is as stated - but the [EDIT] button has some extra features. After you press the [EDIT] button, generally it will change the mode to an edit mode and illuminate the [EDIT] button brightly. This is similar to the assignable switch - which would change some state and also toggle the LED to bright. One difference is that if you press the [EDIT] button while it is bright - a 3rd state will show where it blinks. The "compare" feature.

I do agree that there are instances where the LED state will not be consistent with the general meaning that if it's on (either bright or DIM) then pressing the button should do "something". There are exceptions to this rule and that I have underlined previously. Just that your statement about the assignable switch did not seem to follow a meaningful distinction.

The A.SW LED as dimly lit (meaning switch=OFF) never means that the Performance offsets are at their default un-offset values. Because you can setup a curve where the default controller input value will offset a non-zero offset. This doesn't have anything to do with the LEDs since the LEDs (none of them) change when there is an offset. The offset is, generally, a "hidden" mechanism where you do not have visual feedback of what is going on. LEDs show you what the controller is set to which could be opposite of what the offset will do.

Just like you can have the emergency brake set on a car. If you look at the emergency brake only - you would assume the car is not moving when the brake is set. But set an offset to the incline - and the car could be moving rapidly down a cliff - brake set and all.

But it sounds like in this case even though the AS1 and AS2 buttons are dimly lit, they are still doing something (in the solution above you gave where one configures the start of the curve to not be at zero), unlike the Edit button that is not doing anything when dimly lit other than letting you know you can press it. But by saying the dimly lit Assign buttons "are still doing something" I mean that they are using the "stored value" in the curve to offset the original value that is stored in the Performance. So it seems like all dimly lit buttons don't act the exact same way, except that it mainly means the button is available to do something more, if I am understanding this correctly.

Correct!

Don’t make it more complicated than it need be. If you want to see the impact of this notification, try turning the ‘glow’ feature Off and work with the instrument for a few minutes!

Thanks guys!

I'm new to going on Forums and hope this response goes to "Mad Mister" . Just read your explanation about the switches. Very good information. I just got my MODX6 a few days ago, and am busy starting to program my splits and layer for live performance. I just started working with the A.Switch and figured out how to get it to function as a OCTAVE up (when switch is turned ON) AND Down switch...(when switch is turned OFF) for only the PARTS being played by my right hand. I see the "half-glow" status, and am glad I read your post, so I know what it means) I play Left hand bass with lower two octaves. I became very good at programming my old Yamaha SY-85 and am in the process of jumping thru NEW HOOPS as I am learning the architecture of the MODX. The only thing is that the pitch does not rise a perfect octave when initiated...it's sounding a bit flat...I will have to experiment further and see where it leads me.

That's sort of the problem with "pitch" as an offset. It's not "note shift" which would adjust by semi-tones (12=1 octave) and would not stretch the sample. Pitch as an offset is going to do that (stretch the sample). I'm sort of assuming you're using AWM2 from your description (not FM-X). Because FM-X would be able to sound fine as you don't "stretch" FM. At any rate - assuming AWM2 - you would be better off instead of using A.SW1/2 to offset pitch - to you the assignable switches as XA Control to turn on/off elements. And you would have a set of elements offset by pitch naturally. One set of of elements is your "octave up" set and the other set of elements your "octave down" set. They'd be perfectly in tune and never stretched.