Luminite Graviton M1 MIDI Controller

I experienced this as well. I think right after updated, the flash file system was trying do some maintaining tasks in the background, which will crash if the UI gets updated too often or for too long (LEDs drawing). Waiting a bit for it to finish its job (a few minutes) seems to work, as crashing stopped after for me.
I let the device sit for probably 15 minutes and still crashes if I use the F6W on banks with toggles.
 
I let the device sit for probably 15 minutes and still crashes if I use the F6W on banks with toggles.
Yeah the culprit is definitely the Toggle LEDs. I’ll try to reduce their ‘halo effect’ as its can be quite CPU heavy
 
Yeah the culprit is definitely the Toggle LEDs. I’ll try to reduce their ‘halo effect’ as its can be quite CPU heavy
Wait, the effect itself is making it crash or run out of CPU? I think you could probably just use a bitmap image for that, or just a simple "circle with two different colors".
 
Hold function does not seem to work right. I set two of the F6W buttons for "bank -" and "bank +" on hold. It seems to activate the preset under that footswitch in the current bank first, then do the hold action and go to the next/prev bank. If hold action is set, it should wait longer before activation.

Or it should be the typical "tap does nothing, release activates tap" most MIDI controllers use when hold functions are set.

Same issue seems to happen with double tap where it activates 1st tap, then double tap.

Also, after assigning the action it will show in the block view the action assigned. I think it should show the 1st tap action as usual, but maybe have a button to cycle showing different functions for all the footswitches?
 
Should the footswitch sensitivity setting work? I found it by clicking the name of the footswitch in Footswitch controllers, but there is no value to adjust. I tried pairing the F6W again but still nothing.
 
XY -> Assign action -> Encoder Click -> Fixed preset -> X (cancel) causes it to back out of the view back to the XY controllers view, instead of the encoder click settings view.

Same for selecting any action, e.g None should not back out of the view back to the XY controllers, but just set that function to None. Only the back arrow should go out of the view since you might want to assign all the different tap and hold actions in one go.

Back button in Encoder Click view will also not go back to the previous assign action view, but back to the XY controller view. Any action with the ">" submenu works fine and the back button works as expected there.
 
Is it possible to make XY controller repeat its action when holding? E.g if I want Encoder click to keep incrementing by holding the up direction on the joystick. Or is this just reserved for the EC controllers?
 
Last edited:
Wait, the effect itself is making it crash or run out of CPU? I think you could probably just use a bitmap image for that, or just a simple "circle with two different colors".
LED drawing (with blending effect) seems to take too much CPU time and may interfere with other threads running in parallel. 1-2 LEDs are fine; more than that seems to cause crashes. So yeah I’ll try something simpler.

Hold function does not seem to work right. I set two of the F6W buttons for "bank -" and "bank +" on hold. It seems to activate the preset under that footswitch in the current bank first, then do the hold action and go to the next/prev bank. If hold action is set, it should wait longer before activation.

Or it should be the typical "tap does nothing, release activates tap" most MIDI controllers use when hold functions are set.

Same issue seems to happen with double tap where it activates 1st tap, then double tap.
True, there is no ‘1st click rejection’ implemented yet, but will be. The goal is not causing a delay when the Double Tap and Hold Action are unassigned.

Also, after assigning the action it will show in the block view the action assigned. I think it should show the 1st tap action as usual, but maybe have a button to cycle showing different functions for all the footswitches?
Tapping on it again will back to 1st tap action. Yeah something like a quick glance would be nice. Until then, use long tap menu check these functions.

Should the footswitch sensitivity setting work? I found it by clicking the name of the footswitch in Footswitch controllers, but there is no value to adjust. I tried pairing the F6W again but still nothing.
The sensitive (debounce time) setting is for wired mode only (TRS Jacks or M2 footswitches).

In wireless mode, the footswitches’s CPU is in charge of debouching, and is not adjustable by the M1/M2

XY -> Assign action -> Encoder Click -> Fixed preset -> X (cancel) causes it to back out of the view back to the XY controllers view, instead of the encoder click settings view.

Same for selecting any action, e.g None should not back out of the view back to the XY controllers, but just set that function to None. Only the back arrow should go out of the view since you might want to assign all the different tap and hold actions in one go.

Back button in Encoder Click view will also not go back to the previous assign action view, but back to the XY controller view. Any action with the ">" submenu works fine and the back button works as expected there.
Cool ill fix these
Is it possible to make XY controller repeat its action when holding? E.g if I want Encoder click to keep incrementing by holding the up direction on the joystick. Or is this just reserved for the EC controllers?
Yes it’s totally possible.
 
I just updated and loving the new LED, and now you wanna take it away :( lol
It's cute, but you can probably get the same effect much cheaper by just rendering it as a little "pre-rendered" bitmap image that contains the effect. I assume what Luminite's system is doing is rendering the effect itself which the CPU needs to process "number of enabled LEDs" times with the way it blends and stuff.
 
Can’t you trigger “on release” instead of “on select”?…tbh…if you have longpresses as a function…no way around it right?
Much simpler then “ignore 1st click”.
Yeah that's how I've seen it done on every single MIDI controller I've ever used that has a hold function.

No hold -> use tap.
Hold -> use release in place of tap.

If it helps, Morningstar's documentation for their controllers tells a story on how their solution works: https://morningstarengineering.atlassian.net/wiki/spaces/MMS/pages/181698574/Action+Type+List
 
It's cute, but you can probably get the same effect much cheaper by just rendering it as a little "pre-rendered" bitmap image that contains the effect. I assume what Luminite's system is doing is rendering the effect itself which the CPU needs to process "number of enabled LEDs" times with the way it blends and stuff.

I'm all for whatever works, it's just I've been wanting a better way to see when things are on and the LED things seems great, but if they can make it do the same in a better way without issues then that is good too!
 
@Luminite I really appreciate the speed of updates and the way how you pick up on user requests.

I can’t help to notice that each come with some bugs, which are fixed fast. One can never prevent bugs 100%, for many users it’s no biggy…but maybe some users need 100% reliable 100% of the time.

- Is there a rollback function that 100% mitigates risks of settings / functions changing unwanted?
- would 2 phase roll out be an option? Beta version—> official version. (Kemper does that, the “I want new stuff” peeps jump on the betas, the “I need it to work 100% peeps” wait for the “official”.
Isn't that what they are doing? 4.3.0 did add a rollback feature which you can access from the update software menu.
 
Ahh..ok…missed that one.
Nevertheless, I’d see value in knowing if I’m updating to something “proven”, or something I should expect some findings..
I think the firmware should say something like "4.3.0 beta" in it, but at least for the latest update Luminite did say it was beta so bugs are to be expected.

I hope next week they release a new beta that fixes the crash issues and hold tap problems so I can figure out how I want to really use my F6W.

Initial plan is to have some banks where I could set up presets to act like a pedalboard - just toggling some of my pedals on/off, maybe hold functions to show their control view. Then other banks that act more like scenes, toggling multiple pedals, presets etc.
 
4.3.1 Beta is coming tomorrow:
1) Fixed: Crashing with LEDs
2) Fixed: Hold and Double Tap (won’t trigger the normal tap). Each switch will self-configure for the best possible response timing, depends on which modes are enabled.
3) Fixed: Action Menu navigation.
 
4.3.1 Beta is coming tomorrow:
1) Fixed: Crashing with LEDs
2) Fixed: Hold and Double Tap (won’t trigger the normal tap). Each switch will self-configure for the best possible response timing, depends on which modes are enabled.
3) Fixed: Action Menu navigation.

So do we still get to keep the LEDs? :)
 
4.3.1 Beta is coming tomorrow:
1) Fixed: Crashing with LEDs
2) Fixed: Hold and Double Tap (won’t trigger the normal tap). Each switch will self-configure for the best possible response timing, depends on which modes are enabled.
3) Fixed: Action Menu navigation.
4.3.1 Beta is Online now
 
4.3.1 Beta is Online now
Installed and it no longer crashes. Can also confirm the action menu navigation is fixed.

I like the new LED look, I think it's much cleaner.

Tried setting Bank Inc/Dec to the F6W hold functions for two footswitches.

The behavior is a bit odd compared to other MIDI pedals. Instead of holding activating the function, hold-and-release activates it. This makes it seem like nothing happens when you hold.

Either it should activate on hold like most MIDI controllers (probably leading to the typical "release = tap" functionality) or it should show something on screen to indicate you have held it long enough to activate hold. Now I was unsure how long I need to hold before releasing.

I noticed it malfunctioned quite a few times like this:
  1. I start holding the bank + or bank - footswitch.
  2. Instead the preset under that footswitch gets activated.
  3. Then the UI changes to show "Bank -" for example.
  4. If I hold "Bank +" footswitch, it will do steps 1-3 as well.
  5. If I now just tap the "Bank -" footswitch, it will perform the bank change action.
  6. The Bank + state remains after bank change so the moment I tap that button, it will increment bank.
It doesn't seem to do it every time.
 
The behavior is a bit odd compared to other MIDpedals. Instead of holding activating the function, hold-and-release activates it. This makes it seem like nothing happens when you hold.

Either it should activate on hold like most MIDI controllers (probably leading to the typical "release = tap" functionality) or it should show something on screen to indicate you have held it long enough to activate hold. Now I was unsure how long I need to hold before releasing.

Hmm I’m pretty sure just holding for 1 sec (not hold + release) is enough to trigger the Hold Action. Maybe something was accidentally changed afterward. Should be easy an easy fix.
 
Back
Top