Hi, no snapping back, it's really behaving perfectly, everything you have selected in the automation-menu gets automated properly and when played back the fader moves like it should.

It didn't seem to work the first time i tried to learn the master fader (pitchbend ch9) from the x-touch. i just deleted the learned automation quick access controller in easy mode, relearned it, moved the master fader up and down slowly and it just worked!

My X-Touch One is on 1.07, Logic is 10.4.8, the learned parameters are attached.

Great thanks, I'll give it another shot. It may be your feedback parameters that are making a difference, or I think there was some behaviour differences when I changed firmware version so I downgraded - I'll check which I'm on and see what happens.

I agree though that this is an excellent mod as it enables you to use the fader to write automation for any plugin parameters, rather than just channel volume, and you can't otherwise access plugin parameters from the XTone.

Especially with 'Autoselect Automation Parameter in Read Mode' - just click the parameter in the plug-in, bam, there you go!

You have to have the automation-lane parameters open for this to work, though (keycommand Show/Hide Automation)

Yep, I'm well aware how Logic works, don't worry - I've been using it for nearly 30 years now..

Of course, but maybe somebody else gets something out of this

It's strange that this x-touch one thingie is more or less the controller i wanted for decades.

I'm also on Logic since Logic Audio for the Atari Falcon, btw - which was already in color! well the record button was red.

Indeed - hardware sequencers for me until C-Lab Creator/Notator on the Atari ST, then Logic 1.x on the Atari ST through the last ST version, 2.5 (Logic Audio 2.5 is what you would have been running on the Falcon), then over to the PC until 5.5.1, then over to the Mac for 6.x-10.x...

There's a special place in my heart for the ST, and also a special place in hell for it for the 1 in 5 times it just wouldn't boot unless I'd whisper it sweet nothings and it would finally decide to wake up...!

But yes - good info, and thanks for sharing. I'll let you know how I get on when I have another look at this...

Hmm, it's definitely not behaving like that here and doesn't really work correctly for AQA even those the assignments are good.

I'm on 1.04, will try a firmware upgrade and try again.

Where are you learning your AQA assignment out of interest? In a top level No Zone/No Mode, or somewhere else? Also are you in LC Logic mode? And did you delete the existing Master Fader assignment, or let the AQA command override it?

(In this case, presumably when AQA is off, the assignment might go back to control the Master fader, depending on how Logic prioritises these...)

Edit: Ok, tried again on 1.07, and I still see the same behaviour. With AQA on, the master fader still moves in receipt of the master fader movement, so that assignment had to be deleted. When AQA is on, the fader in master mode will correctly control the currently selected automation parameter, but when I let go of the fader, after a second or so it drops to zero, rather than staying, or following the current parameter value.

I do have a complicated setup, so it's not impossible that the many other MCU's I have installed are interfering with this (it *shouldn't* be the case, but I'm quite experienced in controller assignment funkiness), so I might try again with a temporary virgin set of CS preferences and just this controller - but I think I tried that the first time I was messing with this with no change to behaviour.

So the assignment parameters are identical to yours, possibly the only other difference is where you placed you assignment...

I didn't have to modify/delete anything in expert view, which was surprising to me ...

Now, I also have a macbook with logic 10.4.8 installed, which has never seen the x-touch or any other controller before, so i connected it to the x-touch, which was recognized as a Mackie Control (Logic Control Module).

I opened the automation-lane, enabled 'Autoselect Automation Parameter in Read Mode', clicked on a parameter in a plug-in, the displayed parameter changed to the clicked one of course, learned the x-touch's masterfader in AQA via Mix->Automation Preferences, clicked 'Learn Message', slowly rode the fader up and down 3 times and again, it just worked perfectly.

Then I quit Logic, restarted it, the x-touch one got recognised as a MCU Pro (Logic Control Module) this time (same on my Mac Pro) and worked flawless as before.

AQA seems to rewrite the Parameter for the Master Fader Control everytime you select a new parameter in a plug-in or via dropdown menu.

We both know how flakey the LC implementation of Michael Haydn's genius ideas was at times in some Logic-versions and I hope this will continue to work
Again, you know about 'Rebuild Defaults' and deleting com.apple.logic.pro.cs which was often required in the past and might help here? (And hereby deleting every bit of custom work in the expert view ...)

Anyway, I attached grabs from the first and second run on my macbook and one from my Mac Pro, Zone/Mode view included. The Macbook is running Catalina, the Mac Pro Mojave. Hope this helps!

Thanks for the screenies... yes there's nothing there I haven't already recreated and tried to no avail...

Yep, I am very familiar with controller assignments and the MCU system and am one of the few who has severely hacked the MCU for my own purposes (even to the extent of writing my own virtual MCU), but I can't get past this behaviour of the XT.

By default, even when it's not active by a DAW, if you move the fader, it resets to zero, and I don't really know exactly why it does this or what causes it to stop. I know if I use the XTone this to control my audio interface through a third-party app, it no longer resets the fader and works as expected (and also as an MCU in Logic) so there is *something* the XTone needs, maybe an init string or some feedback to convince it it's active and not to reset the fader.

I'm not the only one who's had this - on another thread here I think, someone else tried to use the XT-ones master fader mode to control AQA and got the exact same fader reset behaviour as me, so I know it's not something unique to my system.

I'll try it with LPX 10.4.x (I'm normally on 10.5.x) just in case that makes a difference too (although again I think I tried this when I was investigating this before), but it is a weird one I haven't managed to get to the bottom of yet.

I don't think in this case it's a problem with buggy controller assignments, it's something to do with the default behaviour of the XT-one and what makes the fader reset I think....

Yes, when the target parameter changes on an assignment, the controller assignment shows the updated current targeted parameter, this works through all controller assignments, eg the VPot assignments will show the current plugin parameter they are currently targetting (and this is great for implementing/testing things etc).

I will try all this again but yes I'm an old hand at this stuff, I even have custom scripts to restore known good CS prefs files from backups if/when they get messed up, so I know the score here... I have to do a lot of this hacking for projects I'm working on, but like I say, I don't actually think it's a controller assignments problem here, other than this fader reset business, the AQA stuff being controlled by the fader does work, it's that the XT-one is behaving in a strange manner that I don't yet understand - and this is why when I first tried this, I ended up leaving it there as I didn't have the time to investigate any further back then... *sigh*

Edit: Same behaviour in LPX 10.4.8.

Appreciate the info - glad it works seamlessly for you!

So you're a real Logic yedi ...

ok, installed logic 10.6.1 on the macbook and here the masterfader gets recognized during the 'learning phase' (the parameter moves on the plug-in), but as soon as i click 'done' the controlled parameter reverts to 'Volume'. This is defenitly different than in 10.4.8. And when i delete this assignment in expert, i also get the fader snapping back after 3 seconds.

Quickly checked in SNOIZE MidiMonitor: Logic always sends a bunch of Mackie sysex strings when a fader is released. If the x-touch does not receive those, the fader snaps back to zero.

It seems that Logic does something different here from 10.5 on ... oh dear .. let me know if you find a way around this, please!

I've been around...

This is I think key - if the master fader assignment is *removed*, I get the snapping behaviour. If I *don't* remove the assignment, then with AQA on, both assignments match, and so the master fader moves when you're also altering a parameter (clearly not ideal! )

So I might be able to try keeping the master fader assignment, but tucking it away in an unused mode or something to try to get round this... (Edit: Tried, doesn't work.)

In fact, even selecting the master fader assignment and command-x cutting it instantly drops the fader to zero, only to be reinstated when you paste it somewhere else. It's like Logic really wants that fader to be there... (I always hated the damn Master Fader anyway...!)

The sysex strings for the MCU are always display updates, as that is how the display is encoded - anything you change on the MCU, or in Logic that requires a display change will send those sysex strings (it's mostly just two lines of 56-char ascii text, or only the chars needed to reduce the MIDI bandwidth). The MCU doesn't use any other sysex strings, it's only for display updates (and a bit of handshaking on initialisation - "Is there an MCU there?" "Yes! I'll tell you what version I am too!" "Ok, gimme me your response to my challenge!" "Here you go!" "Great, you're in!" kinda thing).

Like I say, I still got the fader snapping in 10.4.8, but that was also with the master fader assignment removed.

BTW thanks for your help/back & forth on this, it's quite nice to occasionally find someone who cares about this stuff - it's usually a fairly solitary experience...

Saw this after my post - so i jumped back to 10.4.8 (zipped the program before installing 10.6.1) and it's working fine again, no snapback, perfect & fast control of every plug-in parameter with feedback. i can click through the parameters in a plug-in and the fader jumps to the correct position. this is really weird.

yeah, this makes sense. there are 6 or 7 strings and maybe in one there is some kind of 'active sensing' stuff, that the behringer programmers included in their implementation? anyway, the x-touch expects some kind of command, doesn't get it and resets the fader - that's my thinking here. i have been wrong before.


try the same in 10.4.8 after a 'rebuild defaults' - AQA via masterfader will work

totally & thank you - and it's a strange feeling that i'm gonna loose this new-found functionallity if i finally get a new mac for my studio

I'm wondering if the XT-one uses a bit of a hack to determine the master fader position... I haven't inspected the data flow yet, but it's possible the XT-one is reading the display updates in that "Master" green button mode to update the fader position value - and with no master fader assignment, Logic isn't sending the expected display updates and the XT-one then goes "right, I'm in offline mode, let's reset the fader".

If that *is* the case though, I can't think of an easy Logic-only way to fix that, without something intelligent inbetween (which I don't want to get into at this stage) or some hacky environment stuff (which I also don't want to get into...)

Yes, it's more or less along the same lines that I'm thinking...

Okies, bear with me... let's see what happens...

Edit: So, I don't really see any differences in behaviour. I cleaned everything out, re-added just the XT-one from scratch, verified the master fader worked etc. If I then learn the fader in Master mode to AQA, and with this turned on, it worked with no fader snapping. But, as expected, it also controls the master fader itself when used, because both incoming assignments match.

And as before, if I remove the real master fader assignment, the fader instantly goes a snappin' straight to the bottom.

This whole thing wouldn't be an issue if you could change the MIDI message the fader sends in Master mode, as it wouldn't conflict with the Master Fader assignment. And any interception/conversion between the XT and Logic requires an intelligent middleman, which I can do but need to really solve this with a Logic only solution.

I haven't given up yet, but I'm running into the same walls myself and the other guy both found when we implemented this shortly after we got our units, which makes me want to put it down to a "well, don't have a solution yet, would be great if it worked, but I'll maybe look at this some other time..." type thing...

Our conversation feels like 2005 and Logic 7-8 all over again to me, the weirdness of this genius hack called 'Logic'

Thanks for trying it out, but what can i say, on both computers here the silly horizontal 'Masterfader' does NOT move when assigned to AQA. so strange. When I do a 'rebuild defaults' the damn, useless 'Masterfader' is working again, the learned message for AQA is removed and AQA gets switched off. Then I relearn it and - same result - no movement on the bloody 'Masterfader', it stays put where it was before and i can happily assign anything to the x-touch.

As this only works in 10.4.8, I now checked via SNOIZE what Logic is sending as feedback in this case. I turned off all the 'Help Tags' and 'Value Units' in the Preferences and no sysex strings were sent as a feedback to the x-touch anymore, you were of course right about the sysex being here for display mainly.

Just the simple format:
From X-Touch: Note 112 CH1 VAL127 (Touch Fader)
From X-Touch: Pitchend CH9 many values
To X-Touch: Pitchbend CH9 (Feedback from Logic ) same values, quantized
From X-Touch: Note 112 Ch1 VAL0 (Touch Fader off)

So if the Behringer only needs this kind of feedback, that would obviously be a good sign.

Well, to tell the truth i haven't used it anymore (i found out that i am a lot happier with drawing an automation curve wrt moving the fader in real time), and the one time i tried it again it didn't worked; let me/us know if for you is really working

Yes, controller assignments were *way* more buggy in LP7 back then. I would frequently tear my hair out trying to develop a plugin control system...

Strange indeed. The master fader is a modeless assignment, it never changes in the MCU, so is always active. And if you learn another assignment to that same incoming message, they should both respond (and that's what I see here).

So instead of deleting the standard master fader assignment, I left it intact but changed the input message it responded to, so it would no longer match the fader's messages when moved.

I then learned the fader to the AQA command, and it worked for AQA as expected, and the Logic's master fader did not move - so far so good. But the fader still snaps to some value - it's not even zero - it's like the XT-one is storing the last value it received, or an internal value for the master fader, and resets the fader back to that position.

And thinking about it - this makes sense, because when you press Master on the XT-one, it's screen goes off and the fader moves to the master fader position. I haven't sniffed the traffic yet, but my hunch based on knowing the MCU spec is that it's not Logic that's controlling this (because the MCU doesn't work like this), it's something they implemented for the XT-one - which would imply that the XT-one is internally storing the last master fader position (probably from the feedback messages), and if the fader in master mode is in a different position than the stored value, the XT-one puts the fader back to where it should be in master mode - and this is the behaviour we are seeing. (Or at least *some* of us! )

Edit: Interestingly, if I change the master fader assignments input value (so the fader doesn't control the master fader), with the XT-one in Master mode, changing the master fader in Logic with the mouse *still* sends feedback to the master fader. So it's that master fader feedback that's returning the fader to the master fader's position. Either Logic is sending that when you switch modes, or more likely the XT-one is storing the value internally... (so in order to get around the fader snapping, you'd have to fake feedback the master fader when doing AQA, because in AQA mode where it's not moving the master fader, there will be no master fader feedback coming back from Logic when the fader is moved, hence the internal master fader value is not updated, and the fader snaps back.

I'll have a play, and watch the feedback mechanisms and see if I can deduce anything more about what's happening...

Unfortunately, the feedback mechanism for Logic's controller assignments is a black box, and is actually baked into the compiled control surface module - you can only choose from a variety of feedback types the MCU supports per assignment, you can't change the messages that are sent (eg to fake feedback from a different control, for example)...

yes, or, if it doesn't receive any feedback in around 2 seconds it presumes it is offline and snaps to zero. This is the same behaviour when you adjust the x-touch one's masterfader without any DAW running.
I'll also try this out later with some fake feedback from the environment or a delayed feedback of itself.
I still think Logic has changed its automation from 10.5 on, its also snappier, more responsive to me.

cheers!

Very possibly - but in this specific case it's snapping back to the last master fader value. If it set the master fader to say the middle, then render the master fader inactive and use it in AQA mode, the fader will snap back to that middle position.

Quite possibly both behaviours are true.

Yep.

Thanks, your help and additional findings are much appreciated!

Just tried a simple midiloop with the essential Midipipe from Nico Wald - In Masterfadermode the fader acts like a normal one without snapping back, if it gets the same data from the input.

and yes, the x-touch one stores the last value. when toggling between 'normal' and 'master level' mode it goes back to the last used one in 'master level' mode - which is a cool feature.

To get this working past 10.4.8 (which works for SOME ), i'd filter out the Pitchbend on CH9 & the Touch note for the masterchannel - send both to its own virtual midi input, feed it back to the x-touch and assign this virtual port to AQA. This would be a simple, feedbackless (meaning no parameter feedback from the plug-ins...) system that should work.
All easily doable with Midipipe, which i trust more than the environment and runs systemwide with near to zero system resources.

I'm sure it's not that hard to get parameterfeedback from Logic working also.
Yeah, sure.

Yep.

In essence, this is stopping Logic receiving fader movements in Master mode from the XT-one, and instead routing those via a different MIDI port to Logic to keep them separate, then setting the AQA assignment to that event on your extra port.

Yes, it should work - however, you'd also need to filter master fader output events from Logic back to the XT-one, as otherwise, when you move the master fader with the mouse (not that you should ever really do this in normal stereo use of course!) that will also move the fader.

Yes, this stuff is doable, but all requires a bit of a faff and then breaks when you forget to run that other utility etc, which is more fragile than I'd like my setup to be.

(I actually *could* implement it in a better way, because as I say I have the ability to run an intelligent agent between the controller and Logic, and do all kinds of this stuff there - but that's not what I'm designing that particular tool to do.)

I'm really curious though why in your (working) system, you avoid the snapback of the master fader to it's last held value - and also you have the selected automation parameter playing back via the AQA assignment to the fader which follows it's movements. It's almost like the AQA parameter is correctly updating/feedbacking the XT-one internal "master fader" value - and I've never seen any of those behaviours here in 10.5.x or 10.4.x - it's really weird!

There *might* be an environment solution of tapping off the automation values and mapping them back to master fader feedback messages, but then getting the parameter value for the currently selected AQA parameter isn't that obvious.

So in summary of general events so far, just with a Logic-only solution and no additional "hacky" stuff involving extras ports and transforming etc:

- Without any Master Fader feedback, the XTfader in Master mode will snap back to it's last stored internal value (which will be zero if it's never been set) when it's not being touched

- The feedback that updates the internal fader value comes from Logic sending master fader events back to the XTone - not from the feedback of the assignment itself. As you adjust the master fader, the feedback will be sent by Logic, but if the master fader assignment is removed, the master fader isn't changed, no feedback is sent, and the internal fader value isn't modified and causes the snapback.

(The feedback mechanism is used not to update the control position, but to give the appropriate meter level, so the feedback mechanism doesn't just echo the fader events back, it's sending the lights of the appropriate meter mode back)

- You can stop Logic's Master Fader from responding to the XTfader in Master mode in a few ways (deleting or modifying the assignment)

- You can't stop the fader responding the Master Fader feedback sent from Logic when the Master Fader is moved (eg with the mouse) (because Logic communicates directly through the control surface module for this, not using the assignment)

- You can learn the XTfader in Master mode to AQA, and write automation for the currently selected automation parameter, with all modes, touch working etc - but when you let go of the fader, after a second or so it will revert to it's last master fader internal value.

I'm not sure offhand whether the sent PB events from the fader are the same sent back to it via the Master Fader feedback (I think you said above that it was) so if we just loopback the XTfader events to the XT in Master mode (I think it sends a different MIDI message when in this mode), I wonder whether that would be enough to update it's own internal Master fader value and avoid this conflict...? I think that would work...

It's all fun and games with controller assignments..!

I have to do a last shoutout to Midipipe, which has always running in my fairly complex & big studio-setup (Login item) for 10 years now. It converts & maps all kinds of mididata, especially to control RME's Totalmix via several controllers, also hacking into RME'S MCU implementation - it's just fantastic, totally reliable and also already an ARM-ready universal binary, Nico is just great!

I actually didn't do anything, it just works perfectly.
Last screengrab attached below, because it's easier to explain for me - i can select any of these 3 tracks in either x-touch mode (normal or masterlevel) and the fader always represents the correct value.
i can toggle to my hearts contet, it just works. is that what you mean?

All good points following, let me comment on a few:

In 10.4.8 i'm definitly getting feedback from the assigned parameter i'm controlling. It is really like the master fader assignement got overriden & shut off (which is the way it should be!). Absolutely no idea why this isn't happening in your setup. But yeah, maybe this is really a lucky 'bug'

This is the behaviour i found in Logic 10.5+, not in 10.4.8

yeah, sounds good!
I think, it doesn't even need the Touch-Note, just the pitchbend-data on CH9 and no snapping.

I'll get back to this when i, god knows when, really get a M1x kind of computer, where i will have to update logic :(

cheers man,
thanks a lot for this!!
-b

So, I had another play... (I'm a sucker for punishment!)

- I changed the input message of the Master Fader assignment to not be triggered by the XT's fader

- I used MIDIpipe to sniff the output of the XT (not hijack), with a channel filter only letting CH9 events through (the master fader is the only one that uses this, as it's always CH1-8 for the 8 tracks in a fader bank, and CH9 for the master fader), and then echoing this back to the XT's MIDI port.

So the XT will be sending everything to Logic as normal, but also, any CH9 events (the fader in master mode) will be directly echoed back to the XT via MIDIpipe.

So the MIDIpipe script is:
* MIDI In sniff (XT MIDI port)
* Channel Filter (Let only CH9 events pass)
* MIDI Out send (XT MIDI port)

- I then Learnt the AQA control to the XT's fader in master mode.

A really simple config, and this lets me use the XT fader in master mode to write AQA automation perfectly without any snap back (because the fader in master mode always gets it's own data echoed back via MIDIpipe, which updates it's own internal fader value).

What I don't get of course is the fader playing back the automation moves on that automation parameter, because again, there is no feedback mechanism for this (I still don't know how the hell this would even work on your system!! Maybe through feedback on the AQA assignment perhaps, but I can't get this to work so far) - but this is not a big deal imo.

Just being able to access and write automation for plugin parameters seamlessly is a big enough win, and this MIDIpipe config is not exactly complicated. It's a perfectly good enough solution for now...

great!

yes, also for me the main thing is being able to quickly 'perform' the feedback of delays, filtercutoffs, effect mix values, sends ... parameters like this. if needed, i'll just edit the written automation via the mouse.

being able to edit and record the displayed automation parameter by just pressing 'masterlevel' on the x-touch is a fantastic workflow!

still, emagic (ups..) should really make it an official option to use the ******** 'masterfader' in the MCU implementation for AQA with all the benefits and no need for something like this. let's hit the apple feedback-page yet again
https://www.apple.com/feedback/logic-pro.html



EDIT: gearslutzspace censors: r e t ar d ed. very pc. allright.

Agreed!

Yes, that's not a bad idea.

Although they'd probably be like with a "proper" MCU "er, all the channels in flip mode will write whatever automation you want" and then you'd go "yeah, but these one fader things don't let you access the plugin parameters at all" and they'd be like "well get a proper control surface then" etc...

Remember, it's actually an XT-one feature that maintains the value of that fader internally so it can return there when you hit "Master", which isn't an actual MCU feature, it's a workaround to add an extra feature for the fader.

Anyway, while this implementation is not *completely* ideal, it's fine for now, so thanks for helping me investigate this and find a solution...

that's of course true. having the masterfader on a 'real' MCU device always reflecting the AQA parameter might be a bit stressy at times

you'd definitly need an option to turn it (or at least the feedback) on/off, something we already have.
(AQA on/off as a keycommand already exists, of course)

before the x-touch one i had a mcu&extender and i've never used them for parameter editing in plug-ins. always took too much time for me to find the correct parameter. and the promising new smart controls way to do this also takes a bit of setup time first, but right now it is buggy and often crashes logic ... i really prefer the simple AQA.

Hi all

I'm reviving this old thread because I also am "considering X-Touch One to use with Logic Pro X"!

I also have a Console 1 that I use for EQ and Dynamics and I don't use many other plugins really, so I would only really be using the fader for writing volume automation. This is my main objective in getting the X-Touch One, along with transport control and the jog/shuttle wheel.

The only thing that's not clear to me is whether there is a built-in or easy method for writing send automation? Id like to automate sending to a Delay bus, for example. Or would I have to try to use this clever trick re-assigning the master fader mode to track automation via AQA (assuming that even works now!)?

Thanks

Will

Ok.

Without any modifications as discussed in this thread, the *only* thing you can write automation for is volume for the selected channel, and pan. The XT-one simply cannot automate anything else, it's not a feature it has.

By using the above hack, you can use the fader as AQA, which lets you automate any parameter you select in Logic. So yes, you'd need this hack to do this.

Thanks for the quick and clear reply.
I'll read through the thread again and decide if I want to go down this road...