Announcing ReaLearn: Sophisticated MIDI/MCU/OSC learn for REAPER

[mousi]

Hey helgoboss. I just got a midi controller recently and tried ReaLearn, it is amazing, thank you!

One question in case this is possible. I am trying to "reset" the minilab's LCD screen after adjusting a volume/pan etc. It correctly displays the values I want but then it stays there (showing the pan for example) forever, or at least until a new LCD feedback action happens. With the use of parameters etc, I have already made a mapping that when the display changes, it waits 5000ms and then disables the mapping that sends the pan feedback so it goes back to whatever mapping was active before that. The problem with that is that if I keep adjusting the pan (say 4500ms after the first adjustment) then it only shows the feedback for 500ms because of the first adjustment's delay expiration. Of course if I keep adjusting it goes back to showing the feedback and the cycle restarts.

Quite minor but I was wondering if there's a better way to do this. Could I cancel previous "fire after delay" actions for example? Or maybe there's baked in support for something like that and I'm missing it

[7enz]

Hi, Helgoboss...
just want to express my gratitude again and say thank you for making Realearn...
im looking forward to the release of playtime2, when it is ready *smiles

in the mean time... i wanted to share this video, as it is the result of what realearn has allowed me to establish as far as a looping based template, Reaper & Realearn is central to this all coming together, admittedly its also working alongside 4 instances of the Maschine VST, allowing me to mixnmash up some NI expansion projects...
but again everything gels and works together becos of Realearns midi capabilities..

in this loop based template which is still a work in progress, i've been able to utilise Realearn and map shaperbox2 parameters to buttons to allow me to have rhythmic and drum fill variations in ways this video only touches the surfaces on...

but for those who would suggest Reaper is not a loop based DAW... this is not entirely true when you use the right midi tools & VST's
again thank you Helgoboss for your work in creating and maintaining Realearn!!

heres the video
https://www.youtube.com/watch?v=P9tHYBoM5nE

[Perken]

Hello Everyone,

I've been using some Lua and my Midi-fighter twister over the summer, trying to wrap my head around scripting mappings for Realearn. It's been a ton of fun, and I'd like to share my most ambitious project with you yet: a mapping to control VST synths from the MFT.

DISCLOSURE: if this post seems too long to be in this thread, just let me know and I will move it to its own thread.

Obviously, this isn't an end-all be-all of Synth controllers, but I like it enough to share it.
Here's how it's designed:

2 pages of mappings:
  Synth page:
    One row for oscillators
    One row for Filters
    One row for Envelopes
    One row for LFOs
  FX page:
    One row for Distortion
    One row for Chorus
    One row for Delay
    One row for Reverb

Each row contains 4 encoders, and 1 param is assigned per encoder. Let's take the Oscillators row for example:
If the midi-fighter's row has only 4 encoders, how can I control 3 oscillators X4 params with it? The solution is to use layers. So, pressing the first button of the row will switch between the controls of each of the oscillators of the Synth:

Row1:
  Layer 1: Osc1 (navy LED)
  Layer 2: Osc2 (yellow LED)
  Layer 3: Sub (green LED)
  Layer 4: Noise Osc (cyan LED)
Row2: 
  Layer 1: FilterA (red LED)
  Layer 2: FilterB (purple LED)
Row3: 
  Layer 1: Filter Envelope (red LED)
  Layer 2: Amp Envelope (yellow LED)
  Layer 2: Pitch Envelope (purple LED)
Row4: 
  Layer 1: LFO1 (light blue LED)
  Layer 2: LFO2 (navy LED)

The full schematic of the map:



for each row, the controls of each layer are the same:

Encoders:
Row1 (Osc): Octave | Coarse | wave shape | level/volume
Row2 (Flt): cutoff | res.   | ???        | drive/mix   
Row3 (Env): attack | decay  | sustain    | release / amount(press and turn)
Row4 (LFO): shape  | rate   | ???        | amount / destination(on click)

Now that you get the gist of how I've imagined the layers, here's the layout of the synth page as structured text:

  Synth page:
    OSC 4layers(#1,  #2,  sub, noise) Encoders(Oct,    Coarse, WaveShape, Volume)
    FLT 2layers(#A,  #B,  /,   /    ) Encoders(Cutoff, Res.,   Drive,     Mix)
    ENV 3layers(Flt, Amp, ptch, /   ) Encoders(Atk,    Decay,  Sstn,      Release)
                                      BtnPress(/,      /,      /,         Amount)
    LFO 2layers(lf1, lf2, /,   /    ) Encoders(Shape,  Rate,   ???,       Amount) 
                                      BtnPress(/,      /,      /,         Destination)

The mapping has a second page, dedicated to FX, with the classics: distortion, chorus, delay and reverb.

  FX page:
    row1 (Distortion) Drive | ???    |  ???  | ??? /Mix(press and turn)
    row2 (Chorus)     Rate  | Dly1   | Dly2  | Fdbk/Mix(press and turn)
    row3 (Delay)      Speed | ???    |  ???  | Fdbk/Mix(press and turn)
    row4 (Reverb)     Size  | Densty | HiCut | LoCut/Mix(press and turn)

The mapping provides some visual feedback by showing the LEDs of the layer that are active at the moment:
Layer1 will activate the first LED of the row. When it's active, all the other LEDs are turned off in the row
Layer2 will activate the second LED of the row.
etc.
Also, since the Envelopes and LFOs have an amount button, the color of that button matches the destination: Red is for Filter A, Yellow is for Amp, Cyan is for Pitch, Purple is for Filter B.


Please let me know if this explanation was clear enough, and if you'd like to make a donation, please send it to Helgoboss
You'll find the lua script for this mapping in my reapack handle.

Usage:
- run the script (it will copy the mapping to your clipboard),
- add Realearn to your track,
- in realearn's window, click the «Import from Clipboard» button.
- You'll have to work the assignments from there.

[Cloudswim]

Hi there, is it possible to receive feedback from within a VST plugin to let ReaLearn know which MAPPING GROUP to activate?

for example I'm working with a sampler plugin called Geist 2. Its like a traditional sampler with 16 tracks. On each track the user can have different layers of samples and I'm wondering if it possible to have Realearn acknowledge which layer I'm working on so a specific Mapping Group can be activated.

Hope this makes sense!

[Cloudswim]

I have a Loupedeck controller , the encoder knobs are endless but I'm still getting parameter jumps when I assign the same knob to different parameters.

Is there a way in ReaLearn where I can correct this issue?

thanks!

[7enz]

Originally Posted by Cloudswim

I have a Loupedeck controller , the encoder knobs are endless but I'm still getting parameter jumps when I assign the same knob to different parameters.

Is there a way in ReaLearn where I can correct this issue?

thanks!

have a look at the smart knobs script v1... and use it in conjunction with Realearn, this script is often overlooked by many users of reaper

https://forum.cockos.com/showthread....t+knobs+script

edit: rereading your post.. i noticed you mention 2 assigned parameters to a single encoder...
the behaviour of encoders would be correct if used as absolute... perhaps changing the relative type... might help.. ie +1 and -1

[Cloudswim]

Originally Posted by 7enz

have a look at the smart knobs script v1... and use it in conjunction with Realearn, this script is often overlooked by many users of reaper

https://forum.cockos.com/showthread....t+knobs+script

edit: rereading your post.. i noticed you mention 2 assigned parameters to a single encoder...
the behaviour of encoders would be correct if used as absolute... perhaps changing the relative type... might help.. ie +1 and -1

hi, single knob assigned to 2 different parameters but activated in separate mapping groups. I tried it with several controllers and it dosen't seem to behave smoothly. I will try out your suggestion, thanks!


Edit: Just tired out all the other encoder types and they don't even function properly with my encoders, they all act the same and I've tried out at least 3 different midi-able controllers. When I learn the knobs they initiate as Range Element and this is the one that works flawlessly, the only problem is that I get value jumps if it is assigned to different parameters. I've actually never gotten this to work properly, which controllers is everyone using?

[7enz]

Originally Posted by Cloudswim

I have a Loupedeck controller , the encoder knobs are endless but I'm still getting parameter jumps when I assign the same knob to different parameters.

Is there a way in ReaLearn where I can correct this issue?

thanks!

i do wonder if your controller supports feedback or not??
in your mappings
normally you would have both ticks selected which are normally in the centre... one for control & the other for feedback... if the feedback is unticked then there is no feedback of course, thus you would get parameter jumps
im unfamiliar with the capabilities of the loupedeck...and a big question mark over whether it supports feedback

[kdarkthought]

Originally Posted by Perken

Hello Everyone,

I've been using some Lua and my Midi-fighter twister over the summer, trying to wrap my head around scripting mappings for Realearn. It's been a ton of fun, and I'd like to share my most ambitious project with you yet: a mapping to control VST synths from the MFT.

This is awesome!

[Perken]

Originally Posted by kdarkthought

This is awesome!

There's a template for an overlay with labels for the MFT. I'd like to make some edits to it at some point, to match my mapper. Here's the template

[Miscreant]

It appears since the update to 7.03 my MIDI device ID names changed and now ReaLearn is not triggering assigned actions via external MIDI devices?

For example, I use a Fighter Twister to control my TCP when a floating FX isn't focused. Likewise, I use an external synth with a footswitch to control record functions. Neither is working.

It's not a more general MIDI issue, since I run scripts that send MIDI CCs to my AxeFx and TotalMix, and all is good (though I had to change back my Axe Device ID, the only whose original number I happened to remember).

Anyways, enough going on. If my MIDI device IDs did change, how can I reassign them and get ReaLearn to re-see them?

*I'll also add that Reaper sees and has properly registered my MIDI devices, as I can see activity from my keyboard in the MIDI log. This seems to be a problem strictly between these MIDI devices and ReaLearn, and since I think ReaLearn registers/recognizes devices by ID, if these IDs have changed then ReaLearn won't see them...?

**FIXED: it seems after updating to R7 ReaLearn forgot all the input/output devices for different assignments. AND some of my device IDs changed in Prefs, requiring me either to edit scripts with those IDs or manually change the IDs back.

Anyways, just in case anyone else is having the same problem.

[7enz]

Originally Posted by Miscreant

It appears since the update to 7.03 my MIDI device ID names changed and now ReaLearn is not triggering assigned actions via external MIDI devices?

For example, I use a Fighter Twister to control my TCP when a floating FX isn't focused. Likewise, I use an external synth with a footswitch to control record functions. Neither is working.

It's not a more general MIDI issue, since I run scripts that send MIDI CCs to my AxeFx and TotalMix, and all is good (though I had to change back my Axe Device ID, the only whose original number I happened to remember).

Anyways, enough going on. If my MIDI device IDs did change, how can I reassign them and get ReaLearn to re-see them?

*I'll also add that Reaper sees and has properly registered my MIDI devices, as I can see activity from my keyboard in the MIDI log. This seems to be a problem strictly between these MIDI devices and ReaLearn, and since I think ReaLearn registers/recognizes devices by ID, if these IDs have changed then ReaLearn won't see them...?

**FIXED: it seems after updating to R7 ReaLearn forgot all the input/output devices for different assignments. AND some of my device IDs changed in Prefs, requiring me either to edit scripts with those IDs or manually change the IDs back.

Anyways, just in case anyone else is having the same problem.

this is good to know.... as i tend to be cautious around updating, when there is no immediate and/or major benefit to carrying out an update.
thanks all the same for the feedback

[Cloudswim]

Originally Posted by 7enz

i do wonder if your controller supports feedback or not??
in your mappings
normally you would have both ticks selected which are normally in the centre... one for control & the other for feedback... if the feedback is unticked then there is no feedback of course, thus you would get parameter jumps
im unfamiliar with the capabilities of the loupedeck...and a big question mark over whether it supports feedback

feedback should not be relevant to the actual control.

I got it working, the Catch Up mode works pretty well with the controller.

[Cloudswim]

Had anyone had experience with Global Mouse , Left click ?

I tried using the fire after X ms, but when it’s triggered there is only Press and no release.

How should I set the buttons ? It’s just a midi CC click button

[7enz]

Originally Posted by Cloudswim

Had anyone had experience with Global Mouse , Left click ?

I tried using the fire after X ms, but when it’s triggered there is only Press and no release.

How should I set the buttons ? It’s just a midi CC click button

if you use a midi-note as a trigger its makes effective use of hold until released... if you are using midi-CC then the button needs to have gate type function which mimics a type of midi-note

[TomInBrisbane]

EDIT:Thank you to 7enz for the suggestions in the following post. Researching his method led me to understand the use of bank selectors instead of modifiers. This method is documented in the ReaLearn docs but I didn't understand the significance at the time.

I wish to redefine the arm, solo, mute and select buttons for each track. I will use three (or more) modifier keys to enable other track-based functionality. For example, shift-arm to set automation mode to latch, control-arm to toggle track effect #1 on/off, alt-arm to change the monitor mode, and so on. (There might be other modifier buttons and/or combinations to choose from.)

What seems to be missing is the ability to express how the button should behave in the total absence of any modifier key. I can only express a default function by setting active : when modifiers on/off ; Modifier A : off; Modifier B : off. This appears to preclude the ability to use more than two modifiers. What I wish to express is "active: When all modifiers are off".

Perhaps there's another way to do this through expressions but I don't see how that can scale as new modifiers are added later. Perhaps testing that all modifiers are off....

[7enz]

Originally Posted by TomInBrisbane

I wish to redefine the arm, solo, mute and select buttons for each track. I will use three (or more) modifier keys to enable other track-based functionality. For example, shift-arm to set automation mode to latch, control-arm to toggle track effect #1 on/off, alt-arm to change the monitor mode, and so on. (There might be other modifier buttons and/or combinations to choose from.)

What seems to be missing is the ability to express how the button should behave in the total absence of any modifier key. I can only express a default function by setting active : when modifiers on/off ; Modifier A : off; Modifier B : off. This appears to preclude the ability to use more than two modifiers. What I wish to express is "active: When all modifiers are off".

Perhaps there's another way to do this through expressions but I don't see how that can scale as new modifiers are added later. Perhaps testing that all modifiers are off....

i find that toggling(so it stays on) enable/disable mappings allows for modifier keys to be within the next mapping


for instance
double press a button to activate first mapping layer
the new mapping layer has a different set of mappings plus another modifier which can access another enable/disable mapping...
whether you choose to use exclusive or non-exclusive needs to be though about when doing this type of mapping... as sometimes you require other mappings to be disabled so they will not conflict or activate 2 simultaneous actions
note: the use of tags come in handy when using enable/disable mappings

heres a video: which doesnt actually show the layered mappings... but the implementation of modifiers within mapping is straight forward when utilising the enable/disable mapping

https://www.youtube.com/watch?v=zDYT21fQFDk

[helgoboss]

Just a quick heads-up: There's a bug in the recently released 2.16.0-pre.2 which can let REAPER crash (I mean a hard crash, not just an error message) if you use the Lua-based "MIDI script" source or the Lua-based feedback script. Thanks to Robin for the bug report. It happens for example when duplicating mappings.

The upcoming release will contain a fix. Until then it's safer to stay with 2.16.0-pre.1 (and of course, if you want to be safest, don't use pre-releases at all

[vonglan]

Originally Posted by Perken

Hello Everyone,

I've been using some Lua and my Midi-fighter twister over the summer, trying to wrap my head around scripting mappings for Realearn. It's been a ton of fun, and I'd like to share my most ambitious project with you yet: a mapping to control VST synths from the MFT.

...

You could try to make this into a "controller mapping" for the ReaLearn SY concept https://github.com/vonglan/realearn-sy-presets .
Then you would get the main mappings for many virtual analog softsynths without having to map them manually.
(I made a controller mapping from the X-Touch Mini to the "SY" set of virtual parameters for analog synths, with 8 banks of 8 knobs, somewhat similar to what you did with the MFT.)

[Cloudswim]

Originally Posted by 7enz

if you use a midi-note as a trigger its makes effective use of hold until released... if you are using midi-CC then the button needs to have gate type function which mimics a type of midi-note

Thanks

[Cloudswim]

Hi I noticed that Realearn changes the mapping for a certain FX on a track when the order of the FXs has changed. Ie if I made some mappings to FX A and Moved it below FX B. FX B will now be controlled by the mappings for FX A. Is this the normal behaviour for realearn? Can I have realearn not to respect the order of the FX?

[helgoboss]

Originally Posted by Cloudswim

Hi I noticed that Realearn changes the mapping for a certain FX on a track when the order of the FXs has changed. Ie if I made some mappings to FX A and Moved it below FX B. FX B will now be controlled by the mappings for FX A. Is this the normal behaviour for realearn? Can I have realearn not to respect the order of the FX?

It depends on the FX selector. You need FX selector "Particular" or "Named" (if FX A and B have 2 different names), I guess you have another one right now ("By position")?

[helgoboss]

Released ReaLearn 2.16.0-pre.4 (ReaPack installation instructions)

Changes:
- Improved resource usage (by firing up threads for async logging and metrics only if enabled, by letting ReaLearn and Clip Engine metrics share the same thread, by making app and server share the same one-threaded async runtime and by stopping async deallocation thread and async runtime when last ReaLearn instance gone)
- Improved plug-in scanning speed (ReaLearn won't initialize a few things if it knows it's just being scanned anyway)
- Improved resource clean-up on Windows when REAPER preference "Allow complete unload of VST plug-ins" enabled
- #896 Fixed hard crash when duplicating mappings containing Lua scripts (2.16.0-pre.2 regression caused by changes related to async deallocation)
- Fixed possible audio dropouts due to deallocation in real-time threads when using EEL (2.16.0-pre.2 regression)
- Fixed possible panic when closing REAPER
- Internal change: Use Rust 1.73.0
- Internal change: Added counter of undesired real-time allocations to UI (debug builds only)
- Internal change: Added necessary changes to allow for Playtime early-access program

[Cloudswim]

Originally Posted by helgoboss

It depends on the FX selector. You need FX selector "Particular" or "Named" (if FX A and B have 2 different names), I guess you have another one right now ("By position")?

thanks! I noticed that when a particular mapping is learned for a new target, the default setting for "TRACK" and "FX" is always set at "AT POSITION" is there a way to change this behavior? <- I see there is a FR for this.

I'm trying to receive feedback on my controller for the name of parameter that the knob is currently controlling, is this possible and what would be the easiest way?

Is it possible to set MIDI input from one controller, and feedback of target parameter on another device? Is it possible to use TouchOSC as a display for target FX Parameter for each mapping. But the Source is not from TouchOSC rather from a different Controller.

I hope this make sense.

[jbraner]

Hi,

I'm new with Realearn and I'm missing something.

I've got a new synth (Current by Minimal Audio) and it doesn't have MIDI learn to map CCs to parameters - they tell you to "use your DAW".

I've set up Relearn for a few macro knobs (to get started), but I'm missing something basic.
I can set up the mappings with my MIDI controller, and can control the macro knobs, I can record the CC data into Reaper - but then I can't "play back" the CC data - the synth ignores it.

It's like I need something else between playing the MIDI CCs and the synth - to translate the MIDI CC info into the synth parameter for the macro 1 knob.

The manual says this is "pseudo automation" and the target must be set for "FX parameter: Set value" and it is (it picked this up from "learn").
The target synth is on the same track as Realearn
Relearn -> synth
This is on Reaper 7.05 - so it should work.

The manual says "It is possible to feed ReaLearn with track MIDI items instead of live MIDI data. This results in a kind of pseudo
automation. Some users call this MIDI CC based automation. This feature can be quite interesting and appealing
to MIDI fans."

Does anyone know what I'm missing?

I can set a Reaper automation envelope for the parameter - and play it back, but how do I make Realearn translate the MIDI CC info into this synth parameter - for playback?

[7enz]

Originally Posted by jbraner

Hi,

I'm new with Realearn and I'm missing something.

I've got a new synth (Current by Minimal Audio) and it doesn't have MIDI learn to map CCs to parameters - they tell you to "use your DAW".

I've set up Relearn for a few macro knobs (to get started), but I'm missing something basic.
I can set up the mappings with my MIDI controller, and can control the macro knobs, I can record the CC data into Reaper - but then I can't "play back" the CC data - the synth ignores it.

It's like I need something else between playing the MIDI CCs and the synth - to translate the MIDI CC info into the synth parameter for the macro 1 knob.

The manual says this is "pseudo automation" and the target must be set for "FX parameter: Set value" and it is (it picked this up from "learn").
The target synth is on the same track as Realearn
Relearn -> synth
This is on Reaper 7.05 - so it should work.

The manual says "It is possible to feed ReaLearn with track MIDI items instead of live MIDI data. This results in a kind of pseudo
automation. Some users call this MIDI CC based automation. This feature can be quite interesting and appealing
to MIDI fans."

Does anyone know what I'm missing?

I can set a Reaper automation envelope for the parameter - and play it back, but how do I make Realearn translate the MIDI CC info into this synth parameter - for playback?

dont know if this will help, for what you are asking about...
but do you have the midi-out on Reapers TCP/MCP routing button set to where you want it to go??

[jbraner]

That's just for output to hardware - I just want the MIDI to go back through ReaLearn (changing CC#42 to the macro1 knob) to play the synth.

Maybe I have to do the parameter modulation in Reaper - in which case I don't know if there's any need for ReaLearn. Realearn is just so much easier to set up (but no good to me if it won't work)

I'll have more time tomorrow to try this some more.

I tried the VST2.4 version of the synth, thinking it was a VST3 issue - but got the same results.

This is usually simple to do - and you just "learn" what MIDI CC# to control each parameter right in the synth. This is the first synth I've had which doesn't do this for you.

I'll get there...

[7enz]

@jbraner
so the automation lane whilst in playmode on reaper feeds the synth?
thats what the midi output is for??
unless you are wanting
Automation ---> realearn ---> synth
instead of
Automation ---> midi-out ---> synth

[jbraner]

Yeah I want Reaper to play -> send CC42 through Realearn -> to the synth.
Realearn sits in the FX bay with the synth below it (and nothing else)

EDIT - hmm Relearn is expecting input from the MIDI controller isn't it? Rather than the CC42 data from Reaper.

Maybe this won't work.

[7enz]

Originally Posted by jbraner

Yeah I want Reaper to play -> send CC42 through Realearn -> to the synth.
Realearn sits in the FX bay with the synth below it (and nothing else)

EDIT - hmm Relearn is expecting input from the MIDI controller isn't it? Rather than the CC42 data from Reaper.

Maybe this won't work.

i suppose you could test by trying to set up a midi-mapping of the CC value the automation is using then send the feedback out to CC42
you would need to setup a separate realearn instance and set the input and output to

input - virtual midi port
output - synth

and then you would set the midi-out(inside routing button) port on the track with the automation to output to the virtual port
im assuming you are wanting a conversion process of the automated FXparameter CC values to be converted to CC42 with the corresponding values

so basically
reaper automation FX CC# (which is output to) --> virtual midi port --> realearn(convert FX automation CC+values --> CC42+values) out --> synth

[jbraner]

Thank you 7enz - I'll try that.
I had just assumed that ReaLearn was going to do this simply - but I now realise that ReaLearn is meant more for just *playing* the the soft synth (ie live) than *playing back* the MIDI from Reaper.


here's something really weird - I actually found this old post (in this thread) from 5 years ago
https://forum.cockos.com/showthread.php?t=178015&page=7
and it's from me!!


I must have played with this when I was first learning Reaper - and I completely forgot about it.


So, I'll spend some time this week working out how to set up my MIDI controller and soft synth to play back <and render correctly> from Reaper.


I'm sure one of the combinations will work - and I'll report back.

[jbraner]

OK 7env's solution works (the settings are a tiny bit different)

*BUT* the track doesn't render the MIDI modularion (which is the topic of that original discussion)

I'm getting this to work (including rendering properly) by using the Reaper setting "Parameter Modulation/MIDI Link" to set the CC# for each parameter.
I'm not sure if ReaLearn is even necessary now.

I'll have to wait until tomorrow to spend more time with this.

[helgoboss]

Originally Posted by jbraner

OK 7env's solution works (the settings are a tiny bit different)

*BUT* the track doesn't render the MIDI modularion (which is the topic of that original discussion)

I'm getting this to work (including rendering properly) by using the Reaper setting "Parameter Modulation/MIDI Link" to set the CC# for each parameter.
I'm not sure if ReaLearn is even necessary now.

I'll have to wait until tomorrow to spend more time with this.

ReaLearn supports rendering only under special circumstances. The details are in the docs.

[jbraner]

Originally Posted by helgoboss

ReaLearn supports rendering only under special circumstances. The details are in the docs.

Hi helgoboss,
Yeah I tried what the manual says but it doesn't work for me.
No worries as I can get this working in Reaper
I was just mistaken about how Realearn works.



UPDATE: OK, this works fine without any need for ReaLearn (controlling soft synth parameters with a MIDI controller when the synth doesn't do the mapping for you) - just using Reaper. Of course ReaLearn is a lot easier to set up - but it's not the right tool for the job
Sorry to interrupt this thread - it was my misunderstanding...

[Cloudswim]

Hi I noticed that if a mapping of Global Last Touched exists in a Realearn instance, the ReaLearn: Find first mapping by target action will always locate the GLOBAL LAST TOUCHED as the mapping and not the mapping initially mapped to the source.

Is there a way around this?