I guess I'm not sure yet what you are getting at. Do you want to be able to not just e.g. select 1 track exclusively within a folder, but maybe a group of 2 or more tracks ("multi selection")? If this is what you want, I have another idea: I could add a new option for "Exclusive" named "Within folder, including nested tracks". With such an option you could make all the tracks that you want to be selected at once into one folder. Working on folders should be more stable and clean than affecting previous/next tracks.
Well don't hesitate to call me stupid if it's not the way realearn should work, but what I have in mind is for example 2 folders, one folder per performer, each one playing with his dedicated midi controller. One selected track per folder wich is the chosen sound at the moment. English is kind of making a neurons mixed salad in my brain atm, so not sure "Within folder, including nested tracks" would allow that...
Benefice of track selection would be using Auto-Arm (but not a necissity due to your recent within folder addition) and triggering realearns using corresponding <selected track> option.
Quote:
I added the following FRs:
- #184 (most important requirement for this scenario)
- #191 (2nd most important one)
Well don't hesitate to call me stupid if it's not the way realearn should work, but what I have in mind is for example 2 folders, one folder per performer, each one playing with his dedicated midi controller. One selected track per folder wich is the chosen sound at the moment. English is kind of making a neurons mixed salad in my brain atm, so not sure "Within folder, including nested tracks" would allow that...
Benefice of track selection would be using Auto-Arm (but not a necissity due to your recent within folder addition) and triggering realearns using corresponding <selected track> option.
Thank you!
Wait a moment, now I'm confused ... what you describe is exactly what the already existing option "Exclusive: Within folder" does. It's available for all kinds of targets, also track selection. What am I missing here?
I am using OSCIIBot to have Reaper (i.e. JSFXes I installed in Reaper) communicate with an XR18 Rack mixer via OSC.
My JSFXes communicate with the wolrd outside Reaper by Midi CC messages, and (connected to Reaper via LoopMidi) OSCIIBot converts those from/to OSC packets received from / to be sent to the Mixer.
Now I learned that ReaLearn can bidirectionally communicate via OSC (usually with Control Surface devices).
Is there a chance to use it as a midi/OSC converter so that I can get rid of the external tools LoopMidi and OSCIIBot ?
I am using OSCIIBot to have Reaper (i.e. JSFXes I installed in Reaper) communicate with an XR18 Rack mixer via OSC.
My JSFXes communicate with the wolrd outside Reaper by Midi CC messages, and (connected to Reaper via LoopMidi) OSCIIBot converts those from/to OSC packets received from / to be sent to the Mixer.
Now I learned that ReaLearn can bidirectionally communicate via OSC (usually with Control Surface devices).
Is there a chance to use it as a midi/OSC converter so that I can get rid of the external tools LoopMidi and OSCIIBot ?
-Michael
ReaLearn is not designed as a converter between OSC and MIDI at the moment. What it can do: Using incoming OSC messages to control targets ("=>" control direction) and sending OSC messages in response to parameter changes in REAPER ("<=" feedback direction). Now, if your JSFX is just a means to control some parameter in REAPER, then you could be in luck - because then you might be able to skip JSFX, LoopMidi and OSCIIBot altogether. If it does other things, then maybe not. So the answer to your question is: It depends on what those JSFX effects do.
If I would add a "Send MIDI" target (which is planned), I guess ReaLearn could act as a simple OSC to MIDI converter. If I would add a "Send OSC" target (which is not planned but ... why not, would be consequent), then the other direction would also be possible.
Sounds great !
The current usage case is like that:
With my XTouch Compact I use layer "A" to do a very complex interaction with Reaper to create an "Instrument" playable with 3 keyboards etc. Of course with appropriate feedback to the Control Surface.
Layer "B" is used to control the XR18, simply passing a bunch of faders/rotaries/knobs to the mixer, no complex algorithms involved The values are just filtered / routed as CCs, not stored in JSFXes. Feedback is used here, too, as The XR18 also is controlled by one ore more PCs or Android phones, and all stays in sync.
-Michael
P.S.1:
I once filed a Feature request to add OSCIIBot to Reaper and have it create a Midi Device, so that it can be accessed by JSFXes in tracks.
I now see that this could (rather easily ??) be done as a Reaper extension (in case a Midi device can be created via the API) or as a Reaper aware VSTi (it seems ReaLearn internally provides all which is necessary: Midi Communication by Track I/O, OSC communication with the outside world, EEL handling.)
Are you triggered by that ?!?!?
P.S.2:
It's a known problem with OSCIIBot that it always does OSC Bundling for messages sent directly after one another. There is a new fork that makes this configurabel.
The XR18 and other Behringer gear seem to feature a bug and ignore any bundle OSC messages.
What OSC library do you use ? Can bundling be disabled with ReaLearn ?
P.S.1:
I once filed a Feature request to add OSCIIBot to Reaper and have it create a Midi Device, so that it can be accessed by JSFXes in tracks.
I now see that this could (rather easily ??) be done as a Reaper extension (in case a Midi device can be created via the API) or as a Reaper aware VSTi (it seems ReaLearn internally provides all which is necessary: Midi Communication by Track I/O, OSC communication with the outside world, EEL handling.)
Are you triggered by that ?!?!?
I think it could indeed be done rather easily. Also sounds useful and I love OSCIIBot, but currently I have a bit much work to do with ReaLearn. What I could do though is providing mentioned MIDI and OSC target. Should be enough to translate one parameter per mapping. It won't be a full-blown alternative to OSCIIBot because you cannot program every aspect of it and you have the decoupling between source and target which is typical for ReaLearn.
Quote:
Originally Posted by mschnell
P.S.2:
It's a known problem with OSCIIBot that it always does OSC Bundling for messages sent directly after one another. There is a new fork that makes this configurabel.
The XR18 and other Behringer gear seem to feature a bug and ignore any bundle OSC messages.
What OSC library do you use ? Can bundling be disabled with ReaLearn ?
I use rosc for converting raw byte sequences to and from OSC messages/bundles and the UDP connection part is self-implemented ... so yes, I can disable bundling. I didn't know this makes problems with some devices, just tested it with TouchOSC which is fine with it. I use bundles because this is what REAPER's own OSC implementation does, too. I created an FR.
I've tried to get compartments going. My understanding is that you:
1. Set up a generic Controller mapping with a source for each controller - my encoder 1 for example is set as Channel 16, CC35, range element.
2. Leave the target as, for example, Virtual Multi 1. When finished, save the preset.
3. Now set up a Main mapping. Set each source to Category - Virtual, Type (eg) Multi, Number 1(Enc1).
4. Give each source a target in Reaper. Then save preset.
I cannot get this to work. Whereas if I set the source to Category - Midi, it works fine. Maybe my control inputs are wrong? Is this explained in the manual?
Brilliant piece of software btw.
__________________
Mac Pro 3.1, Emagic AMT8, iconnectivity Audio 4+, Zoom R24, Novation Mininova.
Just a 'in action' report:
I did a new live Dub setup yesterday using ReaLearn and my BCR.
Mass mapping was a breeze with the 'learn many' function (move source, move target, repeat), well done.
Seems all running smoothly so far (but haven't used it really 'live' i.e. in a rehearsal yet).
I've tried to get compartments going. My understanding is that you:
1. Set up a generic Controller mapping with a source for each controller - my encoder 1 for example is set as Channel 16, CC35, range element.
2. Leave the target as, for example, Virtual Multi 1. When finished, save the preset.
3. Now set up a Main mapping. Set each source to Category - Virtual, Type (eg) Multi, Number 1(Enc1).
4. Give each source a target in Reaper. Then save preset.
I cannot get this to work. Whereas if I set the source to Category - Midi, it works fine. Maybe my control inputs are wrong? Is this explained in the manual?
Brilliant piece of software btw.
That's about it, yes. Could you paste the output of "Export to clipboard" or send it or the RPP file to info@helgoboss.org? The process is documented in detail in the "Controller mappings" section in the user guide.
Quote:
Originally Posted by nofish
Just a 'in action' report:
I did a new live Dub setup yesterday using ReaLearn and my BCR.
Mass mapping was a breeze with the 'learn many' function (move source, move target, repeat), well done.
Seems all running smoothly so far (but haven't used it really 'live' i.e. in a rehearsal yet).
Glad it worked! "Learn many" should work especially well if you are work with a controller preset (that contains mappings with virtual targets) - because then ReaLearn doesn't have to figure out the encoder types anymore, which usually means less manual adjustment afterwards. But probably you already did that.
"That's about it, yes. Could you paste the output of "Export to clipboard" or send it or the RPP file to info@helgoboss.org? The process is documented in detail in the "Controller mappings" section in the user guide."
Rebooting everything and suddenly it worked. Oh well. At least I was doing it correctly.
May I ask about the next step. When I press the Projection button, and I press the pencil button in the browser, I get the controls appearing on the right hand side of the browser, but they are labelled with the names set in the Controller mapping not the Main one. But it's the main mapping that will have the specific mappings to proper targets that need to be named accordingly to "pan 1" or "latch mode" or something like that - and so that's what needs to be fed through to the browser?
__________________
Mac Pro 3.1, Emagic AMT8, iconnectivity Audio 4+, Zoom R24, Novation Mininova.
May I ask about the next step. When I press the Projection button, and I press the pencil button in the browser, I get the controls appearing on the right hand side of the browser, but they are labelled with the names set in the Controller mapping not the Main one. But it's the main mapping that will have the specific mappings to proper targets that need to be named accordingly to "pan 1" or "latch mode" or something like that - and so that's what needs to be fed through to the browser?
In edit mode, they are labeled with the controller mapping names (so that you know which control element is which so you can position them correctly) and in the normal mode they are labeled with the main mapping names (so that you know what is controlled).
If I would add a "Send OSC" target (which is not planned but ... why not, would be consequent), ...
Independent of this: In the forums I already read several question about having Reaper send out custom defined OSC messages, and there was no solution. ReaLearn would provider this
Independent of this: In the forums I already read several question about having Reaper send out custom defined OSC messages, and there was no solution. ReaLearn would provider this
-Michael
I never really got into that but I thought the pattern configs of REAPER's built-in OSC control surface can send arbitrary messages? Maybe not.
Please don't get the expectations too high The "Send OSC" target that I have in mind will not be useful for sending arbitrary messages. The way I would implement it as a first step would be similar to the OSC source: By providing an address field that allows you to enter a fixed address, e.g. "/bank1/fader0" and by providing an argument type field which lets you choose the data type according to the OSC spec (float, double, bool, ...). The argument position would be always the first one for now (I don't see yet how I could support transmitting multiple arguments in a reasonable way because one ReaLearn mapping transmits one value only by design). The OSC output device would be the one configured in "Feedback output".
However, apart from the "Send OSC" target I will probably add a way to send an arbitrary number of arbitrarily formed OSC messages on mapping lifecycle events such as activation and deactivation. This is already possible for MIDI in upcoming version 2.5.0 (also sys-ex) so it would make sense to provide this for OSC as well.
Quote:
Originally Posted by mschnell
It rather obviously is a bug in the implementation by Behringer and seems to affect all of there devices.
OTOH, OSC bundle only makes sense to assign a time tag to a bunch of OSC mesdsages. I don't suppose ReaLearn ever will need to do that.
-Michael
True, the time tag doesn't make any sense for ReaLearn. At the moment i use the "Immediately" time tag. I figured that it also could help in terms of performance to send bundles but probably not that much because we are talking about UDP, not TCP. I will provide a checkbox "Send as bundles" for the OSC device which can be unticked. ReaLearn itself can deal both with single messages and bundles. It ignores the time tag.
"Learn many" should work especially well if you are work with a controller preset (that contains mappings with virtual targets) - because then ReaLearn doesn't have to figure out the encoder types anymore, which usually means less manual adjustment afterwards. But probably you already did that.
The way I would implement it as a first step would be similar to the OSC source: By providing an address field that allows you to enter a fixed address, e.g. "/bank1/fader0" and by providing an argument type field which lets you choose the data type according to the OSC spec (float, double, bool, ...). The argument position would be always the first one for now (I don't see yet how I could support transmitting multiple arguments in a reasonable way because one ReaLearn mapping transmits one value only by design). The OSC output device would be the one configured in "Feedback output".
Hi,
Today I am exploring how to set up an X/Y pad in Realearn. I've set up two virtual faders "X-axis" and "Y-axis" in the Controller mapping and then set them up in the main mapping to control something in Reaper.
In the Projection I've set up the objects as two 3x3 squares, with the x box completely covered by the y box. The x label is on the top of the resulting box and the y label down the side.
It looks like an X/Y pad which is pleasing, but I am wondering if I will run into trouble if/when you get the Projection to show values?
__________________
Mac Pro 3.1, Emagic AMT8, iconnectivity Audio 4+, Zoom R24, Novation Mininova.
Hi,
Today I am exploring how to set up an X/Y pad in Realearn. I've set up two virtual faders "X-axis" and "Y-axis" in the Controller mapping and then set them up in the main mapping to control something in Reaper.
In the Projection I've set up the objects as two 3x3 squares, with the x box completely covered by the y box. The x label is on the top of the resulting box and the y label down the side.
It looks like an X/Y pad which is pleasing, but I am wondering if I will run into trouble if/when you get the Projection to show values?
I thought about visualizing the current value by filling the rectangle to n % with some half-transparent color. The circles maybe in some pizza piece style.
So yes, it can happen that it won't look as you desire. But I also plan to make it possible to disable value feedback, globally on the particular mobile device or for specific control elements within the controller preset.
Changes:
- #113 Added foundation for setting advanced aspects of mappings via textual configuration (accessible via new mapping panel button "Advanced settings")
- #113 Added foundation for using very device-specific features by allowing to send a bunch of raw MIDI data (including SysEx!) on mapping activation/deactivation (see user guide section "Advanced settings" for details)
- #196 Added option "Can deal with OSC bundles" to OSC device popup menu which can be unticked in order to achieve compatibility with devices not supporting them (e.g. Behringer mixers)
- Added user guide entries and controller presets for PreSonus FaderPort 8 and FaderPort Classic
- #182 Improved reliability by avoiding memory exhaustion in the presence of potential programming errors (internally use bounded channels instead of unbounded one)
- Improved user guide and projection instructions by explaining how to open a context menu
- #193 Fixed possible folder overflow with the "Exclusive: Within folder" setting introduced in the previous release
- #185 Fixed non-occurring mapping row target label updates when track list changes
- #194 Fixed sometimes non-arriving MIDI feedback on project load
Wow, the "FaderPort Classic" is the controller with the most weird MIDI spec I have seen so far! But with the added preset it's now completely ReaLearn-compatible and should just work. I think this particular controller preset is a good example of how powerful ReaLearn's concepts of virtual controls and control/feedback separation is. It also uses the new feature to be able to send raw MIDI data on mapping/controller activation.
The "Advanced settings" button doesn't seem to work yet on macOS and maybe also not Linux.
I posted this in another forum but got no fix.... maybe you can help?
Quote:
Originally Posted by buschag
What I’m trying to do to speed things up is adjust selected envelope points with an encoder. I have an action that creates 4 points with the middle two selected. After that I would like to move the two points either up or down. I searched and found a custom action along the lines of: skip if <0 cc , move selected points up, skip if >0cc, move selected points down. That’s not exactly it but I’m not at my computer... I’ve found it in these forums. What happens is, when I assign an encoder to the custom action, it only goes the direction I first move the encoder. So if I start by raising it, it only goes up, regardless if I turn the encoder the other way and vice versa. I’ve tried absolute, all relative modes, realearn. Am I missing something? Thanks
Oddly enough the encoder only works in range element or button options. When trying all three relative types, it jumps around and sometimes moves opposite. Xtouch mini endless encoder
Oddly enough the encoder only works in range element or button options. When trying all three relative types, it jumps around and sometimes moves opposite. Xtouch mini endless encoder
Then your encoder is configured to transmit absolute messages instead of relative ones, which is far from ideal (I assume you want relative control = without jumps and with adjustable step sizes). I guess the X-Touch Mini uses the same config tool as the X-Touch Compact, so here's what you need to do. After that check if setting the track volume works with ReaLearn's "Encoder (relative type 1)". Before this basic thing doesn't work, looking into more complicated stuff is kind of hopeless.
Ok I saw the compact instructions and I only have the two modes as well. As far as I understand, the xtouch editor is a windows only program. So it looks like I’ll have to switch back to mc mode or whatever. Because it appears I can’t switch the encoders to relative in standard midi mode. Is that a fair assessment? Will report back later or tomorrow. Appreciate the help as always.
Ok I saw the compact instructions and I only have the two modes as well. As far as I understand, the xtouch editor is a windows only program. So it looks like I’ll have to switch back to mc mode or whatever. Because it appears I can’t switch the encoders to relative in standard midi mode. Is that a fair assessment? Will report back later or tomorrow. Appreciate the help as always.
Windows only! Although I prefer that over Mac only, I still think that's a shame.
Yes, fair assessment, try MC mode. This has the encoders set to relative control for sure.
Wait a moment, now I'm confused ... what you describe is exactly what the already existing option "Exclusive: Within folder" does. It's available for all kinds of targets, also track selection. What am I missing here?
Hmmm Sorry for that I'll test that workflow as soon as I can!
Cheers
So it works in the envelope lane as it should, thanks by the way, but I was hesitant because in the mapping editor window, the 0-100% slider on the right didn’t move. I got it to work by using source knob/fader and target relative. Didn’t work when I used type 1 relative encoder. Hope this makes sense
So it works in the envelope lane as it should, thanks by the way, but I was hesitant because in the mapping editor window, the 0-100% slider on the right didn’t move. I got it to work by using source knob/fader and target relative. Didn’t work when I used type 1 relative encoder. Hope this makes sense
You mean Source Character "Range element"? In that case, you are still absolute, not relative. There's a very simple way to check if your control sends absolute or relative messages: Open up ReaControlMIDI and show the log. If you see increasing values when turning your encoder clockwise, you are absolute. If you see more or less repeatedly the same value, you are relative.
That the slider doesn't move is probably related to the fact that you are triggering actions. Many actions don't support feedback. That's explained somewhere in the user guide in detail.
Changes:
- #47 Added option "Behavior" for "Track solo" target, start using the routing-friendly "Solo in front" behavior by default
- #162 Improved "Track solo" target by making it work with "Options → Solo in front"
Thanks to the explanations in post #685, i was able to change from using several instances of realearn to just one with the groups activation conditions.
Nice !
Quote:
Originally Posted by Arthur McArthur
Yes, tried the InControl/Extended mode but it doesn't solve the issue. When I change the pan, the pad's button does in fact light up. I'll go ahead and throw a smiley face on the pile
Let's hope more people will join in for those linked mappings with common feedback.
Last edited by FarfadetFarfelu; 03-03-2021 at 03:19 PM.
Thanks to the explanations in post #685, i was able to change from using several instances of realearn to just one with the groups activation conditions.
Nice !
That's also explained in the intro video from post #1. Y'all should check it out. Chock full of helpful info.
This thing can do things no one's thought of yet, but helgo will give you many ideas in the vid.
EDIT: My bad, it's not linked there directly, but here you go:
So I used my computer at work to edit the xtouch mini because I only have a Mac. I switched the encoders to relative 1. Hopefully that is correct and I’ll try it when I get home tonight.
I putted my old Faderport Classic into use again - wanted to test ReaLearn with it!
Downloaded the Faderport Classic preset and presto, ReaLearn recognize it!
Now I am trying to let the fader control a FX parameter and thats works also.
BUT, I have no clue how to make the fader "jump to" next or previous FX parameter.
I would like to have two buttons on my Faderport Classic to step thru the
FX parameters in the plugin so my fader can control them (separately).
Is there a way to do this or am I asking for to much?
I putted my old Faderport Classic into use again - wanted to test ReaLearn with it!
Downloaded the Faderport Classic preset and presto, ReaLearn recognize it!
Now I am trying to let the fader control a FX parameter and thats works also.
BUT, I have no clue how to make the fader "jump to" next or previous FX parameter.
I would like to have two buttons on my Faderport Classic to step thru the
FX parameters in the plugin so my fader can control them (separately).
Is there a way to do this or am I asking for to much?
Regards
Tompad
This is currently not possible unless you are willing to create one mapping per FX parameter (that you want to control) and use program-based conditional activation to only ever have one of these mappings active. But that's too much work if you don't generate these mappings using some script or Excel.
However ... I think this question of yours is the final trigger to make me implement a feature that lets ReaLearn determine the target track, target FX or target FX parameter (and maybe other stuff) dynamically based on the value of an internal ReaLearn parameter. It's something I have been thinking about for a while already but I was hesitant to implement it because I wanted to see if user-generated mappings would be a good alternative (see this post). I can see now, especially for your use case, one would have to generate potentially hundreds of mappings, for each track ... that's just not practical, not even when using a generator.
Implementing this feature shouldn't be very hard because almost everything is there already. It would instantly open up a realm of new possibilities and greatly simplify many setups. So ... I suggest you to wait a bit longer (~1 week if nothing interferes with my plan) before investing any effort into alternatives.
Once the feature is finished, you will be able to achieve your goal by doing this:
Assign the Faderport's "Previous" and "Next" buttons to one of ReaLearn's internal freely assignable parameters, let's say "Parameter 1" (using 2 "Incremental buttons" mode mappings, one for each button ... this is possible already now).
Assign your fader to an FX parameter target (using 1 mapping). You would set the "FX parameter" to the new value "<Dynamic>". As a result, a small text field will show up which lets you enter a formula that determines the desired parameter based on internal ReaLearn parameter values. E.g. entering "100 * p1" would have the effect that the controlled parameter will always be bound to the current value of parameter 1.
This is currently not possible unless you are willing to create one mapping per FX parameter (that you want to control) and use program-based conditional activation to only ever have one of these mappings active. But that's too much work if you don't generate these mappings using some script or Excel.
However ... I think this question of yours is the final trigger to make me implement a feature that lets ReaLearn determine the target track, target FX or target FX parameter (and maybe other stuff) dynamically based on the value of an internal ReaLearn parameter. It's something I have been thinking about for a while already but I was hesitant to implement it because I wanted to see if user-generated mappings would be a good alternative (see this post). I can see now, especially for your use case, one would have to generate potentially hundreds of mappings, for each track ... that's just not practical, not even when using a generator.
Implementing this feature shouldn't be very hard because almost everything is there already. It would instantly open up a realm of new possibilities and greatly simplify many setups. So ... I suggest you to wait a bit longer (~1 week if nothing interferes with my plan) before investing any effort into alternatives.
Once the feature is finished, you will be able to achieve your goal by doing this:
Assign the Faderport's "Previous" and "Next" buttons to one of ReaLearn's internal freely assignable parameters, let's say "Parameter 1" (using 2 "Incremental buttons" mode mappings, one for each button ... this is possible already now).
Assign your fader to an FX parameter target (using 1 mapping). You would set the "FX parameter" to the new value "<Dynamic>". As a result, a small text field will show up which lets you enter a formula that determines the desired parameter based on internal ReaLearn parameter values. E.g. entering "100 * p1" would have the effect that the controlled parameter will always be bound to the current value of parameter 1.
Hehe - I found the information about program-based conditional activation in ReaLearn help file
and tried it....and I just gave up because as you say - it will be MANY mappings to do :-)
Reading your answer to my question makes me dance! It would be super-duper-great if
this feature could be added to this amazing work! THANK YOU!
I’m having difficulty getting OSC Pilot ( https://oscpilot.com/ ) to work with ReaLearn.
I’m wondering if this is due to the fact that OSC Pilot uses a Local Port as opposed to a unique device ip?
I’ve got Touch OSC working just fine(!)
In Reaper’s Preferences / Control/OSC/web I can receive and map messages from OSC Pilot.
When I add OSC Pilot to ReaLearn, I get <unable to detect>
When I:
remove OSC Pilot from Preferences / Control/OSC/web
go into Relearn and open the settings for OSC Pilot ( OSC Devices/OSC Pilot/Edit… )
close the settings
<unable to detect> no longer appears and I can receive OSC messages via the Learn function however none of the Target mappings work (unlike TouchOSC).
I’ve held off diving into OSC until ReaLearn supported it and I’m glad it does!
Just gotta see if I can get OSC Pilot working...
I’m having difficulty getting OSC Pilot ( https://oscpilot.com/ ) to work with ReaLearn.
I’m wondering if this is due to the fact that OSC Pilot uses a Local Port as opposed to a unique device ip?
I’ve got Touch OSC working just fine(!)
In Reaper’s Preferences / Control/OSC/web I can receive and map messages from OSC Pilot.
When I add OSC Pilot to ReaLearn, I get <unable to detect>
When I:
remove OSC Pilot from Preferences / Control/OSC/web
go into Relearn and open the settings for OSC Pilot ( OSC Devices/OSC Pilot/Edit… )
close the settings
<unable to detect> no longer appears and I can receive OSC messages via the Learn function however none of the Target mappings work (unlike TouchOSC).
I’ve held off diving into OSC until ReaLearn supported it and I’m glad it does!
Just gotta see if I can get OSC Pilot working...
Changes:
- #5 Added experimental and feedback-enabled (!) "Go to marker/region" target with smooth seek for regions
- #198 Added very flexible feedback-only "Raw MIDI" source in order to support feedback on controllers whose LEDs etc. can only be controlled via system-exclusive MIDI messages (e.g. Arturia MiniLab mkII)
- Added user guide entry and controller preset for Arturia MiniLab mkII (pad feedback now works thanks to ReaLearn's new sys-ex support!)
- Added user guide entry for OSC/PILOT
- #197 Improved usability of min/max text fields on Windows and macOS by changing the value (and correcting the other value) not before leaving the field instead of immediately (the behavior before that was downright annoying as you can see here)
- #113 Improved editing of advanced mapping settings by opening a dedicated small edit window with instant error checking (instead of just opening the system text editor, which sometimes doesn't work)
- #201 Fixed possible "Full" error message when having projects with very many tracks
- #199 Fixed possible error message with "Master tempo" target when mapping panel open
With this release, ReaLearn for the first time conquers REAPER's timeline (*Playtime glancing at ReaLearn slightly nervously*).
The "Send OSC" target that I have in mind will not be useful for sending arbitrary messages. The way I would implement it as a first step would be similar to the OSC source: By providing an address field that allows you to enter a fixed address, e.g. "/bank1/fader0" and by providing an argument type field which lets you choose the data type according to the OSC spec (float, double, bool, ...). The argument position would be always the first one for now (I don't see yet how I could support transmitting multiple arguments in a reasonable way because one ReaLearn mapping transmits one value only by design). The OSC output device would be the one configured in "Feedback output".
Linear Mode
