If lanes are heading towards improving the comping system...

[ferropop]

...maybe we should collect ideas on what the most useful direction might be, to give devs some data of where Reaper users hope it's heading on this super sensitive (and subtly complex) subject.

[ferropop]

My thoughts on it :

The big Reaper comping problem is Splits Hell and Lack of Independence of takes. Lanes has the potential to seriously solve both issues, if certain conditions are there. It very easily actually leads us to Swipe Comping like other modern DAWs now that we have Razor Edits.

1. Takes inside items should be replaced by Lanes inside items.

2. The lanes inside the items should be fully independent, and their content should remain UNAFFECTED by Splits in the Parent Item. Takes stay whole, unless you split them at the Take level where it's actually meaningful. Again -- splits in the parent item do not split the takes inside! But you can split the takes at the lane level independent of the Item splits. Stretch, copy/paste within the lane, etc., as the take stays whole.

3. Razor Edits should be the basis of comping, with suitable actions to Set Area as Active Take, etc. REs are practically made for this - and historically they actually grew out of comping, which funny enough they're specifically not usable for yet in Reaper

4. Razor Edits therefore need to operate inside Take Lanes at the Lane level.

5. When recording passes overlap, takes are created and the bounds of the entire recording pass completely supersede any underlying splits.

This means, if there are 7 splits in an item, and you record overtop those splits, they all disappear and a pair of splits is created instead at the exact new-recording bounds and the recording pass is added to it. That then should become the active take automatically. (Just as Reaper currently does, but with infinitely less split mess).


=== if REs were leveraged for the task, and lanes inside items existed as I described above, we'd have clean swipe-comping type behaviour, and an incredible minimization of unnecessary splits.

[Vagelis]

Originally Posted by ferropop

1. Takes inside items should be replaced by Lanes inside items.

This should be done when the container item is ready, it was introduced a while back for one or two builds and then didn't show up again. Cubase had this feature, you could add items inside the container item, and you could add multiple lanes. If we had this in Reaper would be great to be able and comp inside the container item.

From the lanes in Reaper, i expect to have an option to automute the overlapping item parts from lanes, the mute should be relative to the length of the item from the top or bottom lane.
The exact same system was also in cubase, (the way you add lanes from lane adder area) and automute overlapping parts and was GREAT. I really hope they add this feature.
But would be also nice to have an alternative way to comp with razor edit too.

[paaltio]

Hoping this leads to being able to at least recreate, if not surpass, the current Pro Tools editing workflow, where you can quickly step through the playlists (takes/lanes) and then select a section to be sent to the main playlist when you find something you like. And this happens on all the grouped tracks.

Given Reaper's track/item group division, not quite sure how that latter part should work. But I would really like to see this track-level group editing as core functionality somehow instead of a script.

[colin_maybe]

Originally Posted by paaltio

Hoping this leads to being able to at least recreate, if not surpass, the current Pro Tools editing workflow, where you can quickly step through the playlists (takes/lanes) and then select a section to be sent to the main playlist when you find something you like. And this happens on all the grouped tracks.

Given Reaper's track/item group division, not quite sure how that latter part should work. But I would really like to see this track-level group editing as core functionality somehow instead of a script.

Absolutely agree. It would be very very helpful to be able to work in groups at the track-level for any kind of lane based editing (and in general).

[Triode]

Originally Posted by ferropop

1. Takes inside items should be replaced by Lanes inside items.

.

I think the way it's heading is that the lanes are inside tracks. This is more compatible with takes inside items. If there is a new comping system that works with lanes inside tracks you could add to old projects using this new paradigm and it wouldn't break what you had before.

[Triode]

Originally Posted by colin_maybe

Absolutely agree. It would be very very helpful to be able to work in groups at the track-level for any kind of lane based editing (and in general).

For now I've got some scripts on reapack - search for track-based editing

[bolgwrad]

I've hardly used takes in twelve years, finding other ways to comp tracks. I use (pre-release versions) lanes all the time, with custom actions (solo item in track, etc) to make comping even easier.

[musicbynumbers]

Apart from a grouped item in lanes thing that Schwa experimented with and the good ideas already here...

I'd love to see a mute and solo button for each lane (by the number) as there's a few daws with this and they are useful for lots of things including some kinds of comping

[bolgwrad]

Originally Posted by musicbynumbers

Apart from a grouped item in lanes thing that Schwa experimented with and the good ideas already here...

I'd love to see a mute and solo button for each lane (by the number) as there's a few daws with this and they are useful for lots of things including some kinds of comping

What seems to be emerging is the 'take-as-lane' um, paradigm(?). With uncomplicated custom actions I can record take after take displayed in their own lanes, able to be soloed/muted/dragged elsewhere individually or grouped. That's all I want.

[Ann-82]

It would be really really helpful to have some group option for multitracks editing, kind of pro tools or logic behaviour.

[Funkybot]

For my, use cases, an ideal comping system would check the following boxes:

1. Create basic comps across takes or lanes on a single track (example: a vocal).

2. Create comps across takes or lanes of multitrack audio on one track (multichannel) - should work no different than a single audio track.

3. Create comps across takes or lanes of multitrack audio recorded to multiple tracks (traditional multi-track) if grouped.

4. Work the same for audio and for MIDI comps.

5. Any take/media item lane that vertically overlaps another would be in the pool for potential comps regardless of length or number of takes/lanes. Example: let's say the singer nails verse 1 so we only have 2 or 3 takes, but they were improvising a bit on the Outro vocal so there's 8 or 9 takes (or media items there). That should preset no issue for a comp system.

6. By default, the most recent take would be the active (audible) comp.

7. If audio, such a system would ideally handle crossfades between passages without the user having to stop to adjust fades.

8. Comping should require the least amount of motion (fewest clicks) as possible to select the section of the comp and audition between various takes. Example: click and drag to set comp ranges is faster/smoother than click to set begin point, move to end point, click again to add end point.

9. Active comps should remain in tact if a comp is copied then pasted.

10. Users should have the be the ability to save and recall multiple comp snapshots to audition various comps.

11. Users should be have the ability to view the waveform for the comp.

12. Users should be able to easily visually distinguish between active sections of the comp and inactive sections of the comp.

13. Once comping is complete and the user is satisfied, users should have the options to render the comp in place or to a new track, discarding unused takes/lanes.

[Kenny Gioia]

If we were to divide this into two possible directions, I would think we'd have:

A. A Pro Tools like approach where items are moved to a comp track and can be edited there.

B. A Logic like approach where you swipe sections and only hear the swiped sections with a possibility of also seeing one lane above that represents all the keeper sections.

So if you agree these are the two main possibilities, I would love to hear everyone's pros and cons of each.

Thanks.

[pentecost]

Originally Posted by Funkybot

For my, use cases, an ideal comping system would check the following boxes:
6. By default, the most recent take would be the active (audible) comp.

the thing that I don't particularly like about this with the current system is that any subsequent recordings override your previously selected best take(s). that can require a lot of note-taking, a lot of take cropping, or a lot of re-comping. if there was a way to set the currently selected takes as the A in an A/B/etc system prior to adding takes, and allow a keystroke or modifier to "set current take to preferred" during more comping, this would allow for existing preference to be maintained and modified pretty quickly.

[Vagelis]

Originally Posted by Kenny Gioia

If we were to divide this into two possible directions, I would think we'd have:

A. A Pro Tools like approach where items are moved to a comp track and can be edited there.

B. A Logic like approach where you swipe sections and only hear the swiped sections with a possibility of also seeing one lane above that represents all the keeper sections.

So if you agree these are the two main possibilities, I would love to hear everyone's pros and cons of each.

Thanks.

Logic does both and it does it great. It's the best system imo.

https://www.youtube.com/watch?v=-vepaE-H4EU

https://www.youtube.com/watch?v=X7fPNlm0ots

Having all these features in Reaper with the touch of our devs is going to be magic.

[Triode]

Originally Posted by Kenny Gioia

If we were to divide this into two possible directions, I would think we'd have:

A. A Pro Tools like approach where items are moved to a comp track and can be edited there.

B. A Logic like approach where you swipe sections and only hear the swiped sections with a possibility of also seeing one lane above that represents all the keeper sections.

So if you agree these are the two main possibilities, I would love to hear everyone's pros and cons of each.

Thanks.

AFAIK Logic's quick swipe comping doesn't allow you to edit crossfades until you "flatten" the comp (discard the unused takes). Also I always found it awkward that you have to open each comp container and the containers can't be joined to adjacent containers.

Pro tools is most flexible IMHO especially when combined with track-based group editing. It can get tricky if you try to add new tracks to a currently group (the number of playlists on each track doesn't sync). So there are weak spots that can be improved in either system.

[Teddy]

Originally Posted by Kenny Gioia

If we were to divide this into two possible directions, I would think we'd have:

A. A Pro Tools like approach where items are moved to a comp track and can be edited there.

B. A Logic like approach where you swipe sections and only hear the swiped sections with a possibility of also seeing one lane above that represents all the keeper sections.

So if you agree these are the two main possibilities, I would love to hear everyone's pros and cons of each.

Thanks.

Comping can sometimes be a bit of a jigsaw puzzle, and when I have to use a swiping system, I alway feel limited because changing the timing of clips and altering the fades is harder. I like pro tools way of doing it, because it allows me to easily make multiple comps, try out ideas, change the timing of clips and compare everything. This way, it can also be a way of exploring different musical ideas.

It's also relieving to be able to have the same editing tools available for playlists/lanes as on tracks. No surprises, only the same tried tools you know and love.

In the prerelease-thread, user AZPercussion made an awesome todo list for how a similar system could be implemented within the reaper universe. Hope it's ok if I qoute it here.

Here is my vision about minimum things should be done for track based comping. And they are not technically complicated I hope.
It's not so boring as it look maybe, but it's natural continuation of Media lanes and RE feature.

1. Allow razor edit on lanes, not on whole tracks.

2. To make cursor paste positon sensitive to lanes.

3. To make an actions for toggle solo and hide lanes. (As example click on lane number to hide others, and click with modifier to solo lane. Also it could be done faster via actions in action list looking on mouse position.)

4. To add an action for copying current razor area content to the first line.

5. To make an action for copying current razor area to a new lane or on a first possible line if there is a space. (Auto switch on fixed lanes mode if not yet.)

6. Add track grouping parameter: razor edit.

7. Allow lanes reordering (respect grouping and without)

[Funkybot]

Originally Posted by Kenny Gioia

If we were to divide this into two possible directions, I would think we'd have:

A. A Pro Tools like approach where items are moved to a comp track and can be edited there.

B. A Logic like approach where you swipe sections and only hear the swiped sections with a possibility of also seeing one lane above that represents all the keeper sections.

So if you agree these are the two main possibilities, I would love to hear everyone's pros and cons of each.

Thanks.

I think the Logic style approach is going to be the most efficient, and also used by other comping systems (Studio One, I think Bitwig) so it has the added benefit of being familiar (read: expected).

The thing people mention most often about Pro-Tools comping in these discussion isn't the process (from what I see anyway) but rather the playlist function. No reason one couldn't combine "Logic Style" comping with something like
"Track Comp Snapshots" (PT inspired) and get to a happy place in Reaper.

[Pink Wool]

I think a comp track just makes things (visually) easier! With take colours it's quite easy to see what you've chosen. I mean, we're almost there! The script by embass just needs to work on lanes and respect groups. Possible?

[ferropop]

Originally Posted by Pink Wool

I think a comp track just makes things (visually) easier! With take colours it's quite easy to see what you've chosen. I mean, we're almost there! The script by embass just needs to work on lanes and respect groups. Possible?

It would require a container, no? If you were to move the "top track" slices, everything underneath would move with them, again dragging us into all the same problems we currently have.

[MisterLevy]

Originally Posted by Kenny Gioia

If we were to divide this into two possible directions, I would think we'd have:

A. A Pro Tools like approach where items are moved to a comp track and can be edited there.

B. A Logic like approach where you swipe sections and only hear the swiped sections with a possibility of also seeing one lane above that represents all the keeper sections.

So if you agree these are the two main possibilities, I would love to hear everyone's pros and cons of each.

Thanks.

I’ve not used Pro Tools but if comping is like the Embass script then I think that Logic comping is better.

In Logic after you have swiped a section you can make adjustments to where the change in takes are without having to swipe the whole section again. Not sure if that is how Pro Tools works.

-

[paaltio]

Originally Posted by Kenny Gioia

If we were to divide this into two possible directions, I would think we'd have:

A. A Pro Tools like approach where items are moved to a comp track and can be edited there.

B. A Logic like approach where you swipe sections and only hear the swiped sections with a possibility of also seeing one lane above that represents all the keeper sections.

So if you agree these are the two main possibilities, I would love to hear everyone's pros and cons of each.

Thanks.

Definitely! I feel like the pros and cons are really dependent on the music itself.

Basically I see the suitability of Logic's take folders vs. Pro Tools playlists as a spectrum where, as the stochasticism of the music increases, you move towards PT.

So basically click-based solo performances: take folders are nice. You'll be able to move a little faster, and default crossfades usually work fine, you'll have less second guessing with tuning and so on.

Non-click music, or bigger ensembles with lots of mics: playlists. You trade a little speed, but you can adjust each crossfade separately, apply RX processing, tuning etc. and the original takes are there with the click of a button for reference, or if you just want to go back.

It's hard to really demonstrate why the flexibility is needed, but maybe I can provide an example.

I'm thinking of my recent project where I scored a movie for a 90-piece orchestra where we had all the sections (woodwinds, brass, percussion and strings) separate, each with dozens of mics, then multiply that by overdubs... if we had to edit that whole mess of a session with take folders, I think all of us would've just given up!

Each crossfade has to be custom, so the default crossfades from a take folder are a non-starter for example. Sometimes you'll crossfade across multiple measures in the strings, and then you have something like woodwinds which are an absolute pain to crossfade at all, and you kind of just experiment with different crossfades until something doesn't sound like a phasing mess.

Also, I wanted the timing and tuning of the music to be really tight so it was a ton of manual editing of even single notes at times, and of course it's really hard tuning an ensemble since they won't be off by the same amount, so you kind of just wing it at times. And then something you realize, "you know what, this comp is not working right now and the tuning is just making it worse. Let's go back to a vanilla take and do it again."

At that point it's just a life-saver to have the untouched takes right there on the other playlists in those situations. I would never edit an orchestral multitrack in a take folder system, and it seems like my whole team agrees with me on this!

[Nardberr]

Logic and Protools have very interesting comping and playlist setups.
I especially like the Logic playlist concept but both seem great.

Would have been interesting if Reaper took a unique comping approach similar to a graphics paint program.

Select solo, mute, pan or a plugin eg delay or reverb and swipe paint by lanes, takes or tracks.
You could solo more than one lane simultaneously if you wanted.
Easily paint in fx or fxchain wherever needed.

Don’t know why when playlist was mentioned sub-project came to my mind. Maybe a more
comprehensive combined playlist/sub-project idea.

This is me dreaming though as the requirements might be too demanding.

[paaltio]

Just a couple more thoughts on the take folders vs. playlists issue: I think it's perfectly possible to get both at the same time.

If your basic system is a take folder system, you also get the playlist flexibility if:

1) the comp gets created in its own lane
2) there's an action that turns the comp lane into an independent lane and removes the take markers
3) there's an action to promote sections of takes (e.g. with the razor edit) to the comp lane manually after it's been made independent

I think that's pretty much it, notwithstanding the big question of multitrack editing.

Now this isn't to say that full-blown playlists in REAPER wouldn't be a great addition for a lot of other reasons too. But just going with comping alone, I think this kind of compromise would satisfy both camps.

[Pink Wool]

Originally Posted by ferropop

It would require a container, no? If you were to move the "top track" slices, everything underneath would move with them, again dragging us into all the same problems we currently have.

Well, if it's "only" a copy of the items in lanes, then it should work independently from them too in my opinion.

Edit: Actually, do we have all that we need right now to make the RE script by Embass work on lanes? Or is there API stuff that is missing?