Comping

[Vibrations]

I don't know the possibilities of M4L.
So I have a question.
Is it possible to make a device for recording in comping mode?

[Vibrations]

A simple yes or no would be enough.

[broc]

Maybe you should explain what "recording in comping mode" means?

The common meaning of comping is this (from wikipedia):

Comping (an abbreviation of accompanying) is a term used in jazz music to describe the chords, rhythms, and countermelodies that keyboard players (piano or organ) or guitar players use to support a jazz musician's improvised solo or melody lines.

[Vibrations]

Comping is well known in this forum.
A search founds 526 matches.
Comping is a recording features every DAW has: Cubase, DP, Sonar, Logic, ProTool, Propellerhead Record. But not Ableton.

[prebentious]

The answer is yes, but it probably won't be easy

[stefan-tiedje]

Comping is well known in this forum.
A search founds 526 matches.

But they don't explain it...;-) And who reads 526 matches to find out...;-)
A simple explanation is faster in this case.
You want multiple takes of a recording and choose the best, all being in the same Track...
If I understand that correctly, in Live this is sort of built in already, but not called comping.
You can record a part in a loop, creating a recording containing all versions. Then you choose the one which is best...
Now if the take is longer, and you want to choose only parts of each, its getting more complicated, or even impractical.
Of course M4L could help. I'd use some sort of buffer creation while recording, each turnaround would create a new buffer. Then on replay, you could use two waveform~ objects, to compare two versions and maybe a third waveform~/buffer~ to define the final parts. In the end you need to copy the result into a single clip...
I am fantasizing of course, but this could be a starting point for your own fantasies...

As prebentious already mentioned, yes, but...
I think its worth it...

Stefan

[Khazul]

Depends how you want to go about it.

I wanted to be able to create a multi-take and comping helper in a group tracks that would creatte additional tracks as needed when recording (so that after is become easy to select what you want), then get the tool to create the cross faders over the wanted bits clips etc. another way is to record into sequencial clips (on clip per take), but then I would have still wanted it to do the job of rearrange these into separate tracks in a group (so I can have the processing ont he group track).

With all that - there are alot of required functions that are simply not exposed via M4L, so thought fukit - just use Cubase instead - which sadly seems to be the usual result of wanting to do something with M4L - fukit just do it reaktor / download a reaktor ensemble or whatever, or fukit - write a VST in c++, or just use M4L to spit out data via UDP and write a separate application... or cant do anything with sysex, midi clocks etc... allways down to basic functions that are not exposed and or midi/audio input and output combos that are not supported at all in M4L, yet are common in many plugins, or other stuff that is crippled to the point of being useless.

[broc]

But it should be mentioned that in some cases M4L can be used to overcome inherent limitations of Live.
For example, the midiin/out externals from Leigh Hunt allow unrestricted midi handling and routing.

[stefan-tiedje]

With all that - there are alot of required functions that are simply not exposed via M4L, so thought fukit - just use Cubase instead - which sadly seems to be the usual result of wanting to do something with M4L - fukit just do it reaktor / download a reaktor ensemble or whatever, or fukit - write a VST in c++, or just use M4L to spit out data via UDP and write a separate application... or cant do anything with sysex, midi clocks etc... allways down to basic functions that are not exposed and or midi/audio input and output combos that are not supported at all in M4L, yet are common in many plugins, or other stuff that is crippled to the point of being useless.

I agree, that there are restrictions within Live, which need some creativity to overcome. Personally I was working with multichannel outputs in Max/MSP for years, and Live only knows stereo in a track. But I could create a rack which does perfect spatialisation to eight outputs from within a track. (http://dl.dropbox.com/u/288305/St.pace.zip)
I would much prefer to have different track types in Live than having this hacked rack, but for now it works and the more people are going multichannel, THOUGH they use Live, the more it becomes obvious, that a change is worth to implement on a more basic level in Live directly. If we just start to invent, without bitterness, and throw in our creativity to overcome these limitations, it has a chance to flow into a future version of Live.
Each of us has a preferred tool, one which is used for most of the work, but has some limitations as well. Usually this won't prevent us from simply doing our work, even if its more hassle than with a different tool, we find our workarounds and specialized work flows. These influence in the end the tool we love as well, its a feedback loop between users and developers.
I am sure your wish could be accomplished somehow with M4L. Its not so much my own concern, as I prefer live music, and complete takes in recording for artistic reasons (and out of experience). That's why I haven't come up with a solution for this specific problem, and its hard to imagine how exactly it should look like, as I avoid these "lets fix it in the mix" tools usually, though I know, some times it is useful and exactly what you need...
For me the combination of Max and Live extends Max with powerful sequencing, which was a weak point in Max alone, though totally possible before...

For a pure recording session, I would usually use ProTools, if its at hand (I don't own it...) I guess if you have access to Cubase this is completely reasonable too. You can still dump the tracks later into Live and this can be a fast and effective work flow. It only depends on what you want and what you know, you choose the tools which do it best.
For example, I can't imagine that Reaktor could do anything I can't do in Max, but if you are more familiar with Reaktor than Max, YOU have a good and valuable reason to use it instead of Max or M4L...

just some thoughts to share...

Stefan

[Khazul]

For example, I can't imagine that Reaktor could do anything I can't do in Max, but if you are more familiar with Reaktor than Max, YOU have a good and valuable reason to use it instead of Max or M4L...

Were getting OT, but as you mention that - unlike M4L, with reaktor you can create multi-in, multi-out effects (whch means you can make sidechain fx). You can also create audio+midi-in effects easily which opens up all the midi note performance effects etc - it depends upon which reaktor shell plugin you use.

Also I find seems offer alot mroe options one way or another from creating character stuff rather than just basic algorythmic stuff - ok so alot of that results from stuff that the commnity has created for reaktor, but still seem to exist to a greater extent than such components for MAX/MSP.

The huge strength of max/msp was about being able to access a whole load of os services including network ports, midi ports, audio ports, files, and basically being able to pretty much do what you wanted with real MAX/MSP, but in live so much of the audio and midi (and of course the video side) side is crippled beyond usefulness and so litle of basic live functionality that you can do with a single action in the UI is not accessable and parameter control is useless with racks etc. Yes, I know - you can spend another 600 quid (or maybe it was USD - either way - huge money) to also buy the full MAX/MSP/Jiter in addition to help with some of this.

At the very least it should have included a variety of run time shells for audio in/out, midi in/out, audio+midi in, audio out, multi-channel audio+midi in and multi-channel audio out etc (assuming that live needs to know how many channels something has going in and out when it loads it).

[vicz]

I would normally call what you are looking for 'multiple takes' and 'composite takes'. Comping in music more often refers to keyboard players left hand chord progressions, especially hammond organ for some reason .

As a poor Max programmer I would say it would be a bit of a stretch. Max4L is good for signal processing, drivers and minor interface tweaks. A good challenge though, anyone up for it?

[luddy]

Comping is what every engineer calls it, for better or worse. Short for composite, which isn't a verb, but there it is.

Anyway, real comping is different from selecting one take out of many. It means editing together the best whole from bits and pieces of all the takes. That's why the quickswipe feature in Logic is so great: you can quickly highlight the pieces of each take that you want to keep in the composite, and you can create many different composites for side-by-side comparison. Very nice workflow. If all that were needed was to select one out of many takes then it would be enough to have take folders, no need for quickswipe. Same for PT, DP, etc.

Just say'n that to put together a full-blown comping facility in Live via M4L would be quite a challenge; I'm not sure the API gives you enough of a handle to really do it properly.

-Luddy

[Vibrations]

Comping:
A picture is more worth then thousand words.
Here is a wery simple demonstration of Propellerhead Record's comping editor.
http://www.youtube.com/watch?v=SNRZ0dCq284

[vicz]

Thats the Record Comp Editor, as in Composite (takes).

This is 'Comping' http://www.youtube.com/watch?v=eqMXl-sLWbg

[vicz]

A good challenge though, anyone up for it?

Considering it at the moment.

Right now my main gripes are: "What am I dealing with exactly" and in this case most importantly (ok, selfish comment:) "What's in it for me". And no; I don't mean compensation (as in money or so) here (beer is much better anyway ), couldn't care less; more wondering how this might enhance my current workflow.

Although Stefan summed it up pretty well (I googled and came across the same kind of examples eventually) I still fail to see the big advantage here..

If anyone cares to elaborate; I'm interested!

Problem is without a runtime the marketplace is so limited Big advantage is not having to rely on a singer or band member being able to do an end to end flawless take (though personally I prefer the sound with some flaws left in, but thats another story...