Tag Archives: Add new tag

Cutting video in Ableton Live

I had no idea it is this easy to cut and edit video in the music production application Ableton Live! The lucky coincidence that had me find out was when I was hired to teach a group of film people about Live. live7videoTo prepare the workshop I started playing around with my own crappy cell phone video clips; simply throwing them into Live to see what would happen. What happened was just amazing! I discovered I was able to abuse and mutilate video almost as badly as I usually mangle audio in this application! You may think “so what” but listen here: what whe have here is an exciting new hands-on approach for churning out video and music simultaneously! Not “making video for music” or “composing music for film”, that’s just boring, lame and obsolete by now. The new vision is about being a video musician! Play the shit out  loudly with the right attitude and apply your musician’s first-take approach to video! Ok, so for those of you who haven’t gotten around to try this out for yourself, here’s what you can do.

Drop a Quicktime video clip into Live (Arrange View) to have it create an audio track and open a video display window where you can see the moving picture as the sequencer plays. If double-clicking the video clip you get access to clip properties at the bottom of the screen (just as for audio clips).

If you already have a music mock-up cocking and want to adjust the musical tempo to hook up musical downbeats with cue points in the movie, set the QT clip to “Warp as Master”. Then create warp markers just as you do with audio clips. Drag a warp marker to align a video cue point with a musical timing point.

Also, if you throw in other QT clips they will create their own, new, audio tracks. Mix the volume of these tracks according to how much of the sync sound (original video cam sound) you want to keep (or not). You may also process the video sound with Live’s effects, if you’re so inclined. Anyway, a very cool thing is this: if these extra thrown-in clips are happening while already the master video clip is playing they will just take over the video channel for the clip’s duration. See the point? Instant video cuts with an exact musical timing! And you may do this as you are also creating the music – in the same arrangement window with the same visual timing grid. Awesome! You may even use a copy of one video clip (alt drag and drop it to copy it) to shrink it into a short slice to just beep a 32th note video scene into the overall video flow. And of course this does not even have to show the same moving picture flow as the main video at that particular moment in time, you are free to fetch the cut’s content from earlier or later in the video take.

Here’s another scenario: You have a video with a rhythmic passage that you want to use as the tempo base to create new music. Then do not set the main video clip to “warp” but fool around with Live’s global tempo until you find a tempo that is the same as in that particular video sequence (you may create a playback cycle in Live, around the rhythmic video part, while working out the fitting tempo). You may also want to adjust the video’s starting point in order to set the groove right. Below is a quick mock-up I made in a couple of minutes with this method. I also copied the video and shrinked it to loop only the rhythmic part part a couple of times to make room for adding some drums to the video loop + video sound.

This Man Is So Rude! from Per Boysen on Vimeo.

You can not play video backwards in Live though. But you can do all this:

  • Continuous speed change of video + original sound.
  • Continuous speed change of original sound while video stays normal.
  • Distribute short video slices as “cuts” to take over the video channel from the main video track.
  • Tune the original audio of a video clip into any melodic interval.
  • Change pitch of original video sound without affecting the sync sound timing.
  • Draw rhythmic “pitch change melodies” for the video clip’s original sound.
  • Render both a video + now sound and a 24 bit sound file to give the proper audio mastering treatment. Then you may put that sound channel back into the film.

To wrap this up I have to mention that this was all written regarding Ableton Live 7.0.14. By late spring 2009 we will get access to Live 8.0 and the add-on tool Max For Live. Max For Live is developed especially for Live 8 by Cycling74, based on both their legendary midi and audio manipulation software Max/MSP and the video manipulation application Jitter. Needless to say, Live 8 will bring video musicians some sharp new axes.

Comment here <<<

Steppophonic Looperformer – please steal this!

I’m sharing this awesome idea for an awesome electronic instrument! I’ve been longing for a Steppophonic Looperformer for almost a decade and now I’m giving away the idea for free in the hope that someone will develop it as a software plug-in, a Max For Live mockup for Ableton Live 8 or maybe integrated in Numerology 2.0 Pro. Or whatever… it’s free – grab it!. Creative Commons license applies as stated below.

What is it?
The Steppophonic Looperformer is a step pattern sequencer driven real-time sampler (to be implemented by software). If the sample button is pressed down it samples audio from the system audio input and instantly plays it back according to the looping pattern (patterns can form chords, bass lines… whatever, and you may modify/swap patterns as you go). The idea is that a vocalist, trumpet/clarinet/sax etc player shall play the Steppo to orchestrate multi-timbrally on-the-fly while also playing the lead.

What makes this new and unique is not “looping” or “sampling” but that it’s a real-time system, optimally playable in a musical sense. You can run through chord progression that are composed or improvised on the spot as in Keith Jarret’s legendary piano example, and do it in a techno style sequencing context that still fetches its sounding source component from the acoustic instrument you are playing. So you are totally in control, expressing yourself.

The length of the sample depends on how long the sample button is pressed down. Duration of the note playback is controlled by its own parameter though. When sampled the audio snippet is kept in RAM, looped and split into as many monophonic voices/instances as there are DOTs set up in the grid (or an absolute number, spanning four octaves, if that is easier to program, not having to deal with voice allocation). The vertical grid axis represents pitch (12 half note pitches) and the placement of each dot on this vertical axis controls the playback Rate/Speed of the sample (“pitch”). The horizontal axis represents the loop of the pattern. In the example pattern above we are running a 12 beats long pattern, but the number of steps should be controllable by MIDI (so you can “sweep” the looped pattern’s length continuously while playing, as you can also “sweep” the Steppo’s relation to the global tempo). Normally the grid beats correspond to musical beats, i.e. an eighth note – but that can be changed by a “tempo divisor/multiplier” parameter. A dot may be set to play back at another octave than the grid displayed octave and is then displayed with a special color (in order to keep the graphics minimal). Each note can also be given a Release value to make it fade out slowly (release value “playable by MIDI”). Every voice/pitch is monotimbral – meaning that if using really long samples each new note will overtake a note that is already sounding at the same pitch.

How to use it?
Run it on a laptop while singing, or playing, into the audio input. Kick a foot button to snag a note now and then to feed the pattern. Snagging a different note will result in a parallel pitch transformation of the pattern. Another way to move the music is to change pattern while keeping the same sample. It is a creative way for singers and monophonic instrumentalists to improvise chord patterns and melodies simultaneously.

An interesting aspect is that note pitches are generated as in old-school samplers, by Rate/Speed shifting. This means that if you record a longer snippet where you play a rhythm this rhythm will sound faster at higher pitched notes. But you can also record a short snippet but keep the duration set to longer notes and this will result in the sound of a sequenced old-school sampler (looped short samples playing long notes).

One Bank holds twelve Patterns. These twelve patterns correspond to the twelve notes of the octave. Users should be able to set up the patterns to support any major, minor or personally weird key/scale. That way “chords”, “keys” or “song parts” can be assigned to separate foot pedals. What I think is cool with this is that it opens up for very free multi harmony imrovisation. However, the actual sounding tonal center depends on what sounding pitch is fed into the Steppophonic Looperformer.

Bottom line: While playing a lead instrument or/and singing you can use a simple MIDI foot pedal board to direct the Steppophoner into any sort of chord progression – even on-the-spot improvised. And it really is an instrument because you can change the sound of the whole shebang in a blink by simply exchanging the sample for a note with a different tonal character. And of course you snag the sample seamlessly from your lead singing/playing, making it an instant performance process.

Here are a couple of loose ideas that would be cool to have:

  • Duration value (long/short note with abrupt ending)
  • Release value (fade out ending)
  • Release Pitch Fall (with parameters “amount” and  “speed” of fall, assignable to DOTs or Grid Positions)
  • Release Pitch Rise (with prameters “amount” and “speed” of rise, assignable to DOTs or Grid Positions)
  • Scrolling “Tempo Divide/Multiply” in musical values (i.e. half note, dotted, triad)
  • “Tempo Divide/Multiply” value optional to follow Hard Sync (see below)
  • Hard Sync parameter (how many steps until the downbeat will be forced to happen on the global tempo bar cusp. Great for landing on your feet when returning from “granular rhythm chaos bursts”)
  • The DOTs can be put in with the mose but should also be programmable by live MIDI (as you work with an MPC). If you play a MIDI Note into the Steppo while in “Adapt Mode” a DOT will be placed at the grid cusp closest to where you played that MIDI Note. If you keep the Steppo in “Adapt Mode” and play the same note at the same point in the pattern the DOT will be deleted from the grid pattern. This way a performer can both catch new audio of different pitch into the same pattern, swap to another pattern or PLAY a new pattern as you record drums with MIDI pads.

I’m not a programmer, just a musician that would love to have this instrument (I’d rather spend a day playing music than programming code). So I’m giving away this idea for free to whomever wants to give it a shot. To tell the truth I have been longing for the Steppophonic  Looperformer for almost a decade and even suggested some software companies to pick it up, without any luck so far. But I hope the odds are optimal now, because great programming tools are in the hands of talented individuals and product developers while the commercial players start to support grass root community collaborations (much thanks to the compute gaming industry I would guess).

Existing products, that I know of, that sort of lives in the same building as this idea are GURU by FXpansion, Numerology and the Granulaterre plug-in of Logelloop. But they all miss out on some points (…so far) that I think is important.

Here’s a listening example. The Steppo is what here makes it possible for me to play both the chords and bubbling bass line within the same flute performance.

Ideas? Suggestions? Comment here <<<