Chain Selectors, Rack Activators, Real Time Midi+Processing
Posted: Thu Jan 07, 2016 6:19 pm
I am building my Live Set and doing all my production in Session View, and wanted to post the methodology I’ve come up with to tackle my goals, and see if anyone has any advice or solutions that I may be overlooking to make the process more efficient and easier to manage.
While some elements of my session will be pre-recorded and processed audio clips, most of it I want to keep as midi, triggering samples or soft-synths in real time (either played live by a performer via a midi controller or pre-programmed midi clips). This means that any processing and mixing on these sounds such as EQ, compression, FX, delays, reverbs and automation will also occur in real time.
There are a number of reasons why I’m trying to keep as much in midi/real time as possible, rather than stemming out to audio. For example, taking advantage of “Midi Effects” such as random, velocity, scale, etc. for improvised sections where we want the music to morph in unexpected ways…We also want to have a majority of the sounds/samples available to be played live via midi controllers OR to be triggered by pre-programmed midi clips, depending on the section of the song and how many members are available to perform with us for a given gig (and we want the sounds to be fluid/the same whether pre-programed or played live). Last but not least, our productions are constantly evolving works in progress (it’s part of the nature of the project), and we want the ability to mix and alter the sound design and sample design within the session, between shows, without having to continuously go back and forth bouncing/stemming out audio every time we want to make changes.
Each song requires different soft-synths/presets, drum samples, effects, processing, levels, plugin settings, mix bus settings, etc. In order to accomplish this and conserve CPU, I’ve been using MACROS to map both "Chain Selectors" and "Rack Activators”, so that when one song is playing, all the synths, samples and plugins are “selected” and “activated", and everything else in the entire session is “unselected” and “deactivated”. Each time a new song is launched (via a Scene) midi clip automation is used to automatically change the macro settings to the respective song and select/activate the samples/synths and processing.
In order to show how I’m doing this, I’ll outline what I do for Drum Samples, though the same concept is applied to Bass, other synths, effects, master chain effects, using any combination of instrument racks, drum racks and audio effect racks etc.
For Drums, I have 4 Midi Tracks labeled Kicks, Snares, Hats and Percussion. These are how I like to group all my drums before hitting a Drum Bus.
Let’s say in a single song, I am using 4 different Snare Sounds depending on the pattern/section.
On the Snare Channel, I drop in an Instrument Rack. This is where I will create chains, 1 for each song in the set. I drop a Drum Rack into the instrument Rack and label it “[Song 1] Snares". On that Drum Rack I will drop another instrument rack onto each of 4 pads, corresponding to the midi notes used to trigger each particular snare (i.e. A1, A#1, B1, C1). The reason I put instrument racks on each pad is so that I can layer different samples to create each “snare” sound. I balance the sample’s volume and panning within this instrument rack, and process them individually if needed with audio effects. I then process the entire layered snare sound if needed, and balance/pan it relative to the other 3 snares in the track. I then process the entire Snare Group usually with some glue compression and filter/eq.
For the next song, I will drop another Drum Rack into the Main Snare Group Instrument Rack, which adds it to the chain list. I will ultimately repeat the above process for each song in our repertoire…however, since each song will require different samples and processing, I have to have an approach that turns everything off that’t not needed for that particular song, to conserve CPU.
The way I do this, is I assign a Macro on the Main Snare Group Instrument Rack to control both the Chain Selector, as well as the Activators for the Drum Rack and the Audio Effect Rack that’s on the entire Drum Rack (all the audio effects within the drum rack automatically turn off when I deactivate the drum rack, however audio effects on the entire drum rack/group of 4 snares don’t, so I have to map the activator of the Group Audio Effect Rack as well to the same Macro Values).
The mapping works like this: The Macro gets mapped to the chain selector 0-127, so that each song scrolls to the next drum rack/audio effect rack…then the drum rack/audio effect rack activator buttons get mapped to that same macro, but the first song’s range is set from 0-0, second song from 1-1, third from 2-2; what this does is makes it so the racks/plugins are ONLY on when that song is selected, and everything else in the session is off.
This is the same general idea I use for Kick, Hats, Percussion, Bass Synths, Other Synths as well as Audio Effect Racks on Busses, and the Master Channel. Every piece of processing, if set up as a chain in an Instrument or Audio Effect Rack can be mapped this way, and automation at the start of every scene can take care of automatically activating the respective synths, samples and plugins/settings for each song/scene.
I hope my explanation of how I am approaching this makes sense, so that others can chime in and offer any advice or suggestions to build my session as efficiently/simply as possible, and maximize it’s stability. BTW I am using a handful of third party soft synths and plugins, and will keep track of any third party devices that seem to cause problems, too much latency, or crash the session, and will try to use Ableton’s stock effects as much as possible.
I basically got this idea from an online Lynda course with Daniel Mintseris of the band St. Vincent (who mentions that in their live set he has something like 12,000 elements mapped in this way - everything from drum samples, to synths, to midi changes out to midi controlled guitar fx processors, etc.). I’ve adapted his methods to my specific needs, though before I delve into building the entire sessions this way and mapping everything (hundreds, if not thousands of mappings eventually), I’d like to run this by others and see if there are other methods I might be overlooking
Thanks!
While some elements of my session will be pre-recorded and processed audio clips, most of it I want to keep as midi, triggering samples or soft-synths in real time (either played live by a performer via a midi controller or pre-programmed midi clips). This means that any processing and mixing on these sounds such as EQ, compression, FX, delays, reverbs and automation will also occur in real time.
There are a number of reasons why I’m trying to keep as much in midi/real time as possible, rather than stemming out to audio. For example, taking advantage of “Midi Effects” such as random, velocity, scale, etc. for improvised sections where we want the music to morph in unexpected ways…We also want to have a majority of the sounds/samples available to be played live via midi controllers OR to be triggered by pre-programmed midi clips, depending on the section of the song and how many members are available to perform with us for a given gig (and we want the sounds to be fluid/the same whether pre-programed or played live). Last but not least, our productions are constantly evolving works in progress (it’s part of the nature of the project), and we want the ability to mix and alter the sound design and sample design within the session, between shows, without having to continuously go back and forth bouncing/stemming out audio every time we want to make changes.
Each song requires different soft-synths/presets, drum samples, effects, processing, levels, plugin settings, mix bus settings, etc. In order to accomplish this and conserve CPU, I’ve been using MACROS to map both "Chain Selectors" and "Rack Activators”, so that when one song is playing, all the synths, samples and plugins are “selected” and “activated", and everything else in the entire session is “unselected” and “deactivated”. Each time a new song is launched (via a Scene) midi clip automation is used to automatically change the macro settings to the respective song and select/activate the samples/synths and processing.
In order to show how I’m doing this, I’ll outline what I do for Drum Samples, though the same concept is applied to Bass, other synths, effects, master chain effects, using any combination of instrument racks, drum racks and audio effect racks etc.
For Drums, I have 4 Midi Tracks labeled Kicks, Snares, Hats and Percussion. These are how I like to group all my drums before hitting a Drum Bus.
Let’s say in a single song, I am using 4 different Snare Sounds depending on the pattern/section.
On the Snare Channel, I drop in an Instrument Rack. This is where I will create chains, 1 for each song in the set. I drop a Drum Rack into the instrument Rack and label it “[Song 1] Snares". On that Drum Rack I will drop another instrument rack onto each of 4 pads, corresponding to the midi notes used to trigger each particular snare (i.e. A1, A#1, B1, C1). The reason I put instrument racks on each pad is so that I can layer different samples to create each “snare” sound. I balance the sample’s volume and panning within this instrument rack, and process them individually if needed with audio effects. I then process the entire layered snare sound if needed, and balance/pan it relative to the other 3 snares in the track. I then process the entire Snare Group usually with some glue compression and filter/eq.
For the next song, I will drop another Drum Rack into the Main Snare Group Instrument Rack, which adds it to the chain list. I will ultimately repeat the above process for each song in our repertoire…however, since each song will require different samples and processing, I have to have an approach that turns everything off that’t not needed for that particular song, to conserve CPU.
The way I do this, is I assign a Macro on the Main Snare Group Instrument Rack to control both the Chain Selector, as well as the Activators for the Drum Rack and the Audio Effect Rack that’s on the entire Drum Rack (all the audio effects within the drum rack automatically turn off when I deactivate the drum rack, however audio effects on the entire drum rack/group of 4 snares don’t, so I have to map the activator of the Group Audio Effect Rack as well to the same Macro Values).
The mapping works like this: The Macro gets mapped to the chain selector 0-127, so that each song scrolls to the next drum rack/audio effect rack…then the drum rack/audio effect rack activator buttons get mapped to that same macro, but the first song’s range is set from 0-0, second song from 1-1, third from 2-2; what this does is makes it so the racks/plugins are ONLY on when that song is selected, and everything else in the session is off.
This is the same general idea I use for Kick, Hats, Percussion, Bass Synths, Other Synths as well as Audio Effect Racks on Busses, and the Master Channel. Every piece of processing, if set up as a chain in an Instrument or Audio Effect Rack can be mapped this way, and automation at the start of every scene can take care of automatically activating the respective synths, samples and plugins/settings for each song/scene.
I hope my explanation of how I am approaching this makes sense, so that others can chime in and offer any advice or suggestions to build my session as efficiently/simply as possible, and maximize it’s stability. BTW I am using a handful of third party soft synths and plugins, and will keep track of any third party devices that seem to cause problems, too much latency, or crash the session, and will try to use Ableton’s stock effects as much as possible.
I basically got this idea from an online Lynda course with Daniel Mintseris of the band St. Vincent (who mentions that in their live set he has something like 12,000 elements mapped in this way - everything from drum samples, to synths, to midi changes out to midi controlled guitar fx processors, etc.). I’ve adapted his methods to my specific needs, though before I delve into building the entire sessions this way and mapping everything (hundreds, if not thousands of mappings eventually), I’d like to run this by others and see if there are other methods I might be overlooking
Thanks!