What's new

Key Switches vs. Separate Tracks

agarner32

Senior Member
I have gone back and forth between using key switches vs. separate tracks using expression maps in Cubase. I'm interested in other's workflow on this.

At the moment, I mostly favor using key switches with expression maps in Cubase for a number of reasons. I like having fewer tracks rather than having to edit a line using multiple tracks with various articulations. Plus it's just fewer tracks in my already large orchestral template. It just seems more natural to see one track per instrument. I do have a few extra tracks for libraries that have say special Fx on a different patch or performance legatos on a different patch. Although I usually notate first and then play lines in, with key switches, it's easier to export for notation or just see an entire line in Cubase if I need to.

One drawback is that sometimes short notes will be softer and it's difficult to adjust the volume if I'm using key switches on one track. The separate track way yields more control and the ability to fine tune.

I'm interested in other workflows on this. My conclusion is that different libraries require different workflows and that both methods can work equally well.

Thanks for any input.
Aaron
 
That's been my approach as well. I started writing way before sample libraries so my training is writing things in a score first. That's why it's not natural for me to separate things into various tracks, but I do see how in some instances it works well. Sometimes you have to if for example muted trumpets are in a separate patch or some special effect.
 
There is definitely no One Workflow to Rule Them All.

I like arranging my template like a score pad as well, and so also have one track per instrument. I use Program Change events for switching articulations, and have each track set up in the template so that it's as consistent as it can be (e.g. PC 1 is long, 20 is legato, 40 staccato, etc.). When I want to layer articulations, I use different MIDI channels on the same track.
 
One drawback is that sometimes short notes will be softer and it's difficult to adjust the volume if I'm using key switches on one track. The separate track way yields more control and the ability to fine tune.

Thanks for any input.
Aaron

That's why I took the time to balance all articulations correctly ahead of time and save them as presets in Kontakt, Play, VePro etc within my orchestral template
 
Zenfaced, can you actually balance articulations within one patch that has key switches? For example, if you have say a Spitfire Audio patch from their symphonic strings that has multiple articulations, is it possible to adjust the level on the shorts? I didn't think you could do that within one patch with key-switched articulations.

Thanks,
Aaron
 
I didn't think you could do that within one patch with key-switched articulations.
It depends on the library. In the case of Spitfire's combination palettes, you actually can set the mix per articulation. Although its implementation isn't all that it might be: once you switch to this mode, you have to balance each and every articulation separately. You can't cherry pick individual articulations for special treatment and leave the others at some global default.

If the library offers it -- Spitfire does -- you can always load the individual articulations as separate patches. These can all go on the same MIDI channel if you use UACC to keyswitch. Or you can put them on different MIDI channels and use something like FlexRouter (a KSP multiscript) to handle keyswitch routing for you (so that from the DAW perspective you're still only using one MIDI channel).
 
...One drawback is that sometimes short notes will be softer and it's difficult to adjust the volume if I'm using key switches on one track. The separate track way yields more control and the ability to fine tune.
I use one track per instrument with extensive expression maps in C9. Why is controlling volume on short notes difficult?
 
I actually do have multiple keyswitch tracks (and some single ones) per track. :) When you want to load a library like orchestral tools BST or BBr you really haven't got any other choice. For some people a template might mean to load only the essential articulation. To me it means loading everything (except for stuff I don't like or need). And that means a whole lot of articulations. Essentials, like short notes multi and legato, I load two versions, so I have the ability to set the volume or mic settings differently for using different articulations. Essential for having control over this is the ability to control everything via midi and recording all CCs in front of your music, so Cubase sends all parameters when you reload the project. Otherwise, you will get crazy about your music sounding different every time you load the project.
 
I use one track per instrument with extensive expression maps in C9. Why is controlling volume on short notes difficult?
When you use multiple articulations within one line, you might want to have a different volume on e.g. spiccato then portato short. Also, how do you layer articulations having only one track per articulation? One track per instrument really isn't giving me full control. I need at least two tracks, but I don't want expression maps with 40 articulations loaded, so I split them. Also, some single articulations, mostly legato, offer playability features (OT or spitfire) that are not available in the keyswitch instruments.
 
also looking for one work-flow that fits all<G>...

About as close as I have come is one track per instrument or section, with additional tracks to send key switches or other controller messages.

This solves the clutter, and translation to score problems associated with key switches, and the too many tracks when moving to standard notation of multiple tracks. What it doesn't solve is chasing articulation changes - although that is getting better in Sonar (but not Studio One yet).

Now that I've settled this I am going to look one more time at converting key switches to CC events, and then (if necessary) back again. It strikes me that once I get that working it ought to be once and done.

I should add that I am still (slowly) moving my template into VEPro, which adds to the confusion a little bit.
 
[QUOTE="FriFlo, post: 4069663, member: 51611.

1. When you use multiple articulations within one line, you might want to have a different volume on e.g. spiccato then portato short.

2. Also, how do you layer articulations having only one track per articulation? One track per instrument really isn't giving me full control. I need at least two tracks, but I don't want expression maps with 40 articulations loaded, so I split them.

3. Also, some single articulations, mostly legato, offer playability features (OT or spitfire) that are not available in the keyswitch instruments.[/QUOTE]

1. That is what MIDI ccs are for, no? Or later I can bounce the track to audio and cut the regions to add Gain in Logic.

2. I generally don't, because real players don't. But if i did i would just assign it to the same MIDI channel in a multi in Play or Kontakt.

3. In Logic, using the SkiSwitcher, I can create my own de facto "keyswitch" patches for libraries like that, as you can with Expression maps in Cubase, right?

Anyway, I have no dog in the hunt but separate tracks for each articulation leads to these massive 300-600 track templates that some guys use and I could simply not keep track of it.
 
When you use multiple articulations within one line, you might want to have a different volume on e.g. spiccato then portato short.
This is still possible with a single track. If the articulations are loaded as separate patches, the volume can be adjusted individually.

Also, how do you layer articulations having only one track per articulation?
I use separate MIDI channels on the same track when I want to layer. A side benefit with this approach is that for divisi writing (when I want each part to go to different patches), it looks as you'd expect in the notation view. :)

None of these problems are insurmountable with a single-track-per-instrument strategy. But I recognize that the track-per-articulation approach is easier to understand.
 
I have no idea why I would want articulations on separate tracks. Even if there really, really, really was some odd reason to separate one articulation (separate output because of some additional processing or something), I'll just load that one patch in a second instance - or just route it differently, depending on the sample player - and leave everything else as it is.
 
When you use multiple articulations within one line, you might want to have a different volume on e.g. spiccato then portato short. Also, how do you layer articulations having only one track per articulation? One track per instrument really isn't giving me full control. I need at least two tracks, but I don't want expression maps with 40 articulations loaded, so I split them. Also, some single articulations, mostly legato, offer playability features (OT or spitfire) that are not available in the keyswitch instruments.

Really interesting subject. I control short by velocity, Spitfire is switchable on this and most others do so automatically. When there are dedicated legato patches, I use shared UACC (SF) or K5 instrument banks (OT) to manage these. I've not yet needed to layer arts.

With Sable having over 25 articulations, that made too many slots in the expression map lane, so I've settled on this method, and redone the maps for all my libraries in the same fashion.

My Sable V1 expression map lane looks like this:

LNG (8)
SH (5)
LEG PER
LEG DECO (5)
PIZZ (3)
TR (4)
TREM (6)
FLAG
FX (5)
---------------------

My Tundra H Wild expression map looks like this:

LONG (3)
FLAG (2)
PULSING
T TTEMS
(PIZZ) 3
ARP
COL LEGNO
---------------------

CSS

SUS
SH (4)
LEG
MAR
PIZZ (3)
TREM
MEAS TREM
TR (2)
FLAG
---------------------

VSL

LNG (5)
SH (6)
LEG (2)
DYN
TR (5)
REP
-------------------

The slots can be controlled for playing by soft keys on a my Novation Zero MKII, and programed after playing the midi by clicking in the slot. The number in parenthesis indicates how many of those there are, and those are controlled by a fader controlling a CC (CC9 in my case), always going from shorter to longer, milder to harsh, etc. If there are 6 of those (e.g. shorts) I divide the CC9 into 6 ranges (128 divided by 6 = 21) of 21 each (Stacc SH 0-20, Stacc med 21-42, etc.).

This combo of a limited number of slots in the expression map lane, and a CC9 lane, is common to all my samples. You don't have to remember what each library does or doesn't have, since the names are there in the bottom (Expression map) lane of the Piano roll editor and side panel in the Arrange window.

Making the maps can take some time, but I've made templates of those with all the heavy lifting already done.

I'd be interested to hear what others have done...
 
Zenfaced, can you actually balance articulations within one patch that has key switches? For example, if you have say a Spitfire Audio patch from their symphonic strings that has multiple articulations, is it possible to adjust the level on the shorts? I didn't think you could do that within one patch with key-switched articulations.

Thanks,
Aaron

I used SkiSwitcher in Logic that allows you to create your own custom "keyswitches"which are assigned to different midi channels. But even before that, I will go into east west Play, for example, and balance all the articulations I loaded for, say 1st Violins which I created for that custom "keyswitch" patch. Then when thats done I go even further and balance each instrument with the other instriments so the volumes (loudness) of all orchestral instruments and sections ---the volumes at ff are balanced within the ballpark
 
I use one track per instrument with extensive expression maps in C9. Why is controlling volume on short notes difficult?
Some combination patches don't have balanced articulations and I've found this especially with short notes. If I play a line in with longs and shorts sometimes the short notes will be out of balance. Maybe I'm missing something, but the only way I know how to balance them in a key switched track is futz around with velocities, but then you don't get consistent dynamic layers. When this happens I do the separate track route because I can raise or lower the overall track volume.
 
I use shared UACC (SF) or K5 instrument banks (OT) to manage these. I've not yet needed to layer arts.
Thanks for this detailed post. I'll need to read it a few more times. I am not familiar with UACC. I have some homework to do.
 
UACC is a system developed by Spitfire, where instead of sending keyswitch notes to a patch to get it to switch, you send it CC#32 messages of various values. At the end of the day, it's just an alternative type of MIDI "articulation-switching instruction" to using keyswitch notes.

The advantage of using CC's for articulation-switching is that Logic chases them pretty nicely, as opposed to keyswitch notes, which don't chase. So, when using Spitfire patches without a system like SkiSwitcher, and you start playback from any random point in a project, patches will receive the necessary articulation-switching instructions. That's not the case with keyswitching patches, where it's very common to experience passages playing with the wrong sound because the playhead doesn't pass over the k/switch note needed to select the right articulation.
 
Last edited:
Top Bottom