Which Do you prefer, keyswitches or separate midi instruments?

newbycomposer said:

I recently started writing again after a very long time. I'm using eastwest opus, and have a string section which Has a part playing this rhythm (photo below, speed up the notation it was easier to turn it into a full measure than half measure quickly).

I'm wondering if for parts like this where you have it switching back and forth, if its better to just have each articulation on its own track and not mess with the keyswitches, flipside is then you have to break the part up into a bunch of parts, which has its own problems.


First link is notation, second is first audio clip

Click to expand...

People work effectively both ways. Personally I prefer keyswitches whenever possible. But I use articulation sets to manage the changes.

Key switches, and more precisely, assignable keyswitches.

Keyswitches + Art Conductor FTW. Makes it a lot easier to mentally manage everything. If I need to control separate articulations, or I know I really only need one articulation from a specific library, then I'll add a separate track manually on case-by-case basis.

While there is room for improvement, Spitfire + key switching works fine when you choose UACC KS. That said, I use Cubase, so depending on the DAW you're using your mileage may vary.

If a part were to be started with a rough real time performance, a good way to do this would be to use key switching via clearly labelled buttons on a Stream Deck or equivalent. This is especially useful for libraries that require multiple notes for articulations for which it'd be difficult to memorise the key combinations. However, this complicates things down the line when copying a part with different arts. requiring different negative delays leaving all the notes off grid, so the part would potentially need to be merged with the part prior to its destination to not lose its first note, and this would make further edits more complex. Also the key switches note range would need to be excluded when transposing parts or tracks. If DAWs had a sticky filter within the transpose function to exclude a range at part or track level then this feature would resolve that.

Separate tracks per art. when given their own negative track delay are way more efficient later on when copying or moving parts around.

In Cubase 13 using Babylon Waves, the expression maps can handle limited key switching functionality, but in my experience this has only resulted in incomplete art. entries in the editor unless the notes had already been written into the part. However, expression maps do overcome the issue of transposing using simple key switching, albeit in a clunky way.

As it stands there's no single best answer.

For those of you using separate tracks for separate articulations, are any of you playing the parts in live and then using some smart logic processing to separate the one part into separate tracks?
If so, any hints on how to do this in Cubase?

Key switches with Articulation IDs. I want my Logic Pro projects generally to look like a score page.

BTW, When I want to, I can trigger them from the keyboard in the free Logic Remote on my iPad, leaving my keyboard controller free.

Key switches + my custom keyswitching solution made with Open Stage Control that makes keyswitches so easy to work with

I generally have key switches with articulation maps, though in situations where I need just two or so articulations, I will have one art per track. Digital Performer articulation maps can handle this quite nicely: I can create an articulation map that switches between channels. That way I can have one track in DP that will address two channels/instruments. That makes editing much easier. I imagine other DAWs do this as well.

All in one performance patches / Infinite Winds and Brass are my first choice currently.

But after that, I prefer Keyswitches / Babylon Waves + Sound variations in S1.

I do believe that the future lies with instruments that really respond to your playing, so that you need few if any key switches, like the AcousticSamples V-Horns and saxes.

At least for those of us who are proficient with either keyboards, breath controllers, or MIDI guitars.

Two questions that I really never resolved on this topic, based on my limited knowledge of all of this:

1) if you do one articulation per track, and you export the project for Dorico to create the score (something I haven't had a need for YET), how do you get all first violins to show properly on their staff, V2 on theirs, celli, etc etc?

2) I'm really used to playing in lines with my left hand on my CC faders adjusting CC1/1 on the fly. How do you keyswitch at that point? Do you run KS instead, and then go back in and record CC11/1 after the fact?

For right now I'm playing in strings using one articulation, then fixing all that afterwards, whch is why I like BW in Cubase's expression maps.

gohrev said:

While there is room for improvement, Spitfire + key switching works fine when you choose UACC KS. That said, I use Cubase, so depending on the DAW you're using your mileage may vary.

Click to expand...

UACC is fine. I prefer UACC KS. Both work in every DAW. UACC KS works better when you assign different KS within a chord (mileage might vary depending on the DAW because some are not equipped to deal with articulations on a per-not level).

Logic articulation ID.

I use an iPad, although I simply have a two octave keyboard on it that's set below the range of an '88. This means that the key-switches are out of the way of the normal playing range. If I want to trigger the keyswitches on the physical keyboard, I just use Logic's offset controls to bump them up a couple of octaves.

I'm not averse to a per-track basis either. It depends on the track and what I'm trying to achieve.

Single articulations per track. Keep in mind I'm writing mainly Hybrid Orchestral tracks that have little to do with any kind of realism. Things need to be tight. Once Cubase adopts Dorico's delay per articulation into their expression maps, then I will go back to working with them.

Single articulation per instrument track all the way!

Easier to manage negative track delays (because it's usually different for every articulation). And also you can send different amounts of reverb to each articulation, and it's easier to split stems that way too.

So I personally prefer single articulation per track!

For me (not a professional and I don't play in live, or if I do it's a rough idea):

Exposed part that needs to sound fluid / is coming from the same instrument or section = Keyswitches

Part that isn't exposed or parts that are disconnected = Articulations on different channels but on the same track (preffered)

Same instrument but serving a different purpose = Articulation on separate tracks even when they aren't playing together.

Keyswitches via Program Change.

I got up and running with Art Director but the shortcomings of using midi notes as switches soon became apparent. In the instruments with a larger amount of articulations Art Director will only have a subset of them assigned due to the fact that the switching range would start to impinge on the playable note range.

I like to have EM access to all articulations and keep my midi channel use low, so I moved to EM switching via program change, and doing my own expression maps for the bigger ones. Using instruments banks where necessary. This means I do use the individual articulation patches, but loaded into the instrument bank for EM switching..

If its a SF UACC instrument I dont need to use instrument banks of course.