Context
I’ve been experiencing a long-standing issue in which Cubase doesn’t reliably record the first articulation of a passage on tracks with expression maps. This issue has been mentioned a few times in the past (see here, here, here, and here), but it doesn’t appear to have a well-written bug report or any formal acknowledgement from Steinberg. I’ve spent many hours attempting to reproduce it reliably, but it’s proved difficult to nail down.
The Problem
Consider a track with an expression map. If one records MIDI onto the track, including remote-triggered articulation changes, the first of those articulation changes does not always get written to the articulations lane when recording stops. This is problematic because it’s preferable to avoid having any MIDI regions with notes that lack articulation data. Since I always want every MIDI region to contain articulation data for every note, I’m constantly having to open regions after recording in order to manually draw in the first articulation.
Reproduction Steps
Create a track with an expression map that has multiple sound slots/articulations of type direction (not attribute).
Record MIDI onto the track, ensuring that the first MIDI message is a remote-triggered articulation change. For example:
press record
trigger whichever articulation is top-most in the articulations list (not necessarily the same as the top slot in the Sound Slots list)
press and release a note
trigger a different articulation
press and release a note
stop recording
Most of the time, the result will be a MIDI region in which the first articulation was not recorded (see picture below). The expected result would include a recorded articulation (in this case, Mallet) before the first note.
Narrowing it down
As mentioned, this issue doesn’t always occur; but when it does, it seems to be correlated with whatever articulation was last active. Since the top-most articulation serves as the “default” and is switched to automatically in many situations, it is usually active and as such it’s the most likely articulation to not get recorded. Occasionally while recording I can see the first articulation get recorded as I play, but once I hit stop, all recorded articulations slide slightly to the left, and the first one disappears completely. It also seems to occur more frequently with retrospective record, but it can occur with regular record as well.
Other factors
None of the following options or configurations seem to have any affect on the outcome; the issue can happen regardless:
Expression map sound-slot activation mode: program change
Recording MIDI via normal record
Recording MIDI via retrospective record
Expression map with empty first sound slot with a remote trigger assigned
Expression map with empty first sound slot with no remote trigger assigned
Recording in the project window
Recording in the key editor
Transport > Return to start position on stop enabled
Transport > Return to start position on stop disabled
System Info
I’ve done exhaustive testing on Cubase v15.0.30 and Windows 10, but I’ve experienced the issue going back to at least Cubase v9.5 on both Mac and Windows.
Technically you don’t really need an articulation in the first slot at all. If all groups are in a “reset” state it will go back to the first slot. So you can reduce lane clutter by not bothering to assign an articulation to the first slot and using “reset” to go back to that. Although, it is true, some people might prefer to have an articulation in the first slot. But in the example above, you don’t really need the “mallet” lane at all.
There are so many reasons why it’s vastly preferable to have articulation data for ALL MIDI notes, with no exceptions. What if I have a region that lacks the initial articulation, and I move/copy it to a later point in the project (possibly merging it with whatever region now precedes it?) There’s a good chance I will then have an incorrect articulation.
Whether or not one could conceivably work around this is a distraction from my initial post, which describes the bug quite succinctly.
There are always exceptions. A correct-looking score is one of them. You wouldn’t have a “nat.” or “ord.” above each staff at the beginning of an orchestral score for instance indicating that the notes are to be played in the normal way, you just wouldn’t put anything. The default articulation is the equivalent in this way. Similarly your screenshot is probably from some kind of mallet percussion library like a vibraphone etc, and you would not want to have to write anything on the first page of a score to have it play in the default way (with the regular mallets for the instrument).
I’m also not saying you shouldn’t want this, I very clearly stated that I can understand why people would want this. Myself, I would personally want it to record the necessary "reset"s to get back to the default slot correctly. But it could make sense to have a different behaviour for users depending on whether they have elected to have an articulation in that slot or not.
There are quite a few things I would like to see improved in the way articulations get recorded, and I’m not entirely clear on what the expectations of the default behaviour are. I’ve just been recording all articulations with the mouse myself, which is time consuming.
There’s also often a bit of uncertainty on whether to record something as a bug or a feature request because it isn’t always clear what the intended behaviour actually was. In my experience there can be a lot of things where it is working as intended but people feel it is a bug because it doesn’t do what they expect. The special status of the first slot was often viewed as a bug for instance (helped in 15 now by the clear label). I just wish we didn’t have to guess so often as to what precisely the intentions were.
