Add Grand-Staff splitpoint as Configurable Item to MIDI import Dialogue
Type
Graphical (UI)
Severity
S5 - Suggestion
Status
active
Regression
No
Workaround
No
Project
When importing a midi piano track the splitpoint between lower and upper staff is fixed to C3. In some cases this leads to unsatisfactory results. In this case it is better to forgo the splitting option and to manually split after import. It would be desireble to change the check box "split staff = checked / unchecked" in MIDI Import by a dropdown list directly indicating the split point. The top most entry should be "none", being also the default value. This is equivalent to the box unchecked.
Comments
As far as I know, it's not true that the split point is hardwired to C3 - it's actually flexible and determined based on the content of the file, changing over the course of the music to try to keep everything playable. So one passage might have, say, the E above middle C in the right hand but another might have that same note in the left hand if that seems more playable based on the other notes in that measure. It's far more sophisticated than any single split point could be and produced better results in the majority of cases.
Nonetheless, I'm sure there are a few exceptional cases where a single split point might happen to produce better results, so it would indeed be nice to provide that option int he MIDI import panel.
BTW, see the following for an example showing how much better the current scheme can be than a single split point for the whole file:
split.mid
I am surprised that the splitpoint ist determined dynamically, but you must be right. I get different results when I split manually at c3. I attached some files to demonstrate what happens:
1) Reference: how it should look like
2) Result using auto split in MIDI Import
3) Result importing MIDI to single staff and splitting manually at f2
4) Result importing MIDI to single staff and splitting manually at c3
5) The midi file causing all that trouble ;-)
The score is very simple to my opinion, so I wonder how it manages to fool the auto split. I get closest to the desired results (1) when I use (3) and remove all those additional keys.
I'm coming across a similar thing, where the auto-split does not account for which hand the note is closest to. The worst offenders are marked in red.
![Capture.PNG]( "Capture.PNG")
Attached is the midi file for reference. I imported with one voice, no staccato, and split staff.