Measure rythm incorrectly shown as wrong

[olin256]

It's the measure 4 in this OMR file. It is displayed as "pink", although the dump shows no issue:

MeasureStack#4
    |0       |3/4     |1       |5/4     |3/2
--- P1
V 1 |Ch#3869 |Ch#5018 |Ch#5020 |Ch#3870 |3/2
V 2 |Ch#3868 |Ch#5161 |Ch#5163 |Ch#3871 |3/2
V 5 |Ch#3909 |Ch#5019 |Ch#5021 |Ch#3911 |3/2
V 6 |Ch#3908 |Ch#5165 |Ch#5167 |Ch#3910 |3/2

(Note that there are "hidden", i.e. overlapping quarter rests; I don't know any other way to assign a single rest to both voices.)

[SofNotenwart]

I'm by no means a regular user or expert in Audiveris, but opening your omr-file in Audiveris and exporting it to MXL and finally opening in MuseScore as well as in Lilypond, both show a correct measure, see below. In fact, the Lilypond result is identical to the input file. Maybe the colouring in Audivereis is just incorrect. Hence, one may conclude that the MXL-file is correct and the colouring is just an artefact.

For shared heads, a line is drawn diagonally through the head, see handbook https://audiveris.github.io/audiveris/_pages/main/voice_colors/ . Maybe a similar feature is embedded for rests although, in this case, there is no diagonal line visible.

By the way, you have exactly the same "constellation" in measure 8 but without any colouring.

Result in MuseScore

Result in Lilypond

[hbitteur]

Hmm, I don't like your "hidden" rests.
In the measure 4, I have deleted your manual rests and entered new separate ones.
Then, with a bit of voice linking (using the Chords popup menu), I got this correct result:

[olin256]

I don't like the hidden rests either, perhaps it's better to put new ones separately (a feature for sharing rests between voices would be nice btw…).

However, I've tried to delete them (both the original ones and the manual ones) and replace them with differently placed ones as you did, but the measure is still pink – which, I believe, is clearly a bug…
019_new.zip