Support for cleaning up imported PDF score

• Oct 25, 2022 - 22:06

I recently tried the "read PDF" service on a score. The result looks very promising and would save me a lot of work keying in parts of the score (for personal use certainly). The "notation" looks quite good, but the service missed a lot of the n-toles, i.e. we have quite a lot of irregular bars as the notes do not sum up correctly.

I assumed, there is something like "mark these notes and make an n-tuple of it" what would get the timing right again.

I hope I may post three lines of the score here without going to jail.

original score:
2022-10-25_22-50.png

converted result:
2022-10-25_22-52.png

There are some missing rests and beat changes too, but these are easy to insert. It's mostly the n-toles that I cannot create from existing notes. I just can delete them and re enter them manually. Can that be? That would result in deleting most of the already "good" score.

I had worked with Sibelius for quite a while where I also can OCR a PDF to a score - which is of similar quality. But there I just mark the notes, type ctrl-6 and voila, the wrong part is a 6-tole again, the "bad" bar sums up to a nice 4/4.

Did I miss something? Is there any add-on that does this? Am I out of luck? Nobody else had this kind of problem before?


Comments

It's a common problem indeed, given that PDF import tools are ttill pretty primitize. And there is a plugin "duration editor" that can help to some extent with this. But realistically, most of the time it's faster to simply enter the music normally than to try to correct the many errors in a PDF import.

In reply to by Marc Sabatella

Thanks for the tip. I will search and try that one.

Meanwhile I found that Musescore is quite stubborn when I try to correct some of the "simple" errors, like a wrong time signature. It does not simply change the symbol but obviously tries to force the rest of the score into new measures, setting barlines anew "at will" ultimately ending on a "cannot span n-tole over barline".

So maybe the add-on will not help me out on that problem at all.

To correct such a score it would be much preferable to simply turn off all such safeguards and automatism.

Marking "irregular bars" would be enough to spot remaining errors.

The PDF scanning service does not match well with the editing capabilities that are necessary, since nobody can expect a perfect result from the OCR.

Maybe I'm using the wrong tool. :-( Thanks for the help anyway.

Do you still have an unanswered question? Please log in first to post your question.