Incorrect (?) measure lengths on import of xml file with changes in divisions value
Hello
Whether on MuseSore 2 or 3, I have a recurring problem reading xml files from Neuratron PtotoScore Ultimate.
I use this OMR (Music OCR) program a lot, which allows me to transcribe any PDF file into a music file. In terms of OMR it is expensive - 269 € - but it is clearly the most successful. All the opera choirs I put on MuseScore have been transcribed with this program.
I scan a quality paper score into PDF or upload a PDF found on the net directly.
PhotoScore recognizes it, I correct the remaining errors and export in xml. This exported xml opens well in Finale, Sibelius, Harmony Assistant but not in MuseScore! Why? I can't find it.
I am therefore obliged to make a second xml export from one of the programs that opened the first one but it is tedious and increases the risk of errors.
In addition, I am more and more addicted to MuseScore and I would like not to have to use the money pumps that Finale and Sibelius have become.
Who can help me?
Attachments :
1- The original PDF from the RICORDI 42602 edition of Aïda piano-chant Aïda58-62H.pdf Aïda58-62H.pdf
2- The.opt file created by PhotoScore (jpeg of the first page) 
3- The.xml file exported by PhotoScore Aïda58-62H.xml Aïda58-62H.xml
4- This.xml file opened in MuseScore (2 or 3) and saved without editing Aïda58-62H.mscz Aïda58-62H.mscz
5- This.xml file opened in Sibelius and saved without editing (jpg of the first page) 
6- This.xml file opened in Final and saved without editing (jpg of the first page) 
7- This.xml file opened in Harmony Assistant and saved without editing (jpg of the first page) 
Apparently MuseScore's mistake is made here : 
Thanks for your help
Comments
Looking at the file, I see it changes "divisions" value several times, and I'm not really sure what the correct behavior is supposed to be. But for the record,
That said, adding a "divisions 12" at the beginning of the fourth part (the one showing the one very long rest per measure at the beginning) does solve the problem. Not sure if this should have been required or not.
In reply to Looking at the file, I see… by Marc Sabatella
Thank you, Marc, for looking at this closely. Unfortunately, I'm not knowledgeable enough to be sure I understand your explanations. I can fix the partition by imposing a 4/4 at the beginning and then deleting the empty measures it creates but I would prefer that MuseScore manages to open this xml file correctly.
Indeed, my comments were as much for other programmers' benefit as yours. What I will say is, to me there is 50/50 chnce the fault lies with the program that exported the file, but anyhow, in order to work around the problem, you would need to edit the file in a text editor to add the line I mentioned. Or, just use this one I edited in my testing :-)
In reply to Indeed, my comments were as… by Marc Sabatella
Thank you, I understand better:)
I agree that the fault may come from PhotoScore. Unfortunately I have no way to set up the xml export in this program.
Thank you for your corrected attachment. I personally solved the problem, as I said above, by opening the PhotoScore xml file with Sibelius and then exporting it to MuseScore. It works very well and here is the finalized file attached.
But it seems to me that there is still a MuseScore problem since it is the only music editing program that does not open this xml correctly
Aïda58-62.mscz
Could be, or it could be that we are the only ones interpreting the standard correctly :-). Wouldn't be the first time. Most likely you are right, but I just don't know enough to say for sure.