Routines, using OMR (Optical Music Reading) to turn pdfs from IMSLP Archive into correct Lilypond.files

Steps in short:

• Choose a pdf, image with readable music. If needed improve the quality of the file.
• Choose an OMR app. (for me: Audiveris or SheetMusicscanner (IOS))
• Use Audiveris to create MXL file(s).
• Import MXL into Frescobaldi, use optimal CLI settings
• In Frescobaldi improve the quality untill satisfied.
• Finished

Setup of apps

• Folderstructure
• Audiveris 5.6.1 (special update because of MacOs error)
• Frescobaldi latest version
• xml2ly.sh to set xml2ly import to correct settings.
• Arranger.ly from Gilles Thibault https://github.com/gilles-th/arranger.ly
• Several snippets in Frescobaldi + Lilypond

Tips & Tricks

PS

My idea to work with arranger.ly was not succesfull up to now. Maybe it was not such a good idea. Inputting the instructions is maybe more work, then just editing what is wrong :-(

For now I work like this:

• The imported into lilypond transferred xml file is turned into 3 versions:
• the imported one
• a version with only the notes, no slurs, nothing (can be done with Frescobaldi easyly)
• a version with only s notes + slurs + fingerings + all . I use the regex string to select all notes and and change this in one single click. I have these 3 in one combined staff. And the original pdf on my ipad , so I can compare and edit easily. Specially all the small articulations like the dots . . . . are much easier like this. In the end I will combine the last 2 into a new file. << 1 \ 2 >>