Unable to convert pdf
Apologies if this has been covered already. My pdf score was pronounced unsuitable though it's clear, 2 voice song. Is there an interim format I can convert the pdf to before trying Musescore with that? Many thanks
| Attachment | Size |
|---|---|
| AgniParthene.pdf | 126.25 KB |
Comments
"My pdf score was pronounced unsuitable though it's clear, 2 voice song"
It is a sad fact that the MuseScore function File > Import PDF... is relying on a very out-of-date version of a third-party application for optical music recognition (OMR). The OMR software is written by Audiveris.
Updating the Audiveris-based online function Import PDF is really not a priority for the MuseScore developers, who are fully occupied with writing code for MuseScore Studio itself.
But you can download Audiveris for free and set it up on your own PC:
https://github.com/Audiveris/audiveris
Meanwhile, attached is the result of running your PDF through a commercial OMR programme. I haven't made any correction, I just imported the MusicXML output by the OMR.
In reply to "My pdf score was pronounced… by DanielR
I ran it through SmartScore Pro 64. Corrected a couple of very minor items, but I didn't go through the entire score. Text was funky in a couple of places, but the notes and rests and most of the lyrics were where they belonged. Five minutes, tops.
There are a number of music OCR apps, but they are not free. I use SmartScore 64 Pro, but there are others.
I have seen references to downloading the open-source app that MS uses, and running it on your computer, with better results, but I don't recall the name. Maybe someone else will be able to help.
(Daniel beat me to it!)
You wrote:
My pdf score was pronounced unsuitable though it's clear...
Much of the available OMR software produces output which needs corrections. Many errors involve incomplete (or even overfilled) measures. Your pdf does not show time signatures at all - and also contains pickup (incomplete) measures. This could confuse an OMR from its very first stages of processing which is why a "one pass" conversion is suboptimal. It does not allow for making corrections before final output to ensure better results.
I tried this PDFtoMusic software:
https://www.myriad-online.com/en/products/pdftomusicpro.htm
It has a free trial version (which only converts a single page). I put your pdf file through 8 times and stitched the xml output into a single MuseScore file.
Here it is:
AgniParthene-needs_work.mscz
Compare it to your pdf to see where I left off formatting the system breaks and further add your own. Also check to see that the lyrics are okay.
You also wrote:
Is there an interim format I can convert the pdf to before trying Musescore with that?
There is no interim format before trying the MuseScore "import" of a pdf. MuseScore does not read the pdf, it reads the "one pass" conversion to xml which is performed by Audiveris. So actually, xml may be considered the interim format before MuseScore can use it.
Regarding pdf...
A pdf can be made from a scanned image (e.g., of a paper score), or it can be exported from a score writing software as a pdf document.
See the picture here to determine the difference when viewing a pdf score in a pdf reader:
https://musescore.org/en/node/261949#comment-806203
Some OMR's are designed for converting pdf's made from a score writing software. Others are good at image recognition.
Regarding Audiveris:
https://musescore.org/en/node/265326#comment-808903