Importing xml files from playscore2 into musescore not even close to accurate
Recently, I've tried importing xml files derived from playscore2 (correct in playscore2) into musescore with horrible results. A single instrument part (band music) contained mostly eighth note, quarter note, or mixed note triplets with interspersed rests. The playscore2 file (from scanning is excellent). I output xml files from playscore2 and import them into musescore yielding a useless result. For instance, triplets are completely lost and the time signature changes from 4/4 to 6/4. This renders musescore useless for light cleanup or playback or audio file construction.
I have difficulty discerning if this is a playscore2 xml file output problem (which I doubt) or a musescore input issue. I've done alot of hand editing to produce choral audio files and want the playscore2 to musescore bridge to work flawlessly as possible.
What is going on? If this is a glaring musescore issue with processing xml files, why is it not corrected?
I am not uploading the file due to copyright issues.
Bill L
Comments
Not much anyone can do without a score. Perhaps you could create one that has this problem and post it.
"I am not uploading the file due to copyright issues"
This is just one of very basic information that may help people to help you. Basically, you're on your own unless you can provide better information.
"For instance, triplets are completely lost and the time signature changes from 4/4 to 6/4."
PlayScore 2 claims to handle the following features :
- Tuplets: triplets, duplets, quintuplets, septuplets etc (both marked and implied)
-Time signatures: system, inline, cautionary and implied
See https://www.playscore.co/blog/faq-items/specification/
So why not show us the results of a test run with plenty of typical "problem" notation such as triplets?
a) the original PDF score
b) the MusicXML generated by PlayScore2
Once we have some MusicXML to examine, it should be easy to identify whether the problem lies with PlayScore's MusicXML export or MuseScore's MusicXML import.
In reply to "For instance, triplets are… by DanielR
Please give me a day or so to contact the concert band coordinator as the group has very strict rules on copyright related issues. Once I hear from them on what they will allow me to provide I will do that.
THanks for your patience, it was a fairly significant failure.
I've done alot of musescore editing of choral music (to produce audio files with piano part) and I've noticed that triplets are often done wrong in musescore input of xml files (I switched to playscore 2 xml input from direct pdf scan input into musescore this fall as the direct pdf iconversion into musescore is even worse than my example for you.
In reply to Please give me a day or so… by wjlutter
OK, I will send a short segment to illustrate. It is a clarinet instrumental part for Pirates of the Caribbean: At World's End. All triplets are lost in translation from playscore 2 (OK) to musescore via output xml which is then input into musescore producing the musescore file type file. In the full pdf, the musescore conversion shows correct time signature of 3/4 at beginning and then changes a 4/4 to 12/8 at where measure 27 should be (it messes up the 9 measure rest). Following 27 are the lost triplets. For the segment that I've attached, I whited out lower first page using sedja free pdf editor, ran it thru playscore2 (correct). Now the musescore time signature is 3/2 at beginning and undetermined where measure 27 should be. Triplets lost again.
In reply to OK, I will send a short… by wjlutter
The PDF shows 36 bars, no repeats and tie dotted half notes in bars 24 and 25.
The xml file shows 26 bars, a Start repeat but no End repeat and dotted quarter notes instead of dotted halves.
The mscz file largely follows the xml.
The PDF starts with Time signature 3/4
The xml says "divisions 4", "beats 3" (so far, 3/4) then "beat-type 2" (making it 3/2?)
The mscz file starts with Time signature 3/2
In the PFD, the dotted half notes are indistinct and easily mistaken for doted quarter notes by a human and very likely to be misinterpreted by a machine.
Unless I'm wrong, these are different files or the scanning from PDF or conversion to xml (or both) is flawed and MuseScore is trying to make the most of interpreting inaccurate input data.
Checked the xml file in detail: what MuseScore shows is exactly what is in the xml file. Used MuseScore 4.4.4.