incorrect interpretation of triplets
took a bar of (4) 3 1/8th note triplets and made it 9 1/8 notes and a 1/4 note triplet of 3. That equals 6.5 beats and to its credit, it added rests to the right hand to also equal 6.5 beats
Obviously unusable
| Attachment | Size |
|---|---|
| shenandoah.mscz | 26.04 KB |
Comments
I find it hard to understand this score. There are many measures with more notes than the Time Signature dictates. Your "triplets" seem to be ordinary eighth notes that you have placed in groups of three. Did you follow the advice in the Handbook for creating triplets? Basically, you create a note or a rest of the duration of all the tuplets in a group and then divide the time taken by that into the number of notes you want in the tuplet (often 3 but can be different).
https://musescore.org/en/handbook/4/tuplets
In the attached file, is the measure with the notes in red something like what you wanted to achieve?
In reply to I find it hard to understand… by underquark
yes your 'red' measure is exactly like the scanned in music that I sent to be 'converted' to musescore.
I'm attaching the choral PDF and you can plainly see that the triplets are indicated as you did. My question is why did the conversion to musescore not pick those up?
Also, 4 part choral writing on two staves shouldn't be that hard to parse, but maybe I'm missing something. All of the up and down stems have the correct beats/rests in each measure. I'm guessing the program had issues with the whole notes.
In reply to yes your 'red' measure is… by rheckman3
There are a lot of erroneous measures in the scan result. E.g. all measures with the small gray '+' in the upper right corner are faulty, they are too long. Triplets are such a problem case, also the grace notes (e.g. measure 2 from piano) are often not recognized and shown here e.g. as rest.
> 4-part choral singing on two staves should also not be that difficult to analyze.
It is difficult. Recognizing notes from a scan is far from perfect, even if you create a relatively simple and perfect PDF score with Musescore and then scan it again, it will probably have errors. And it's even worse if you have a scan from a paper copy.
And the time it takes you to correct the errors afterwards can be longer than if you had simply copied the notes by hand.
Unfortunately, that's the way it is.
In reply to yes your 'red' measure is… by rheckman3
The transformation from a graphical to a music format is always problematic and you will need manual work even when using commercial programs. I tried your file with a locally installed Audiveris and most of the triplets will not be recognized and have to be corrected.
In reply to The transformation from a… by Mr Fox
Thanks for the input. I had used musescore 5 years ago with a full orchestra score and spent a lot of time correcting (though not as much as re-entering from scratch!). Was hoping that in the intervening 5 years, music OCR would have gotten better. It is an inexact science to be sure.
Thanks again