Hacker News new | ask | show | jobs
by jhasse 1506 days ago
IMHO that's a very bad example because there's a big loss of information with Google Translate (you wouldn't get the Urtext when translating back), but there isn't with Clairnote.
1 comments

Yeah, if the original score is digitally encoded such that everything really is a first-class citizen (and not, say, a GIF of some scribbles) -- a very big if -- but if so, it's not obvious that any information has to be lost in translation to Clarinote, nor that Clairnote would impose any additional information. If you want to know the scale you could encode the key signature, which Clairnote seems to make possible, just unnecessary.
From the earlier reply:

distinguished voices (stem up v down)

That's a very important one that might not survive across re-transcription. Although in this case, Clairnote is similar enough that it might generally be possible to retain all the stem directions.