Voce to text/musical notation

• Jan 27, 2011 - 00:33

I know I have dropped several feature requests since I aquired this software, but I am so infatuated with it zI cant hep but! I have searced for years for a program that even comes close to this. Just a couple of more thoughts, If I was fluent in Javascript I could probably do it, I could in visual basic. 1.)is there a way to incorporate a speech to text/musical notaion simula to inputing from midi, and 2.)is there any way to convert from wma format back to muscore format. I do LOve this program, and give thanks to all who made it possible, Have a wonderfull day
David Smith


Comments

WMA is an audio format, is it not? If so, then there is not going to be a way to automatically convert to MuseScore - that's like trying to convert from a picture of a car to a car.

A single melody line, played with a flute or sung, is not that hard to turn into midi data, if the audio signal doesn't contain anything else. The more stuff you have there, the trickier it gets. There's no way to "learn" a computer to listen to music and really have it distinguish the single tones that should be turned into midi note events. Just think of how a church organ sounds. One single key stroke causes several pipes to sound, depending on which stops you use. The human musical ear thinks of it still as one musical note. If you play a single melody line with a complex registration, you listen to it as if it were one single melody line, though each note is produced by several pipes. On the other hand, some times one single tone, produced with one instrument, can create such a rich sound spectre, that some partial tones pop out as if they were played by another instrument.

There's some development going on in this field, but the bottom line is that it will never work like text recognition software (where you place a sheet of printed text in your scanner and your software translates it into a Word document without errors). Instead the developers are working on tools that can only help out at their best. You import some audio, you get some working screen, where you can confirm or disqualify some suggestions that the software makes for you. You might have to adjust audio filters to emphasize things in the audio that are relevant and remove things that are not.

I'm a very impressed new user but also just migrated to Win 7, which has built-in spech recognition.

I was wondering if anyone has used Win 7 speech recognition (or other such as Dragon) to enter notes by getting to position on the score and then simply saying the note lengths and notes, e.g. 5AB6C5DEFG, etc.

It could be quite quick and accurate compared with keyboard/mouse entry and could vastly help people with limited dexterity.

Richard G

Do you still have an unanswered question? Please log in first to post your question.