As I related in a previous article, we have been working on making MuseScore accessible. It is a large project as MuseScore was never designed with accessibility in mind, and the Qt framework used by MuseScore has a number of issues that make it difficult to produce fully accessible applications of any complexity. Plus of course, simply deciding how to present standard music notation in an accessible way is a difficult problem in itself.
However, as of yesterday, the MuseScore 2.0 Beta 1 release is available, and the first round of accessibility support is in! It has been tested with NVDA on Windows and works well with that. There are mixed reports using Jaws, VoiceOver (Mac), and Orca (Linux). We have Andrei Tuicu – a Romanian student working through the Google Summer of Code – to thank for the new accessibility features. I (Marc Sabatella) served as his mentor for the project.
So far, all that is fully accessible is score reading functionality – the ability to load a score and navigate through it while hearing the contents of the score read aloud. The basic menu commands should be self-explanatory, although there are still some glitches with how the menus read due to what appear to be Qt issues. It seems no worse than many other programs, though. You can also use the File / Export menu to convert to other formats, hit Space to listen to play back, or hit Tab to access the program toolbar.
With a score loaded, the right and left cursor keys move forward and backwards through the score a note at a time, and the screenreader should read the notes aloud (pitch, duration, etc). In addition, when using the cursor keys with Ctrl+Alt+Shift, it will move not just note by note, but will also include the clefs, key signatures, time signatures, and barlines. Information about articulations, dynamic markings, lyrics, and so forth are read along with the notes they are attached to.
The MuseScore team welcomes feedback on the new accessibility of MuseScore 2.0 Beta 1. The best way to do that is to post a thread to their Technology Preview forum.
We all realize, of course, the score reading is only part of the battle. Next up will be to make score creation and editing accessible. Actually, it is partially accessible already, but many important markings – key signatures – will be difficult or impossible to create. Still, when MuseScore first starts up, there is a blank 32-bar score already loaded (key of C, 4/4) and you should be able to start typing in notes at least. Press “N” to enter note input mode, then select a duration using the number keys on your keyboard (3-7 are sixteenth note through whole note), then type a letter name for pitch. There is of course a lot more to note entry than this, and again, only some of it will work without use of the mouse, but that should get you started and give you an idea of the potential.
The Accessible Music Notation Project 
I know this is not the right place to ask, however I have not been able to get a single note read to me neither by NVDA nor by Jaws. I was overly excited to hear MuseScore providing this functionality in the beta version. Of course it might be expected that the stable version will have overcome this little issue, but is there a specific way to get around to reading the notation? For the moment navigating with arrow keys speaks out things like “selected checked”, title of the file, some numbers etc.
It is indeed better to use the Technology Preview forum on musecore.org to report issues with this Beta release.
Anyhow, I think the issue you are having is getting the focus onto the score in the first place. Depending on how you start, the focus may or may not be there. That is a known issue. The most reliable way to get focus into the score is to press “N” to enter note input input mode, then immediately “Esc” to exit. Now focus should be where you need it. Left and Right arrow keys will now move note by note, or Ctrl+Alt+Shift+Left and Ctrl+Alt+Shift+Right to move “element” by element (to hear about clefs, key signatures, barlines, etc).
NVDA should work well. Jaws currently does not – you here the first status update, then no more after that.
BTW, Andrei is hard at work implementing accessibility for actual score *editing* (as opposed to just reading) functionality. We don’t at this point know if any of this work this will make in into the official MuseScore 2.0 release, but it seems quite possible some of it could be working soon afterwards and made available as a series of “special” builds, and then incorporated into the next official release (ie, 2.1).
Mark, many thanks for taking time to give an extremely helpful explanation. Just for the benefit of people following your blog, is the technology preview forum the only place of reference for those wishing to join in the testing of upcoming beta releases? I have such a good feeling about this: If we manage to get MuseScore as functional as one of those free music editting softwares all over the internet, I believe commercial titles with unrealistic accessibility concerns might also think about investing more effort into the accessibility of their software, Which would of course increase the overall quality of blind musicianship
Ugur 21.9.2014 00:20 tarihinde, The Accessible Music Notation Project yazdı: > WordPress.com Marc Sabatella commented: “It is indeed better to use > the Technology Preview forum on musecore.org to report issues with > this Beta release. Anyhow, I think the issue you are having is getting > the focus onto the score in the first place. Depending on how you > start, the focus may o” >
The Technology Preview forum on MuseScore is the *right* place to discuss the beta releases of MuseScore. It is of course possible people will see and respond to comments elsewhere, but less likely. Why take chances?
wow, I am very, very impressed with this work, and over all, with this blog and even more with the really sensitive aspects you have as a teacher, Mark. I stumbled upon your jazz improv. primer years ago and now I am quite shocked at the possibility that accessilibity in notation would have interested you so much. I am a blind jazz musician too myself, and I have been using lilypond for years. The options that are specific for the visually impaired that I know of for producing music are about 7 or so, between ridiculously high-cost and often ineffective programs (expecially in the world of jazz writting) and open source or free, most of which have been discontinued. In light and spite of all of this, i have been using lilypond for years, since at least 2010. The changes are ever evolving (no musical name pun intended) and quite great on many aspects. Lilypond has been my number one method of producing scores for publishing, selling, giving to other musicians, composing, etc. It has been so engrained into my way of thought that I have created: 1. A blog in spanish especially adapted for the visually impaired, in progress; 2. Many templates for my own personal use that I think could benefit others; 3. Ways of working with different text editors with functions such as macros that can speed up some tasks when coding. 4. I have gathered and found different ways to convert existing materials that a teacher could produce without leaving finale, sybelius, etc. into lilypond, making the score a completely described rendering on what was notated before, plus there is always an existing library of growing classical and public domain work available at the mutopia project.
Anyway, for entering chords and the like I find that I can probably transcribe a form with its corresponding harmony and rhythms or full ensemble stop, et. in under 10 minutes using lilypond. Previously I used to do this in band in a box (I would write the melody in lilypond, then import the midi and type in the chord) because it was a fun and interactiva way to do it, though of course everyone would just use a Word processor for entering chords only. However, I have ditched anything that has a GUI and even though I could understand how museScore editting would have been a nice starting point, I still think that lilypond is definitely way faster tan anything i have ever tried. Plus, its output and methodology for entering music are very flexible. Maybe I can, if you allow and wouldn’t mind, bring this alternative into consideration as a post on this blog. I think both museScore and lilypond could be bridged and bring endless possibilities to the table.
Thanks for your comments! FWIW, I do mention LilyPond in the resources page. I had considered it when first working with a blind student a few years back, but found that ABC was more easily readable / writable / learnable, so tht was what we settled on. Either way, indeed, for the right type of musician, this can be very efficient and perhaps preferable to a GUI-based program. But I think not everyone is that type of musician. Obviously, some of the benefits of a GUI are not realized if the user is blind, but the musicians we have worked with during MuseScore development really rely on the ability to hear instant feedback on each note of the score. Anyhow, it is certainly good to have options!
of course. I used BME and was interested at the possibility of hearing a not e right away, but then as time went by I started to rely less and less on the auditory feedback-and that has helped me grasp music better in a more intuitive level. I think that the auditory feedback is better off when composing or arranging, but I believe that there are certain things you wouldn’t have control over when using a gui interface–for instance, the exact Font size for each note (in some cases that is recquired) or the ability to easily format text, color notes, change titles or rehearsal mark text, etc. Still, with museScore support for lilypond export and import I would think that would be posible to do in museScore in the no too distant future. I admit lilypond can carry over a bit of a learning curve, but the obtained results are certainly well deserved
today mussescore is 2.0.3 version. I would like ask if with next version will be possible edit scores?, are there people working yet in this progect?.
It’s still a goal for a future version. The next version will probably be 2.1 and it will have a few more accessibility features, but still not enough to make full editing feasible, unfortunately.