[MusicTlk] In-accords/Multiple Voices in Advanced Piano Repertoire

[Ella Yu]

Hey everyone,
Something that has recently been on my mind is how in-accords (both full
measure and part measure) are used in braille music and how automated
conversion tools handle this sort of thing, as I do find this to be one of
the most cumbersome/difficult aspects of braille music, at least for me
personally. I think it's because with in-accords, voices are written one
after the other, so you really have to be able to analyze a line and figure
out what goes where before piecing it together in your head, which gets
more difficult with more complex music, whereas the more spatial nature of
print music perhaps makes this sort of thing somewhat easier to digest.
Anyhow, my main question is, is it reasonable to say that deciding when to
use full measure in-accords and when to use part measure in-accords
particularly in complicated situations is one of those things that is very
discretionary, meaning it is ideally up to the transcriber to examine each
situation on a case by case basis to decide what is the best way to write
it out so it's easy to read? There are definitely carefully laid out rules
for using in-accords in braille music codebooks, which is what automated
conversion tools are based on, but still, there is lots of room for
interpretation, right?
I'm using a very specific example to illustrate this issue, which is why I
have attached several .brf files. The piece in question is Ravel's Sonatine
for piano, which is a pretty advanced/complicated piano piece I learned a
while ago. From a reading standpoint, it was probably one of the most
difficult/mentally taxing pieces I've learned due to the complex interplay
of voices between the two hands and how that is presented notationally. I
know there are plenty of piano pieces that are similarly difficult, if not
more difficult, than this. The three .brf files, which are all braille
music transcriptions of the Sonatine, are:
1. A transcription from brailleorch.org from the year 2018. I believe the
person who made this transcription uses automated conversion tools, but
does manual editing afterwards to optimize things. It's the easiest to
read/the most clean, but back when I learned the piece, I didn't really
know brailleorch.org existed, like, I knew it existed, but didn't really
dive deep enough to figure out what is in its collection, though now I know.
#2 and #3 were created from the following source on musescore.com, which
looks to be a well done/high-effort transcription:
https://musescore.com/user/3797871/scores/5111461
2. A version converted via braillemuse.net, which is the copy I learned the
piece from. I edited it to try and clean up some things and stuff. Back
then, Sao Mai Braille (which is a much better music braille translator
overall than BrailleMuse) wasn't really available yet. It is the hardest to
read.
3. A version that I quickly made today using the latest version of Sao Mai
Braille. I apologize, it has lots of unwanted tempo indicators, a product
of certain musescore.com uploads that have extra tempo marks to optimize
playback, which is something I cleaned up in the BrailleMuse version back
when I learned the piece. Sao Mai has tons of settings to optimize the
results, so part of it is knowing what settings to adjust to get what you
want, but because the program has so many of them, it can be overwhelming.
For handling in-accords, it has three transcription modes: part-measure
with stem signs, full and part measure, and full measure only. I forget
what the default is or what is most ideal when, but I chose the "full and
part measure" option, thinking it was the best. I know in-accord stuff is
something the Sao Mai team has continued to optimize over time through
software updates. Anyhow, in terms of how in-accords are handled, the Sao
Mai version looks a lot more like the BrailleOrch version, so still easier
to read than the BrailleMuse one. I did notice that if I translated the
piece using "part measure and stem signs" in-accord mode using Sao Mai, it
would look much more like the BrailleMuse version, making it harder to read.

So as you can see, with complex multi-voice music, there are multiple ways
to interpret it when translating from print music to braille music, and
some representations are more ideal/easier to read than others. The
BrailleMuse version is the hardest to read in my book because many measures
constantly switch between part measure in-accords, no in-accords, and then
part-measure in-accords again all within a single measure. On top of that,
there are a bunch of 16th note stem signs that clutter things up. All of
this makes many measures much longer than they need to be in braille, at
least based on my reading experience and what I know about the code rules.
To sum up: Is complex multi-voice music one of those things that is often
up to interpretation in terms of transcribing it in a way that is easier to
read more than other items/situations in braille music? I definitely feel
like the answer is yes. I love automatic MusicXML to braille programs, as
they have given me access to braille music much more easily than if they
didn't exist. They do a remarkably good job these days with standard
notation and easy to higher-intermediate complexity material, especially
Sao Mai Braille, but I'm definitely thinking for certain situations where a
much greater degree of discretion in interpretation is involved from a
transcription standpoint, such as less/non-standard symbols, complex
multi-voice material, and perhaps certain situations in vocal music and
percussion music, having the input of an experienced braille music
transcriber or proofreader is generally invaluable. What do you all think?
All thoughts are appreciated.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Ravel Sonatine BrailleOrch Version.brf
Type: application/octet-stream
Size: 16243 bytes
Desc: not available
URL: <http://nfbnet.org/pipermail/musictlk_nfbnet.org/attachments/20241110/19f43fa9/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Ravel Sonatine BrailleMuse Edited.brl
Type: application/octet-stream
Size: 22387 bytes
Desc: not available
URL: <http://nfbnet.org/pipermail/musictlk_nfbnet.org/attachments/20241110/19f43fa9/attachment-0001.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Ravel Sonatine Sao Mai Unedited.brf
Type: application/octet-stream
Size: 28630 bytes
Desc: not available
URL: <http://nfbnet.org/pipermail/musictlk_nfbnet.org/attachments/20241110/19f43fa9/attachment-0002.obj>