Re: Ancient tablatures

[Laura Conrad]

>>>>> "Marc" == Marc Hohl <address@hidden> writes:

    >> I've made some efforts to transcribe the tablature, but what I want
    >> ideally is to transcribe what's there, in an input form that doesnt'
    >> require me to translate the tablature into notes, and then use that
    >> transcription plus the tuning of the strings to produce both a
    >> tablature that looks like the one in the facsimile and standard
    >> notation that a modern keyboard player could deal with.
    >> 

    Marc> That's an interesting point - I think Dana Emery posted to
    Marc> the users list that writing tablature as normal notation and
    Marc> letting lilypond do the translation into tablature is at
    Marc> least not always the best way.

As a transcriber, I would *always* prefer a transcription method that
doesn't depend on me doing complicated calculations from my input.  In
the case of tablature, lots of the real users have never used it in a
way that went through note names in their heads, and then there are
transcribers like me who've never used tablature to play from at all.

Of course, if you're transcribing from standard notation, and just
want the tablature as output, you have a different problem.

If you've ever known a guitar player who has trouble playing a G when
asked, you can imagine that a lute player might well be playing a
tablature of a 4-part polyphonic piece without really knowing all the
note names.

So an input method that says "play the fourth fret on the fourth
string" is definitely better than one that says "play A", even if
you're also saying "the fourth string is tuned to F".

    Marc> Hm, then let's try to nail it down: how would you like to input
    Marc> tablature? 

I think people should look at abctab2ps, which I mentioned the other
day, but when I was actually entering that, I found it a little
cumbersome.  And of course, it's based on ABC syntax, and we'd want
something based on lilypond syntax.

The feature that was cumbersome was that you had to treat every
"column" in the tablature as an ABC chord.  It seems to me that the
natural input method for tablature would be something like:

        here's a new column

        Here's the length of this column (actually the time between
        the start of this column and the next one, as strings that
        aren't refretted by the next column can be left to vibrate).
        This would be optional, as it wouldn't be necessary to repeat
        it if the "length" were the same as the previous column.

        Here are the strings that are fretted in this column. I think
        the abctab2ps for that is quite compact and straightforward.
        A string that's skipped is notated with a comma; one that has
        a letter uses that letter; if there are unfretted strings
        below the last fretted one, you don't have to specify them. 

So here's some abctab2ps:

   [,db,,d2][,,,,,d1][,b][,ab,,d2][,db,,d1][a]

Note that in ABC, there's a default note length, so the "chords"
without a length are the default length (in this case, 4), not the
length of the previous note.  And in ABC the [] surround a chord.

I found writing all those square brackets to indicate new columns a
pain, but of course the right editor macro could have fixed that.  And
of course, if we were doing it as part of lilypond, we'd want to use
lilypond-like chord and length indications.  

And of course instead of just asking someone who's transcribed a few
French Tablature pieces in abctab2ps, you should also ask people
who've used the other tablature generating programs (Fronimo,
Finale...), and people who've used Italian and German tablatures.

    Marc> or we can use some converters which translate the lute
    Marc> tablature into lilypond syntax, which again translates this
    Marc> into a nicely formatted tablature.

I'd find that somewhat baroque, which might be suitable.  I did think
for a while about writing a python program that would produce ABC (and
then lilypond) from abctab2ps.

I wrote most of this yesterday, before the discussion between Carl and
Mark had continued.  I agree that it's more important to get the
output right than to have an input mode *right now*.  Maybe the
tab2abc and abc2ly route does make sense, if lilypond gets to where it
can produce something like the original.

-- 
Laura   (mailto:address@hidden)
(617) 661-8097  233 Broadway, Cambridge, MA 02139   
http://www.laymusic.org/ http://www.serpentpublications.org

I've been singing. I've been taking a part in White sand and grey
sand.  I don't know anything about it. Never mind. I'll take any part
in anything. It's all the same, if you're loud enough.

Charles Dickens, _Little Dorrit_