Re: Gt-tabs and drum-notation

[David Raleigh Arnold]

==========Rune Zedeler
That's what I have in mind. Don't want to enter the same piece of music
twice.(i.e. what note should be transferred to
fret-0)
---------yhs
Sorry to have crossed messages.

Yes you do, sort of. The tab is *not* music. Only the location of the
finger on the string is given. You don't see what it sounds like.
Furthermore, the time value is not given to an
individual note. An e on a guitar (I always use G_8 for guitar music.)
can be played as the open 1st string, the 5th fret of the 2nd string,
the 9th fret of the 3rd string, the 14th
fret of the 4th string, and the 19th fret of the 5th string, at least.
Only the last is unusual. By
including a LH finger indication you can rule out only the open string.
To be really specific, which you must be to be useful, you need to
specify a string for every note. I
don't think you (or Han-Wen) want that scrambled into the \notes,
especially if the music has multiple parts on a single staff, or has 4
or even more parts on two staves as lute music does. If anything
additional should go in the notes IMHO it is midi velocity, which is an
unambiguous quality of the note itself rather than an instruction to a
human being telling how to play the note on a particular  instrument,
and I'm totally outvoted on that one. 

Also, what about instruments which have two unpaired strings tuned the
same? How can a program guess the correct string? It can't.

Notes, *only* two parts:
|   |
o   o   
* o   *
| |   |

as tab:

0---5-- 
0-1---- 
------2 
            

Could equally well mean:
|   |
* | * |
* *   *
 or a number of other things.

So tab only has one part, no matter what \notes data is parsed.

So multiple voices are often impossible for a real human being to sort
out of tab.
English lute music often came with a complete arrangement of vocal
parts, making it possible to
find the composer's intent. The exact correct interpretation of some of
it without vocal or keyboard versions will never be known.

It would be easier to give the string a letter and the fret a number.
z0 z2 z4 should work. Or just z0 2 4.
1-0 1-2 1-4 is nasty, isn't it? 
zg za zb  or  zsol zla zsi would be better for real musicians. :-) zg a
b would work too, or gsol la si, even gg a b would be nice, if the first
g indicated a g string.

tablines = nopqrs 
would designate the 1st string as n, 2nd as o, etc., for easy data
entry.

tablines = nopqrs ( e b, g, d, a,, e,,)
would also give an open string note for each string, from which the fret
number may be easily calculated, to quote David Roundy.

tabline x = e f# g# a b c  or even tabline x = bok frba zbic afd
would designate one string x  as e or bok, the first fret as f# or frba,
the second fret as g# or zbic, for anyone who was using different
fretting and/or oriental notes, for example. This is to simply set a
note for each fret on a string, one line for each string with the 1st
string the 1st line. This is a bell and whistle, but why not have an
alternative to the default of calculating the frets chromatically? And
of course one could always just do the numbers instead for hairy
instruments.

The 1st string on any instrument which has a neck is the one closest to
the left hand. Consequently it is usually, but not always, the highest
in pitch. Some of the exceptions include 5 and 7 string banjo, guitarra
battente, *renaissance* *and* *baroque* guitar, and ukulele (which is a
renaissance guitar). No program can automake good tab out of music for
those, ever.

Thanks a lot for asking. I hope this helps clear things up a bit. If
not, I am completely willing to keep at it for as long as it takes, and
consider that you are doing me a favor. :-)