Re: feature request - lilypond completion of note octave and string

[David Kastrup]

bb <address@hidden> writes:

> Am 02.08.2013 20:39, schrieb Pierre Perol-Schneider:
>> \version "2.17.23"
>>
>> MyTuning = \stringTuning <g, d g b d'>
>>
>> \new TabStaff {
>>    \key g \major
>>    \set TabStaff.stringTunings = #MyTuning
>>    \tabFullNotation
>>    \stemDown
>>    {
>>      %1
>>      \times 2/3 { <d\5>4^"P" (c)^"H" (<d\5>) }
>>      <d\5>4^"P" (c)
>>      %2
>>      \times 2/3 { gis'4^"H" (f')^"P" (g') }
>>      g8^"H" (a) <d b g>8 g
>>    }
>> }
> Thank's for your response. I do know how to do tabs with lilypond. The
> example is a constructed one to show some problems.
>
> I think there is some misunderstandig about what I really proposed. I
> do NOT want lilypond to do any composition or transcription work
> automatically! I DO WANT lilypond to use it's very profound musical
> knowledge to support a composer or transcriber!
>
> I. e. your (my reworked) example. A banjo is tuned (instandard tuning)
> <g' d g b d'> with a reentrant 5th string not <g, d g b d'>. The 5th
> string usually is not fingered but a drone. If you try the given
> example you will find that there are omitted some notes and the tab is
> crippled because there are fingerings wich needs to use negative
> frets. Lilypond is messaging warnings and tells to calculate a new
> string - but it does not. But it could do!

No, it couldn't.

> Why does'nt it do it, that is my question?

Because it is not possible to play c (namely C3) on a banjo unless you
retune it.

> Banjo or Guitar do have 3 octaves available (Ukulele even less), in
> three octaves each note exists just one time, so lilypond has the
> choice out of only three notes!! To make a choice should be to much
> for it?

You are suggesting that LilyPond should just play a different note than
requested?  In a different octave?  Seriously?

> I believe in many things and in most I believe not and one
> thing I do not believe in is, that is impossible for lilypond to find
> an alternate note out of three in a fitting range.

There _is_ no alternate note in a fitting range.  c is lower than any
available note.

> How? Lilypond finds that there is used an open string wheere there is
> no pull off (- always means lower note) possible. The required lower
> note is a c. Where is the next possible c closest to the d available?
> Answer: 4. string 5th fret one octave higher or equivalent or 2nd
> string 1 fret. So there might be defined a rule: i. e. in doubt use
> the possibility with the lowest fret number. (Others: 10th fret of the
> c-drone, about two octaves higher.) So the only solution is 4. string,
> 5th fret.

That's not a solution.  It's the wrong octave.

> As I pointed out in a feature request the first choice concerning to
> musical statistics is the one closest to the note just before.

The closest note is a c (c below middle c').  c does not exist on a
banjo.

> But I found as a result of the responses that this idea is not liked
> very much,

The idea of just playing a different note than requested?  Yes, it would
appear that this idea is not liked very much.  At least not by me.

> so at best let's stop this meaning- and senseless discussion.

I don't consider it a discussion since that would require me getting
your point.

-- 
David Kastrup