Re: uniform-stretching

[Trevor Bača]

On 12/5/06, V!ctor address@hidden <address@hidden> wrote:
> hello Orm, Trevor...
>
> Sorry I have not been helping here. I had not checked my lilypond mail
> for weeks! Thankfully we have Trevor, the sponsor of proportional
> notation!
> So, coincidentally, I just finished a piece for flute where I do
> *exactly* the same thing Orm is trying to do: align a lilypond score
> with some other graph above. My alignment came out almost perfect. There
> are a couple of places where the score is just slightly off and I'm not
> sure why this is, but you can barely notice it.
>
>
> Trevor has pointed out the main elements to adjust for strict
> proportional notation:
>
> 1. \set proportionalNotationDuration = #(ly:make-moment n d) (in the
> Score)
> 2. \override SpacingSpanner #'uniform-stretching = ##t (in the Score)
> 3. \override PaperColumn #'used = ##t (in the Score)
> 4. \remove Separating_line_group_engraver (in the Staff)
> 5. \override SpacingSpanner #'strict-note-spacing = ##t (in the Score)
>
> Out of these I only used no. 1,2 and 5.
> Staffs in my score do have Separating_line_group_engraver, and I didn't
> know about PaperColumn.
>
> So are these two strictly necessary? In what cases would they be? Could
> removing Separating_line_group_engraver have unwanted secondary
> consequences... this creates the StaffSpacing grob correct?
>
> Not sure I understand what PaperColumn #'used does...

Hi Víctor,

Great to see you on the list again.

Han-Wen put together a great explanation of PaperColumn this morning
that's probably further down in your inbox, but I'll paste it in here
anyway:

%%% PaperColumn explanation %%%

Normally, space between 2 columns (two left-edges of
notes/rests) is thought to be composed of

(FIXED + STRETCHABLE)

space, where FIXED is the size of the symbol, and STRETCHABLE
determines how things scale up for shorter or longer lines.

uniform-stretching

sets FIXED to 0.0, and refrains from optical tweaks to STRETCHABLE

un#'used columns are normally thrown out of the spacing problem,
because they distort things.  In this case, they should stay.

%%% END %%%

So prior to looking at Orm's piece I was thinking exactly what you're
thinking, ie, that PaperColumn doesn't matter because in the last
piece I put together I left PaperColumn out completely and got
absolutely fabulous proportional results anyway. (The piece for Carin
that I mailed to you earlier.)

BUT, if you check the flurry of mail I sent out yesterday, there is in
fact a situation under which we need #'used PaperColumns: if there are
skips at the beginning of a measure.

Try it; run a proportional testfile with a measure-initial skip and
watch how the spacing collapses. The solution is to \override
PaperColumn #'used = ##t. (And so I bet that if you check over your
flute piece you'll find that you have no measure-initial skips and can
therefore get away without the PaperColumn stuff.)


> I will try applying numbers 3 and 4 to my score and see what happens.
> (and will be checking my email more often :|  )

As far as \remove Separating_line_group_engraver goes ... it
definitely moves stuff around in some of Orm's music that I was
looking at yesterday, but I haven't yet had time to sit down and work
out a minimal set of permutations to show the exact effect. That
sounds like a good task to put on the list for the proportional HOWTO.


> Trevor, I'm happy to help put together a tutorial on proportional
> notation and temporal spacing issues in general.

OK, good. I want the thing to wind up as one, solid, well laid-out PDF
with the notation and Lily input side-by-side that we can then post on
the Lily wiki (or just email out to whoever's interested). I'm almost
at stopping point before the next piece so maybe we can Skype and draw
up a table of contents to divide up the work.

(Is the flute piece done? Mail me a copy if it's gone to the printers.
And if you have a nice snippet at 400 x 600 or less, send it over and
I'd love to post it on the blog.)


--
Trevor Bača
address@hidden