[EMAIL PROTECTED] writes:
> Currently, Lily spaces notes by starting with a basic distance,
> arithmetic_multiplier, which it applies to the minimum duration note
> of the bar. Then she adds a logarithmic increment, scaled from
> arithmetic_basicspace, for longer notes. (Then, columns are aligned
> and justified.) Fundamentally, this matches visual spacing to musical
> weight and works well.
Almost correct, you interchanged the meaning of basicspace and multiplier.
> A lot of the time in music, I see a section basically in melodic
> notes that occasionally has a rapid ornamental run (scale). So, there
> will be a section in 1/4 notes, then a brief passage in 1/32nds, then
> a return to long notes. Currently, Lily gives the same horizontal
> space to the 1/32nd notes in their bar (even if set in small size as
> is commonly done for cadenzii) as she gives to 1/4 notes in bars
> where 1/4 note is the minimum duration. The resulting visual weight
> method used, arithmetic_multiplier could be applied referred to the
> same duration throughout a piece. Of course, the current method
actually, the minimum for spacing comparisons is 1/8, not 1/4 (so the
spacing algorithm always works as if there are 1/8 notes or shorter in
a measure). But I am not so sure that un-hardcoding this will be
satisfying in the end.
On the other hand, having arbitrary constants in the lily code is bad
engineering, so I made this runtime settable for .73. Have a look at
this
http://www.cs.uu/~hanwen/lilypond/scarlatti-32.ps.gz
for a base of 32nds (so every measure is assumed to contain 32nd or
shorter notes). Do you think it looks better than the original ?
--
Han-Wen Nienhuys | [EMAIL PROTECTED] | http://www.cs.uu/~hanwen/
