Re: polytempi in Lilypond - is it possible?

[Alexander Kobel]

On 2016-10-26 14:11, Bálint Laczkó wrote:
> Hi Alexander,
>
> Thanks for the fast reply!
>
> "Essentially, you will have one "master staff" that stays true to time,
> everything else is scaled to fit." -- so I can change the spacing(?) of
> the staves individually in reference to an original spacing? So if
> staff1 is in default spacing, then can staff2 could be in spacing 1.2
> (so the overall horizontal length of staff2 would be the length of
> staff1 times 1.2?)

With the default method, you change the note /durations/; the appearance 
(e.g., quarters will remain quarters, no matter whether their execution 
is sped up) stays the same, but their spacing is appropriately adjusted. 
Each staff itself will be as long as necessary; if you want a shorter 
staff1 in your example, you can issue a \stopStaff for it after the 
music is over, add a spacer rest of length music*0.2, and call 
\startStaff again. If you don't do it manually, all staves are simply 
filled as long as there is music for them; you certainly know that 
concept from unfinished pieces.

> Would this also mean different barline positions and
> everything?

Not sure about "everything", but for barlines: no. That's because, as I 
said, a bar length in Lilypond is global, as in "per entire score". I 
know that you can technically "fake" bar lines per staff (by 
synchronizing all voices of a score at one beat and hiding the bar line 
in all but one staff). YMMV, and appropriately styled breathe marks 
could be a sensible alternative.
E.g., in the small example you posted, I'd have no proper idea what 
"measure number 5" means... - each voice has the same number of bars in 
that system, but that's more or less accidental. For Lilypond, 
"beginning of measure 8" (or rather something like "73.75 quarters from 
t=0, the beginning of the score") is an /absolute/ point in time, and 
the same for all parts of one score.

I see a very good reason for this approach, and this is why I don't 
consider my question a pure practical one rather than an engraving 
question (albeit not entirely Lily-technical):

> [...]
> As for the practical side you brought up: of course it depends on the
> musical material, so yes in some cases it would need multiple
> conductors, in others just a bit more practise -- but I suggest that we
> focus on the LilyPond-side of the issue for now. :)

Maybe replace "conductor" by "reference clock". My opionion is that if 
there is no such "reference clock", there are to some extent different 
scores in the piece, each for one group. Hence you might want several 
scores, each with the necessary number of staves, and the case is solved 
with strictly proportional notation AFAICS. But you are right, Lilypond 
does not offer a lot of support for intertwined scores because - ahem - 
it's not meant for several scores being part of one score.

Otherwise (and I think that's closer to your situation), there is one 
score and one common time given by one "reference clock". Even if there 
are "timeservers" (assistant conductors), there is that one universal 
time server who insists he's always right. The bar lines correspond to 
(major) ticks of that clock. Trivial cases are the 3/4 vs. 9/8 as in 
J.S.Bach's "Wohl mir, dass ich Jesum habe" (just to avoid writing out 
the 3/2 tuplets); simple cases are e.g. a 4+4 vs 3+3+3 timing (where I'd 
have no idea as a conductor except to ignore one voice, and/or have 
common beats on the first note of each bar); more difficult ones are 
completely polyrhythmic. But they have in common that they more or less 
agree on some ticks of the clock.

If the rhythms are entirely unrelated (or the ratios are too complex for 
a sane human to understand and/or conduct), I propose not to use bars at 
all. Then you are in a situation where I find it most appropriate to 
have a cadenza section with notes proportionally spaced to reflect the 
different speeds, and bar lines only at synchronization barriers, where 
they are plausible between the cadenza parts.


HTH,
Alexander


> 2016-10-26 10:15 GMT+02:00 Alexander Kobel <address@hidden
> <mailto:address@hidden>>:
>
>     On 2016-10-25 21:48, Bálint Laczkó wrote:
>
>         Hey Everybody,
>
>         I would like to engrave a polytempical musical material in
>         LilyPond. [...]
>         Now ideally I would like to see this engraved in a proportional
>         way, so
>         that the lower staff has a wider spacing, and so the notes are
>         distributed in the two staves illustrating their exact relation to a
>         mutual time-scale (so in short: sort of like on the picture above).
>         Is there any way I can do this?
>
>
>     Hi Bálint,
>
>     have a look at section 1.2.3, Polymetric notation in the Notation
>     Reference (NR):
>
>     
> http://lilypond.org/doc/v2.18/Documentation/notation/displaying-rhythms#polymetric-notation
>     
> <http://lilypond.org/doc/v2.18/Documentation/notation/displaying-rhythms#polymetric-notation>
>
>     There is one difference to your example: in the usual approach,
>     there is only a per-score tempo; the note durations (not appearance)
>     are scaled to fit this reference metrum. Essentially, you will have
>     one "master staff" that stays true to time, everything else is
>     scaled to fit.
>
>     If you want even more control over horizontal positioning of the
>     notes, also consider NR sec. 4.5.5:
>
>     http://lilypond.org/doc/v2.18/Documentation/notation/proportional-notation
>     <http://lilypond.org/doc/v2.18/Documentation/notation/proportional-notation>
>     Especially the uniform stretching is highly useful in many
>     contemporary works; your case ask for strict proportional spacing.
>
>     I'm not 100% sure how to fit unique bar line positions per staff
>     into that picture, though; and I'm not sure whether I like it. To
>     me, measure numbers and bar positions are a global thing, so I'd
>     rather have a rest with fermata at the end and a common final
>     barline; in between, I'm unsure.
>     Is this piece supposed to be conducted by several conductors? Then,
>     different bar lines make sense, but maybe they could be faked with
>     breathe marks or something similar; if there is only one conductor,
>     I feel that the bar lines should be with him, for the entire score.
>
>
>     HTH,
>     Alexander