Re: Changing autobeaming for 4/4 time

[Trevor Daniels]

Carl

Yes, that looks good.  Paraphrasing the beam settings, they say:

"break beam at beats, but break 8th-note beams at 1/2", exactly what 
Ross wants.

So these rules are fine, but they are not correctly implemented when 
the final two beats have a 8 8 8 16 16 pattern.  This is the bug 
that has always existed, and can't be fixed by beaming rules using 
the present time-step by time-step approach.  It will need either a 
longer look-ahead or retrospective modification of beams.  I've no 
idea whether either of these are feasible.  That's what I meant when 
I said earlier that this is a more fundamental problem.

Trevor

----- Original Message ----- 
From: "Carl Sorensen" <address@hidden>
To: "Trevor Daniels" <address@hidden>; "Neil Puttock" 
<address@hidden>; "Patrick McCarty" <address@hidden>
Cc: "lilypond-devel" <address@hidden>
Sent: Monday, August 10, 2009 11:25 PM
Subject: Re: Changing autobeaming for 4/4 time


OK, I've found a setting that works *pretty well*, with one bad 
case:

\relative c'' {
 \overrideBeamSettings #'Score #'(4 . 4) #'end
#'((* . (1 1 1 1))
  ((1 . 8) . (4 4)))
 a8 a a a a a a a |  % OK
\break
 a16 a a8 a a
 a a16 a a8 a |  % OK now, was wrong, should be ...
 a16[ a a8] a a
 a[ a16 a] a8 a |
\break
 a8 a a16 a a8
 a a a a16 a | % wrong, should be ...
 a8 a a16[ a a8]
 a a a[ a16 a] |
\break
 a8 a16 a a8 a16 a
 a8 a16 a a8 a16 a | % OK now, was wrong, should be ...
 a8[ a16 a] a8 a16 a
 a8[ a16 a] a8 a16 a |
\break
 a32 a a16 a a a a32 a a16 a
 a a a32 a a16 a a a a32 a | % OK now, was wrong, should be ...
 a32[ a a16 a a] a[ a32 a a16 a]
 a[ a a32 a a16] a a a a32 a
}


I think I can fix this by tweaking with the START rule; it appears 
to me
that what fails in measure 4 above is that the beam should be broken 
at the
final beat because the beaming in the final beat contains a 1/16 
beam, so
the 1/8 rule shouldn't apply.  But we don't check for that at the 
start of
the final beat; we only get that check at the end of the final beat.

I'll do some investigation and see if I can fix that.

But for now, I think this beaming rule for 4/4 time is the best I've 
seen
yet.  What do you think?

Thanks,

Carl