On Saturday, October 01, 2011, Yves Guillemot wrote:

> I don't think I ever implemented any sort of "compression".
> Are you talking of displaying the repeated segments in notation ?
> I see it rather as an expansion than as a compression.

What I'm referring to as "compression" is something new that has started 
happening since your recent work, Yves.

In anacrusis-reference-2.rg the time stream runs to the end of bar 18.  If you 
play this (random notes, don't listen) you're going to hear the initial 
segment repeat itself, and then finish off with the second segment.

If you export this to LilyPond, it comes out almost as one would want to see 
this written down in practice.  The notation ends at bar 9, and the bars are 
numbered correctly.  Making a little tweak to the \skip again gets it even 
closer to coming out right.  There is still a problem after the final double 
bar line, but the last bar has a repeat in the middle of it, and no gap 
between that repeat and the final two beats in bar 9, so the mission is mostly 
accomplished.

--- anacrusis-reference-2.ly.orig       2011-10-02 11:47:05.000000000 -0400
+++ anacrusis-reference-2.ly    2011-10-02 11:49:49.000000000 -0400
@@ -58,7 +58,7 @@
                 \override Voice.TextScript #'padding = #2.0
                 \override MultiMeasureRest #'expand-limit = 1
 
-                \skip 2*35                 \skip 64*32
+                \skip 2*34                % \skip 64*32
 
                 \clef "treble"
                 \key c \major

In the old days (I'm 95% sure, but didn't compile an old version of Rosegarden 
to confirm), LilyPond export would have expanded that repeated segment and 
written all of this out to bar 18, in addition to having the same sort of 
\skip problem I had to correct manually in this example.  This "compression" 
is something new.  You could theoretically have a composition that sounded all 
the repeats, and yet printed out with correct bar numbers and no redundant 
notes put on paper.

Of course the bar numbers would not match Rosegarden bar numbers.  To do that, 
you'd still have to go back to the old way of writing out an example like 
this, old-anacrusis-example.rg

That example, interestingly enough, now prints out correctly without 
modification!  So at least if we did go back to the traditional way of having 
to have one composition to print and one composition to listen to, at least 
the printed composition containing anacrusis might actually come out correctly 
on paper now.

I hope this explanation and these examples make it more clear what I'm talking 
about.

> So it should be possible to define the segno as a local event which resides
> in a linked segment without being duplicated in the others.

I'll leave that sort of thing to the two of you to figure out, but I can 
vaguely see the corner of it now, and it sounds like it might be possible.
-- 
D. Michael McIntyre

Attachment: old-anacrusis-example.rg
Description: audio/rosegarden-composition

Attachment: anacrusis-reference-2.rg
Description: audio/rosegarden-composition

------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
http://p.sf.net/sfu/splunk-d2dcopy2
_______________________________________________
Rosegarden-devel mailing list
[email protected] - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-devel

Reply via email to