Re: Improbable offset for stencil

Yotam Medini יותם מדיני <[email protected]> Sat, 20 Jul 2024 00:31:35 +0300

Newsgroups	gmane.comp.gnu.lilypond.bugs
Message-ID	<CAAE-6rb64zDqooOYRP=gcO8KvSMntj1RNGfx+1OthHjZ6c9JYA@mail.gmail.com>

Hello Werner,

Thanks for your reply.
Does the (bug?) report I sent get a ticket in
https://gitlab.com/lilypond/lilypond/-/issues ?

I ran your suggested - "branch regression test":
  [@master, test-baseline, bytecode, @MyBranch, bytecode, check]
Indeed lots of regtests failures.

Another fix attempt is attached.
If I knew how to debug Scheme code, I might have come
with a finer solution.
But with the (C++) suggested fix, the 'make check' produced
much smaller (attached) build/out/test-results/index.html.
To my surprise, also (attached) build/input/regression/midi/out/midi.diff
Could it be that I missed something in the "branch regression test" ?

regards -- yotam

On Mon, Jul 15, 2024 at 10:10 AM Werner LEMBERG <[email protected]> wrote:

>
> > programming error: Improbable offset for stencil: -inf staff space
> > [...]
> > Also attached is a log and a suggested fix.
>
> Thanks for the report.  Your suggested fix essentially undoes commit
> 6ac913499ad6 (from 2013) and unfortunately it makes many regtests
> results worse.
>
> You might run `make bytecode` followed by `make test-baseline` on
> 'master', then applying your change, then running `make bytecode`
> followed by `make check` to create `out/test-restuls/index.html`,
> which you can visit in your browser and which shows you all regtest
> differences.
>
>
>     Werner
>

index.html (text/plain, 6.9 KB)

   click to filter rows by type: [1]ly / [2]midi / [3]log / [4]gittxt /
   [5]reset to all
     __________________________________________________________________

   4282 unchanged

   distance before after
   3.741657 [6]input/regression/scheme-engraver.log
@@ -13,9 +13,9 @@
 #<Mom 0>        : (process-acknowledged)
 #<Mom 0>        : (stop-translation-timestep)
 #<Mom 1/8>      : (start-translation-timestep)
-#<Mom 1/8>      : detected breathing event in #:once listener: #<Prob: Stream_e
vent C++: Stream_event((music-cause . #<Prob: Music C++: Music((length . #<Mom 0
>) (origin . #<location /home/yotam/pub/lilypond/input/regression/scheme-engrave
r.ly:67:3>))((display-methods #<procedure method (a)>) (name . BreathingEvent) (
types event breathing-event) (midi-length . #<procedure breathe::midi-length (a
b)>)) >
+#<Mom 1/8>      : detected breathing event in #:once listener: #<Prob: Stream_e
vent C++: Stream_event((music-cause . #<Prob: Music C++: Music((length . #<Mom 0
>) (origin . #<location /home/yotam/pub/lilypond/input/regression/scheme-engrave
r.ly:67:3>))((display-methods #<procedure method (event)>) (name . BreathingEven
t) (types event breathing-event) (midi-length . #<procedure breathe::midi-length
 (len context)>)) >
 ))((class breathing-event music-event StreamEvent) (length . #<Mom 0>) (origin
. #<location /home/yotam/pub/lilypond/input/regression/scheme-engraver.ly:67:3>)
) >
-#<Mom 1/8>      : detected this rest event: #<Prob: Stream_event C++: Stream_ev
ent((music-cause . #<Prob: Music C++: Music((length . #<Mom 1/8>) (duration . #<
Duration 8 >) (origin . #<location /home/yotam/pub/lilypond/input/regression/sch
eme-engraver.ly:68:3>))((display-methods #<procedure method (a)>) (name . RestEv
ent) (iterator-ctor . #<procedure ly:rhythmic-music-iterator::constructor ()>) (
types event rhythmic-event general-rest-event rest-event)) >
+#<Mom 1/8>      : detected this rest event: #<Prob: Stream_event C++: Stream_ev
ent((music-cause . #<Prob: Music C++: Music((length . #<Mom 1/8>) (duration . #<
Duration 8 >) (origin . #<location /home/yotam/pub/lilypond/input/regression/sch
eme-engraver.ly:68:3>))((display-methods #<procedure method (rest)>) (name . Res
tEvent) (iterator-ctor . #<procedure ly:rhythmic-music-iterator::constructor ()>
) (types event rhythmic-event general-rest-event rest-event)) >
 ) (length . #<Mom 1/8>))((class rest-event general-rest-event rhythmic-event mu
sic-event StreamEvent) (length . #<Mom 1/8>) (duration . #<Duration 8 >) (origin
 . #<location /home/yotam/pub/lilypond/input/regression/scheme-engraver.ly:68:3>
)) >

 #<Mom 1/8>      : created this grob: #<Grob TextScript >
@@ -28,7 +28,7 @@
 #<Mom 1/8>      : (process-acknowledged)
 #<Mom 1/8>      : (stop-translation-timestep)
 #<Mom 1/4>      : (start-translation-timestep)
-#<Mom 1/4>      : detected breathing event in #:once listener: #<Prob: Stream_e
vent C++: Stream_event((music-cause . #<Prob: Music C++: Music((length . #<Mom 0
>) (origin . #<location /home/yotam/pub/lilypond/input/regression/scheme-engrave
r.ly:70:3>))((display-methods #<procedure method (a)>) (name . BreathingEvent) (
types event breathing-event) (midi-length . #<procedure breathe::midi-length (a
b)>)) >
+#<Mom 1/4>      : detected breathing event in #:once listener: #<Prob: Stream_e
vent C++: Stream_event((music-cause . #<Prob: Music C++: Music((length . #<Mom 0
>) (origin . #<location /home/yotam/pub/lilypond/input/regression/scheme-engrave
r.ly:70:3>))((display-methods #<procedure method (event)>) (name . BreathingEven
t) (types event breathing-event) (midi-length . #<procedure breathe::midi-length
 (len context)>)) >
 ))((class breathing-event music-event StreamEvent) (length . #<Mom 0>) (origin
. #<location /home/yotam/pub/lilypond/input/regression/scheme-engraver.ly:70:3>)
) >
 #<Mom 1/4>      : (pre-process-music)
 #<Mom 1/4>      : (process-music)
@@ -41,15 +41,14 @@
 #<Mom 1/4>      : (process-acknowledged)
 #<Mom 1/4>      : (stop-translation-timestep)
 #<Mom 3/8>      : (start-translation-timestep)
-#<Mom 3/8>      : detected breathing event in #:once listener: #<Prob: Stream_e
vent C++: Stream_event((music-cause . #<Prob: Music C++: Music((length . #<Mom 0
>) (origin . #<location /home/yotam/pub/lilypond/input/regression/scheme-engrave
r.ly:73:3>))((display-methods #<procedure method (a)>) (name . BreathingEvent) (
types event breathing-event) (midi-length . #<procedure breathe::midi-length (a
b)>)) >
+#<Mom 3/8>      : detected breathing event in #:once listener: #<Prob: Stream_e
vent C++: Stream_event((music-cause . #<Prob: Music C++: Music((length . #<Mom 0
>) (origin . #<location /home/yotam/pub/lilypond/input/regression/scheme-engrave
r.ly:73:3>))((display-methods #<procedure method (event)>) (name . BreathingEven
t) (types event breathing-event) (midi-length . #<procedure breathe::midi-length
 (len context)>)) >
 ))((class breathing-event music-event StreamEvent) (length . #<Mom 0>) (origin
. #<location /home/yotam/pub/lilypond/input/regression/scheme-engraver.ly:73:3>)
) >
 #<Mom 3/8>      : (pre-process-music)
 #<Mom 3/8>      : (process-music)
 #<Mom 3/8>      : (process-acknowledged)
 #<Mom 3/8>      : (process-acknowledged)
 #<Mom 3/8>      : (process-acknowledged)
-#<Mom 3/8>      : (process-acknowledged)
-#<Mom 3/8>      : (stop-tran
+#<Mom 3/8>      : (process-ack
 Preprocessing graphical objects...
 Calculating line breaks...
 Drawing systems...
@@ -58,5 +57,6 @@
 Layout output to `scheme-engraver-1.eps'...
 Writing scheme-engraver-systems.texi...
 Writing scheme-engraver-systems.tex...
-Writing scheme-engraver-systems.count...slation-timestep)
+Writing scheme-engraver-systems.count...nowledged)
+#<Mom 3/8>      : (stop-translation-timestep)
 #<Mom 3/8>      : (finalize)
[7]

   input/regression/scheme-engraver.log
   0.300030
    BRANCH: master
      HEAD: 5828262395bedbbe72cabae29b951160d2cd4d70
MERGE_BASE: 5828262395bedbbe72cabae29b951160d2cd4d70

    HISTORY
    =======
      HASH: 5828262395bedbbe72cabae29b951160d2cd4d70
   SUBJECT: 61f-Lyrics-GracedNotes.xml: Remove erratic ties

Fri Jul 19 06:19:39 PM IDT 2024

[8]

   input/regression/tree.gittxt
    BRANCH: yotam/translate-minus-infty-to-zero
      HEAD: 773afabd1752786be6073b755057b73798a25bc9
MERGE_BASE: 5828262395bedbbe72cabae29b951160d2cd4d70

    HISTORY
    =======
      HASH: 773afabd1752786be6073b755057b73798a25bc9
   SUBJECT: Stencil::translate modify -inf to zero

      HASH: 5828262395bedbbe72cabae29b951160d2cd4d70
   SUBJECT: 61f-Lyrics-GracedNotes.xml: Remove erratic ties

Fri Jul 19 06:27:36 PM IDT 2024

[9]

   input/regression/tree.gittxt

References

   1. file:///tmp/tmp4lX9LZ.html
   2. file:///tmp/tmp4lX9LZ.html
   3. file:///tmp/tmp4lX9LZ.html
   4. file:///tmp/tmp4lX9LZ.html
   5. file:///tmp/tmp4lX9LZ.html
   6. file:///tmp/input/regression/out-test-baseline/scheme-engraver.log
   7. file:///tmp/input/regression/out-test/scheme-engraver.log
   8. file:///tmp/input/regression/out-test-baseline/tree.gittxt
   9. file:///tmp/input/regression/out-test/tree.gittxt

58282623.diff (text/x-patch, 627 B)

commit 773afabd1752786be6073b755057b73798a25bc9
Author: Yotam Medini <[email protected]>
Date:   Fri Jul 19 18:12:35 2024 +0300

    Stencil::translate modify -inf to zero

diff --git a/lily/stencil.cc b/lily/stencil.cc
index 156b48bf14..1c7bfa7c1e 100644
--- a/lily/stencil.cc
+++ b/lily/stencil.cc
@@ -150,7 +150,11 @@ Stencil::translate (Offset o)
 {
   for (const auto a : {X_AXIS, Y_AXIS})
     {
-      if (std::isinf (o[a])
+      if ((o[a] < 0.0) && (std::isinf (o[a])))
+        {
+          o[a] = 0.0;
+        }
+      else if (std::isinf (o[a])
           || std::isnan (o[a])
 
           // ugh, hardcoded.

midi.diff (text/x-patch, 80.1 KB)

--- /home/yotam/pub/lilypond/input/regression/midi/articulate-repeat.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/articulate-repeat-midi.ly	2024-07-19 18:27:39.958361727 +0300
@@ -1,23 +1,72 @@
-\version "2.25.0"
+% Lily was here -- automatically converted by midi2ly.py from out-test/articulate-repeat.midi
+\version "2.14.0"
 
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/articulate-repeat.header
 \header {
-  texidoc = "@code{\\articulate} correctly handles repeats
+texidoc="@code{\articulate} correctly handles repeats
 with alternatives.  They are played in the same order as their
 visual rendition suggests."
+options=""
+}
+% end
+
+trackAchannelA = {
+  
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 60 
+  \skip 1*4 
+}
+
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
+
+
+trackBchannelA = {
+  
+  \set Staff.instrumentName = "\new:"
+  \skip 1*4 
+}
+
+trackBchannelB = \relative c {
+  ais'8*7 r8 
+  | % 2
+  b8*7 r8 
+  | % 3
+  ais8*7 r8 
+  | % 4
+  c8*7 
 }
 
-\include "articulate.ly"
+trackB = <<
+
+  \clef bass
+  
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+>>
+
 
 \score {
-  \articulate {
-    \repeat volta 2 {
-      bes1
-      \alternative {
-        \volta 1 b!
-        \volta 2 bis
-      }
-    }
-  }
-  \midi { }
-  \layout { }
+  <<
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
+  >>
+  \layout {}
+  \midi {}
 }
--- /home/yotam/pub/lilypond/input/regression/midi/articulate-trill.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/articulate-trill-midi.ly	2024-07-19 18:27:40.314359632 +0300
@@ -1,17 +1,72 @@
-\version "2.23.9"
+% Lily was here -- automatically converted by midi2ly.py from out-test/articulate-trill.midi
+\version "2.14.0"
 
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/articulate-trill.header
 \header {
-  texidoc = "@code{\\articulate} correctly handles trill spans ending on multi-measure rests."
+texidoc=""
+options=""
 }
+% end
 
-\include "articulate.ly"
-\score {
-\unfoldRepeats \articulate {
-   b1 \startTrillSpan |
-   R1 \stopTrillSpan |
-   b2 r |
-   b1
+trackAchannelA = {
+  
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 60 
+  \skip 1*4 
 }
-\midi {}
+
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
+
+
+trackBchannelA = {
+  
+  \set Staff.instrumentName = "\new:"
+  \skip 1*4 
 }
 
+trackBchannelB = \relative c {
+  c'64 b c b c b c b c b c b c b c b c b c b c b c b c b c b c 
+  b c b c b c b c b c b c b c b c b c b c b c b c b c b c b c b 
+  c b c b 
+  | % 2
+  r1 
+  | % 3
+  b16*7 r16*9 
+  | % 4
+  b8*7 
+}
+
+trackB = <<
+
+  \clef bass
+  
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+>>
+
+
+\score {
+  <<
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
+  >>
+  \layout {}
+  \midi {}
+}
--- /home/yotam/pub/lilypond/input/regression/midi/crescendo-gap-compatible-target.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/crescendo-gap-compatible-target-midi.ly	2024-07-19 18:27:40.646357676 +0300
@@ -1,18 +1,64 @@
-\version "2.19.45"
+% Lily was here -- automatically converted by midi2ly.py from out-test/crescendo-gap-compatible-target.midi
+\version "2.14.0"
 
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/crescendo-gap-compatible-target.header
 \header {
-  texidoc="When there is a gap between the end of a crescendo and a
-  subsequent explicit dynamic, the dynamic performer uses the explicit
-  dynamic as the target of the crescendo."
-
-  %% Note: Choosing this behavior simplified the implementation.  In
-  %% the developer's opinion, it is difficult to argue that choosing a
-  %% target dynamic that under- or overshoots the explicit dynamic is
-  %% more correct.
+texidoc=""
+options=""
+}
+% end
+
+trackAchannelA = {
+  
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 60 
+  \skip 2. 
+}
+
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
 
+
+trackBchannelA = {
+  
+  \set Staff.instrumentName = "\new:"
+  \skip 2. 
+}
+
+trackBchannelB = \relative c {
+  c4 c c 
 }
 
+trackB = <<
+
+  \clef bass
+  
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+>>
+
+
 \score {
-   { c\mf\< c\! c\f }
-   \midi {}
+  <<
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
+  >>
+  \layout {}
+  \midi {}
 }
--- /home/yotam/pub/lilypond/input/regression/midi/crescendo-return-crescendo.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/crescendo-return-crescendo-midi.ly	2024-07-19 18:27:40.982355697 +0300
@@ -1,12 +1,70 @@
-\version "2.19.44"
+% Lily was here -- automatically converted by midi2ly.py from out-test/crescendo-return-crescendo.midi
+\version "2.14.0"
 
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/crescendo-return-crescendo.header
 \header {
-  texidoc="The dynamic performer chooses a reasonable peak dynamic and
-  returns to the original dynamic.  The latter, extreme crescendo does
-  not affect the former."
+texidoc=""
+options=""
+}
+% end
+
+trackAchannelA = {
+  
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 60 
+  \skip 1 
+  | % 2
+  
 }
 
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
+
+
+trackBchannelA = {
+  
+  \set Staff.instrumentName = "\new:"
+  \skip 1 
+  | % 2
+  
+}
+
+trackBchannelB = \relative c {
+  c4 c c c 
+  | % 2
+  
+}
+
+trackB = <<
+
+  \clef bass
+  
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+>>
+
+
 \score {
-   { c\mf\< c\> c\< c\sf }
-   \midi {}
+  <<
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
+  >>
+  \layout {}
+  \midi {}
 }
--- /home/yotam/pub/lilypond/input/regression/midi/crescendo-return-louder-target.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/crescendo-return-louder-target-midi.ly	2024-07-19 18:27:41.314353742 +0300
@@ -1,11 +1,64 @@
-\version "2.19.45"
+% Lily was here -- automatically converted by midi2ly.py from out-test/crescendo-return-louder-target.midi
+\version "2.14.0"
 
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/crescendo-return-louder-target.header
 \header {
-  texidoc="The dynamic performer chooses a reasonable peak dynamic and
-  returns to the specified dynamic."
+texidoc=""
+options=""
+}
+% end
+
+trackAchannelA = {
+  
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 60 
+  \skip 2. 
 }
 
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
+
+
+trackBchannelA = {
+  
+  \set Staff.instrumentName = "\new:"
+  \skip 2. 
+}
+
+trackBchannelB = \relative c {
+  c4 c c 
+}
+
+trackB = <<
+
+  \clef bass
+  
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+>>
+
+
 \score {
-   { c\mf\< c\> c\f }
-   \midi {}
+  <<
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
+  >>
+  \layout {}
+  \midi {}
 }
--- /home/yotam/pub/lilypond/input/regression/midi/crescendo-return-softer-target.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/crescendo-return-softer-target-midi.ly	2024-07-19 18:27:41.646351787 +0300
@@ -1,11 +1,64 @@
-\version "2.19.45"
+% Lily was here -- automatically converted by midi2ly.py from out-test/crescendo-return-softer-target.midi
+\version "2.14.0"
 
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/crescendo-return-softer-target.header
 \header {
-  texidoc="The dynamic performer chooses a reasonable peak dynamic and
-  returns to the specified dynamic."
+texidoc=""
+options=""
+}
+% end
+
+trackAchannelA = {
+  
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 60 
+  \skip 2. 
 }
 
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
+
+
+trackBchannelA = {
+  
+  \set Staff.instrumentName = "\new:"
+  \skip 2. 
+}
+
+trackBchannelB = \relative c {
+  c4 c c 
+}
+
+trackB = <<
+
+  \clef bass
+  
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+>>
+
+
 \score {
-   { c\mf\< c\> c\p }
-   \midi {}
+  <<
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
+  >>
+  \layout {}
+  \midi {}
 }
--- /home/yotam/pub/lilypond/input/regression/midi/crescendo-return-unspecified-target.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/crescendo-return-unspecified-target-midi.ly	2024-07-19 18:27:41.982349809 +0300
@@ -1,11 +1,64 @@
-\version "2.19.44"
+% Lily was here -- automatically converted by midi2ly.py from out-test/crescendo-return-unspecified-target.midi
+\version "2.14.0"
 
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/crescendo-return-unspecified-target.header
 \header {
-  texidoc="The dynamic performer chooses a reasonable peak dynamic and
-  returns to the original dynamic."
+texidoc=""
+options=""
+}
+% end
+
+trackAchannelA = {
+  
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 60 
+  \skip 2. 
 }
 
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
+
+
+trackBchannelA = {
+  
+  \set Staff.instrumentName = "\new:"
+  \skip 2. 
+}
+
+trackBchannelB = \relative c {
+  c4 c c 
+}
+
+trackB = <<
+
+  \clef bass
+  
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+>>
+
+
 \score {
-   { c\mf\< c\> c\! }
-   \midi {}
+  <<
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
+  >>
+  \layout {}
+  \midi {}
 }
--- /home/yotam/pub/lilypond/input/regression/midi/crescendo-single-compatible-target.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/crescendo-single-compatible-target-midi.ly	2024-07-19 18:27:42.310347878 +0300
@@ -1,11 +1,64 @@
-\version "2.19.44"
+% Lily was here -- automatically converted by midi2ly.py from out-test/crescendo-single-compatible-target.midi
+\version "2.14.0"
 
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/crescendo-single-compatible-target.header
 \header {
-  texidoc="The velocity of notes during a crescendo is linearly
-  interpolated between the starting and target dynamics."
+texidoc=""
+options=""
+}
+% end
+
+trackAchannelA = {
+  
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 60 
+  \skip 2. 
 }
 
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
+
+
+trackBchannelA = {
+  
+  \set Staff.instrumentName = "\new:"
+  \skip 2. 
+}
+
+trackBchannelB = \relative c {
+  c4 c c 
+}
+
+trackB = <<
+
+  \clef bass
+  
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+>>
+
+
 \score {
-   { c\mf\< c c\f }
-   \midi {}
+  <<
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
+  >>
+  \layout {}
+  \midi {}
 }
--- /home/yotam/pub/lilypond/input/regression/midi/crescendo-single-unspecified-target.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/crescendo-single-unspecified-target-midi.ly	2024-07-19 18:27:42.646345901 +0300
@@ -1,11 +1,64 @@
-\version "2.19.44"
+% Lily was here -- automatically converted by midi2ly.py from out-test/crescendo-single-unspecified-target.midi
+\version "2.14.0"
 
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/crescendo-single-unspecified-target.header
 \header {
-  texidoc="If no explicit dynamic follows a crescendo, the dynamic
-  performer chooses a reasonable target dynamic."
+texidoc=""
+options=""
+}
+% end
+
+trackAchannelA = {
+  
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 60 
+  \skip 2 
 }
 
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
+
+
+trackBchannelA = {
+  
+  \set Staff.instrumentName = "\new:"
+  \skip 2 
+}
+
+trackBchannelB = \relative c {
+  c4 c 
+}
+
+trackB = <<
+
+  \clef bass
+  
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+>>
+
+
 \score {
-   { c\mf\< c\! }
-   \midi {}
+  <<
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
+  >>
+  \layout {}
+  \midi {}
 }
--- /home/yotam/pub/lilypond/input/regression/midi/decrescendo-multiple-compatible-target.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/decrescendo-multiple-compatible-target-midi.ly	2024-07-19 18:27:42.978343945 +0300
@@ -1,12 +1,68 @@
-\version "2.19.45"
+% Lily was here -- automatically converted by midi2ly.py from out-test/decrescendo-multiple-compatible-target.midi
+\version "2.14.0"
 
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/decrescendo-multiple-compatible-target.header
 \header {
-  texidoc="The dynamic performer apportions changes to consecutive
-  decrescendi in proportion to their duration.  In this case, 1/3 of
-  the change occurs over the first decrescendo."
+texidoc=""
+options=""
+}
+% end
+
+trackAchannelA = {
+  
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 60 
+  \skip 1*2 
 }
 
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
+
+
+trackBchannelA = {
+  
+  \set Staff.instrumentName = "\new:"
+  \skip 1*2 
+}
+
+trackBchannelB = \relative c {
+  c4 c c2 
+  | % 2
+  c1 
+  | % 3
+  
+}
+
+trackB = <<
+
+  \clef bass
+  
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+>>
+
+
 \score {
-   { c\f\> c\! c2\> | c1\p }
-   \midi {}
+  <<
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
+  >>
+  \layout {}
+  \midi {}
 }
--- /home/yotam/pub/lilypond/input/regression/midi/decrescendo-single-contrary-target.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/decrescendo-single-contrary-target-midi.ly	2024-07-19 18:27:43.310341992 +0300
@@ -1,15 +1,64 @@
-\version "2.19.45"
+% Lily was here -- automatically converted by midi2ly.py from out-test/decrescendo-single-contrary-target.midi
+\version "2.14.0"
 
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/decrescendo-single-contrary-target.header
 \header {
-  texidoc="When a decrescendo is followed by an explicit dynamic that
-  is louder than the starting dynamic, the dynamic performer chooses a
-  reasonable target dynamic.  The velocity of notes during the
-  decrescendo is linearly interpolated between the starting and target
-  dynamics, with the explicit dynamic taking effect at the last
-  moment."
+texidoc=""
+options=""
+}
+% end
+
+trackAchannelA = {
+  
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 60 
+  \skip 2. 
 }
 
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
+
+
+trackBchannelA = {
+  
+  \set Staff.instrumentName = "\new:"
+  \skip 2. 
+}
+
+trackBchannelB = \relative c {
+  c4 c c 
+}
+
+trackB = <<
+
+  \clef bass
+  
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+>>
+
+
 \score {
-   { c\mf\> c c\f }
-   \midi {}
+  <<
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
+  >>
+  \layout {}
+  \midi {}
 }
--- /home/yotam/pub/lilypond/input/regression/midi/dynamic-initial.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/dynamic-initial-midi.ly	2024-07-19 18:27:43.642340036 +0300
@@ -1,15 +1,64 @@
-\version "2.19.44"
+% Lily was here -- automatically converted by midi2ly.py from out-test/dynamic-initial.midi
+\version "2.14.0"
 
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/dynamic-initial.header
 \header {
-  texidoc="The MIDI performer uses a default initial dynamic"
+texidoc=""
+options=""
 }
+% end
 
-\score {
-   { c\< c\f }
-   \midi {}
+trackAchannelA = {
+  
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 60 
+  \skip 2 
+}
+
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
+
+
+trackBchannelA = {
+  
+  \set Staff.instrumentName = "\new:"
+  \skip 2 
+}
+
+trackBchannelB = \relative c {
+  c4 c 
 }
 
+trackB = <<
+
+  \clef bass
+  
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+>>
+
+
 \score {
-   { c c\< c\f }
-   \midi {}
+  <<
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
+  >>
+  \layout {}
+  \midi {}
 }
--- /home/yotam/pub/lilypond/input/regression/midi/dynamic-voices-sequential.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/dynamic-voices-sequential-midi.ly	2024-07-19 18:27:43.978338059 +0300
@@ -1,15 +1,70 @@
-\version "2.19.47"
+% Lily was here -- automatically converted by midi2ly.py from out-test/dynamic-voices-sequential.midi
+\version "2.14.0"
 
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/dynamic-voices-sequential.header
 \header {
-  texidoc="The MIDI performer operates in Voice context by default,
-  so dynamics in different voices are independent."
+texidoc=""
+options=""
 }
+% end
+
+trackAchannelA = {
+  
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 60 
+  \skip 1 
+  | % 2
+  
+}
+
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
 
-\score {
-  \new Staff {
-    \new Voice = "A" c'2\p
-    \new Voice = "A" c'2 % default dynamic expected
-  }
 
+trackBchannelA = {
+  
+  \set Staff.instrumentName = ":A"
+  \skip 1 
+  | % 2
+  
+}
+
+trackBchannelB = \relative c {
+  c'2 c 
+  | % 2
+  
+}
+
+trackB = <<
+
+  \clef bass
+  
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+>>
+
+
+\score {
+  <<
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
+  >>
+  \layout {}
   \midi {}
 }
--- /home/yotam/pub/lilypond/input/regression/midi/dynamic-voices-simultaneous.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/dynamic-voices-simultaneous-midi.ly	2024-07-19 18:27:44.310336105 +0300
@@ -1,15 +1,61 @@
-\version "2.19.47"
+% Lily was here -- automatically converted by midi2ly.py from out-test/dynamic-voices-simultaneous.midi
+\version "2.14.0"
 
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/dynamic-voices-simultaneous.header
 \header {
-  texidoc="The MIDI performer operates in Voice context by default,
-  so dynamics in different voices are independent."
+texidoc=""
+options=""
+}
+% end
+
+trackAchannelA = {
+  
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 60 
+  \skip 2 
 }
 
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
+
+
+trackBchannelA = {
+  
+  \set Staff.instrumentName = ":A"
+  \skip 2 
+}
+
+trackBchannelB = \relative c {
+  <e' c >2 
+}
+
+trackB = <<
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+>>
+
+
 \score {
-  \new Staff <<
-    \new Voice = "A" e'2\p
-    \new Voice = "A" c'2 % default dynamic expected
+  <<
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
   >>
-
+  \layout {}
   \midi {}
 }
--- /home/yotam/pub/lilypond/input/regression/midi/key-initial.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/key-initial-midi.ly	2024-07-19 18:27:44.810333163 +0300
@@ -1,19 +1,13 @@
-% Lily was here -- automatically converted by ../../../scripts/midi2ly.py from out/initial-key.midi
-\version "2.19.21"
+% Lily was here -- automatically converted by midi2ly.py from out-test/key-initial.midi
+\version "2.14.0"
 
 \layout {
   \context {
     \Voice
-    \remove "Note_heads_engraver"
-    \consists "Completion_heads_engraver"
-    \remove "Rest_engraver"
-    \consists "Completion_rest_engraver"
-  }
-}
-\midi {
-  \context {
-    \Score
-    midiChannelMapping = #'instrument
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
   }
 }
 
@@ -29,16 +23,14 @@ trackAchannelA = {
 
   \key d \minor
     
-  % [SEQUENCE_TRACK_NAME] control track
-  
   % [TEXT_EVENT] creator: 
   
-  % [TEXT_EVENT] GNU LilyPond 2.13.54          
+  % [TEXT_EVENT] LilyPond 2.25.18              
   
   \time 4/4 
   
   \tempo 4 = 60 
-  
+  \skip 1*2 
 }
 
 trackA = <<
@@ -46,32 +38,33 @@ trackA = <<
 >>
 
 
-trackBchannelA = \relative {
+trackBchannelA = {
   
   \set Staff.instrumentName = "trackB:voiceA"
-  d'4 e f g 
-  | % 2
-  a bes cis d
-  | % 3
-  
+  \skip 1*2 
 }
 
-trackB = <<
-  \context Voice = voiceA \trackBchannelA
->>
-
-
-trackCchannelA = {
+trackBchannelB = {
   
   \set Staff.instrumentName = "trackB:"
   
 
   \key d \minor
+  \skip 1*2 
+}
+
+trackBchannelC = \relative c {
+  d'4 e f g 
+  | % 2
+  a bes cis d 
+  | % 3
   
 }
 
-trackC = <<
-  \context Voice = voiceA \trackCchannelA
+trackB = <<
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+  \context Voice = voiceC \trackBchannelC
 >>
 
 
--- /home/yotam/pub/lilypond/input/regression/midi/key-option-all-staves.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/key-option-all-staves-midi.ly	2024-07-19 18:27:45.266330479 +0300
@@ -1,23 +1,84 @@
-\version "2.19.21"
+% Lily was here -- automatically converted by midi2ly.py from out-test/key-option-all-staves.midi
+\version "2.14.0"
 
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/key-option-all-staves.header
 \header {
 texidoc="Midi2ly --key works on all staves, this is G major (--key=1)"
 options="--key=1"
 }
+% end
+
+trackAchannelA = {
+
+
+  \key g \major
+    
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 60 
+  \skip 4 
+}
+
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
+
+
+trackBchannelA = {
+  
+  \set Staff.instrumentName = "treble:one"
+  \skip 4 
+}
+
+trackBchannelB = \relative c {
+  fis''4 
+}
+
+trackB = <<
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+>>
+
+
+trackCchannelA = {
+  
+  \set Staff.instrumentName = "bass:two"
+  \skip 4 
+}
+
+trackCchannelB = \relative c {
+  fis 
+}
+
+trackC = <<
+
+  \clef bass
+  
+  \context Voice = voiceA \trackCchannelA
+  \context Voice = voiceB \trackCchannelB
+>>
+
 
 \score {
   <<
-    \context Staff = "treble" <<
-      \context Voice="one" \relative {
-	fis''
-      }
-    >>
-    \context Staff = "bass" <<
-      \context Voice="two" \relative {
-	\clef bass
-	fis
-      }
-    >>
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
+    \context Staff=trackC \trackA
+    \context Staff=trackC \trackC
   >>
   \layout {}
   \midi {}
--- /home/yotam/pub/lilypond/input/regression/midi/key-option.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/key-option-midi.ly	2024-07-19 18:27:45.766327536 +0300
@@ -1,19 +1,13 @@
-% Lily was here -- automatically converted by ../../../scripts/midi2ly.py from out-www/option-key.midi
-\version "2.19.21"
+% Lily was here -- automatically converted by midi2ly.py from out-test/key-option.midi
+\version "2.14.0"
 
 \layout {
   \context {
     \Voice
-    \remove "Note_heads_engraver"
-    \consists "Completion_heads_engraver"
-    \remove "Rest_engraver"
-    \consists "Completion_rest_engraver"
-  }
-}
-\midi {
-  \context {
-    \Score
-    midiChannelMapping = #'instrument
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
   }
 }
 
@@ -27,16 +21,16 @@ options="--key=-1"
 trackAchannelA = {
 
 
-  % [SEQUENCE_TRACK_NAME] control track
-  
+  \key f \major
+    
   % [TEXT_EVENT] creator: 
   
-  % [TEXT_EVENT] GNU LilyPond 2.13.54          
+  % [TEXT_EVENT] LilyPond 2.25.18              
   
   \time 4/4 
   
   \tempo 4 = 60 
-  
+  \skip 1*2 
 }
 
 trackA = <<
@@ -44,9 +38,13 @@ trackA = <<
 >>
 
 
-trackBchannelA = \relative {
+trackBchannelA = {
   
   \set Staff.instrumentName = "trackB:voiceA"
+  \skip 1*2 
+}
+
+trackBchannelB = \relative c {
   f'4 g a bes 
   | % 2
   c d e f 
@@ -56,18 +54,7 @@ trackBchannelA = \relative {
 
 trackB = <<
   \context Voice = voiceA \trackBchannelA
->>
-
-
-trackCchannelA = {
-  
-  \set Staff.instrumentName = "trackB:"
-  
-
-}
-
-trackC = <<
-  \context Voice = voiceA \trackCchannelA
+  \context Voice = voiceB \trackBchannelB
 >>
 
 
--- /home/yotam/pub/lilypond/input/regression/midi/lyric-combine-switch-voice.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/lyric-combine-switch-voice-midi.ly	2024-07-19 18:27:46.246324712 +0300
@@ -1,29 +1,72 @@
-\version "2.21.6"
+% Lily was here -- automatically converted by midi2ly.py from out-test/lyric-combine-switch-voice.midi
+\version "2.14.0"
 
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/lyric-combine-switch-voice.header
 \header {
-  texidoc = "It is possible to switch the voice that @code{\\lyricsto}
+texidoc="It is possible to switch the voice that @code{\lyricsto}
 follows."
+options=""
 }
+% end
 
-\layout {
-  indent = 0
-  ragged-right = ##t
+trackAchannelA = {
+  
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 60 
+  \skip 2 
 }
 
-music = \fixed c' <<
-  \new Staff <<
-    \new Voice = Va \with \voiceOne { f4 s }
-    \new Voice = Vb \with \voiceTwo { s4 e' }
-  >>
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
 
-  \new Lyrics \lyricsto Va {
-    \set Lyrics.associatedVoice = Vb
-    Hoo hah.
-  }
+
+trackBchannelA = {
+  
+  \set Staff.instrumentName = ":Va"
+  \skip 2 
+}
+
+trackBchannelB = \relative c {
+  f'4 e' 
+}
+
+trackB = <<
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+>>
+
+
+trackCchannelA = \lyricmode {
+  "Hoo" "hah." 
+}
+
+trackC = <<
+  \context Lyrics = voiceA \trackCchannelA
 >>
 
+
 \score {
-  \music
+  <<
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
+    \context Lyrics=trackC \trackC
+  >>
   \layout {}
   \midi {}
 }
--- /home/yotam/pub/lilypond/input/regression/midi/lyric-repeat-count.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/lyric-repeat-count-midi.ly	2024-07-19 18:27:46.606322593 +0300
@@ -1,33 +1,89 @@
-\version "2.23.7"
+% Lily was here -- automatically converted by midi2ly.py from out-test/lyric-repeat-count.midi
+\version "2.14.0"
 
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/lyric-repeat-count.header
 \header {
-  texidoc = "This exercises @code{GregorianTranscriptionStaff} and
-@code{GregorianTranscription} with MIDI output enabled."
+texidoc=""
+options=""
+}
+% end
+
+trackAchannelA = {
+  
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 60 
+  \skip 1*7 
+}
+
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
+
+
+trackBchannelA = {
+  \skip 1*7 
 }
 
-#(ly:set-option 'warning-as-error #t)
+trackBchannelB = \relative c {
+  f'2 g 
+  | % 2
+  a g 
+  | % 3
+  f e 
+  | % 4
+  f1 
+  | % 5
+  g1*2 f1 
+  | % 8
+  
+}
 
-music = \fixed c' {
-  \repeat volta 2 { f2 g }
-  \repeat volta 3 { a g }
-  \repeat volta 4 { f e }
-  \repeat volta 1 f1
-  g\breve f1
-}
-
-words = \lyricmode {
-  \repeat volta 2 { Lo2 -- rem. }
-  \repeat volta 3 { Ip -- sum. }
-  \repeat volta 4 { Do -- lor. }
-  \repeat volta 1 Sit.1
-  A\breve -- met.1
+trackB = <<
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+>>
+
+
+trackCchannelA = \lyricmode {
+  "Lo"2 "rem." 
+  | % 2
+  "Ip" "sum." 
+  | % 3
+  "Do" "lor." 
+  | % 4
+  "Sit."1 
+  | % 5
+  "A"1*2 "met."4 " "2. 
+  | % 8
+  
 }
 
+trackC = <<
+  \context Lyrics = voiceA \trackCchannelA
+>>
+
+
 \score {
   <<
-    \new GregorianTranscriptionStaff \music
-    \new GregorianTranscriptionLyrics \words
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
+    \context Lyrics=trackC \trackC
   >>
-
+  \layout {}
   \midi {}
 }
--- /home/yotam/pub/lilypond/input/regression/midi/lyric-tie.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/lyric-tie-midi.ly	2024-07-19 18:27:46.942320615 +0300
@@ -1,12 +1,82 @@
-\version "2.23.13"
+% Lily was here -- automatically converted by midi2ly.py from out-test/lyric-tie.midi
+\version "2.14.0"
 
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/lyric-tie.header
 \header {
-  texidoc = "In MIDI, lyric ties are rendered as Unicode undertie characters
-(U+203F)."
+texidoc=""
+options=""
 }
+% end
+
+trackAchannelA = {
+  
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 60 
+  \skip 1*2 
+}
+
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
+
+
+trackBchannelA = {
+  
+  \set Staff.instrumentName = ":uniqueContext0"
+  \skip 1*2 
+}
+
+trackBchannelB = \relative c {
+  c'1 
+  | % 2
+  c 
+  | % 3
+  
+}
+
+trackB = <<
+
+  \clef bass
+  
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+>>
+
+
+trackCchannelA = \lyricmode {
+  "a~~~a" 
+  | % 2
+  "a~~~a~~~a"4 " "2. 
+  | % 3
+  
+}
+
+trackC = <<
+  \context Lyrics = voiceA \trackCchannelA
+>>
+
 
 \score {
-  { c'1 1 }
-  \addlyrics { a~a a~a~a }
-  \midi { }
+  <<
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
+    \context Lyrics=trackC \trackC
+  >>
+  \layout {}
+  \midi {}
 }
--- /home/yotam/pub/lilypond/input/regression/midi/lyrics-addlyrics.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/lyrics-addlyrics-midi.ly	2024-07-19 18:27:47.466317532 +0300
@@ -1,19 +1,13 @@
-% Lily was here -- automatically converted by ../../../scripts/midi2ly.py from out/initial-key.midi
-\version "2.19.21"
+% Lily was here -- automatically converted by midi2ly.py from out-test/lyrics-addlyrics.midi
+\version "2.14.0"
 
 \layout {
   \context {
     \Voice
-    \remove "Note_heads_engraver"
-    \consists "Completion_heads_engraver"
-    \remove "Rest_engraver"
-    \consists "Completion_rest_engraver"
-  }
-}
-\midi {
-  \context {
-    \Score
-    midiChannelMapping = #'instrument
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
   }
 }
 
@@ -17,23 +11,79 @@
   }
 }
 
-% included from ./out/initial-key.header
+% included from ./out-test/lyrics-addlyrics.header
 \header {
 texidoc="Lyrics are preserved"
 options=""
 }
 % end
 
+trackAchannelA = {
+
+
+  \key g \major
+    
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 6/8 
+  
+  \tempo 4 = 60 
+  \skip 8*11 
+}
+
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
+
+
+trackBchannelA = {
+  
+  \set Staff.instrumentName = ":uniqueContext0"
+  \skip 8*11 
+}
+
+trackBchannelB = \relative c {
+  d''4 b8 c4 a8 
+  | % 2
+  d4 b8 g4 
+}
+
+trackB = <<
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+>>
+
+
+trackCchannelA = {
+  
+
+  \key g \major
+  \skip 8*11 
+}
+
+trackC = <<
+  \context Voice = voiceA \trackCchannelA
+>>
+
+
+trackDchannelA = \lyricmode {
+  "Girls" "and"8 "boys"4 "come"8 
+  | % 2
+  "out"4 "to"8 "play,"4 
+}
+
+trackD = <<
+  \context Lyrics = voiceA \trackDchannelA
+>>
+
+
 \score {
   <<
-    \relative {
-      \key g \major
-      \time 6/8
-      d''4 b8 c4 a8 | d4 b8 g4
-    }
-    \addlyrics {
-      Girls and \markup \bold boys come | out to play,
-    }
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
+    \context Lyrics=trackD \trackD
   >>
   \layout {}
   \midi {}
--- /home/yotam/pub/lilypond/input/regression/midi/midi-overlapping-notes.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/midi-overlapping-notes-midi.ly	2024-07-19 18:27:47.802315553 +0300
@@ -1,22 +1,66 @@
+% Lily was here -- automatically converted by midi2ly.py from out-test/midi-overlapping-notes.midi
+\version "2.14.0"
+
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/midi-overlapping-notes.header
 \header {
+texidoc=""
+options=""
+}
+% end
+
+trackAchannelA = {
+  
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 60 
+  \skip 1*3 
+}
 
-  texidoc = "Notes with equal pitch overlapping in time on the same
-MIDI channel should produce multiple MIDI note-on events, but no
-isolated note-off events before the last of these notes ends."
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
 
+
+trackBchannelA = {
+  
+  % [INSTRUMENT_NAME] church organ
+  \skip 1*3 
+}
+
+trackBchannelB = \relative c {
+  c'4 c c1. g1 
+  | % 4
+  
 }
 
-\version "2.19.45"
+trackB = <<
+
+  \clef bass
+  
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+>>
+
 
 \score {
-  \new Staff \with { midiInstrument = "church organ" }
-  \new Voice \relative c' <<
-    % This combination of music expressions should not produce audible
-    % pauses between notes in the MIDI output, in particular, before the
-    % 'g1' note in the first expression.
-    { c1 s1 g1 }
-    { s4 c4 }
-    { s2 c2~ c1 s1 }
+  <<
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
   >>
-  \midi { }
+  \layout {}
+  \midi {}
 }
--- /home/yotam/pub/lilypond/input/regression/midi/midi-polymeter.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/midi-polymeter-midi.ly	2024-07-19 18:27:48.146313530 +0300
@@ -1,16 +1,90 @@
-\version "2.23.7"
+% Lily was here -- automatically converted by midi2ly.py from out-test/midi-polymeter.midi
+\version "2.14.0"
 
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/midi-polymeter.header
 \header {
-  texidoc = "@code{\enablePolymeter} can also be used with MIDI output.
-This test should not produce bar check errors."
+texidoc=""
+options=""
+}
+% end
+
+trackAchannelA = {
+  
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 60 
+  \skip 1*3 
+}
+
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
+
+
+trackBchannelA = {
+  \skip 1*3 
+}
+
+trackBchannelB = \relative c {
+  c'2. c4 
+  | % 2
+  c2 c 
+  | % 3
+  c4 c c c 
+  | % 4
+  
 }
 
+trackB = <<
+
+  \clef bass
+  
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+>>
+
+
+trackCchannelA = {
+  \skip 1*3 
+}
+
+trackCchannelB = \relative c {
+  c'1 
+  | % 2
+  d 
+  | % 3
+  f 
+  | % 4
+  
+}
+
+trackC = <<
+  \context Voice = voiceA \trackCchannelA
+  \context Voice = voiceB \trackCchannelB
+>>
+
+
 \score {
-  \midi {
-    \enablePolymeter
-  }
   <<
-    \new Staff { \time 3/4 c'2. | 4 2 | 2 4 | 4 4 4 }
-    \new Staff { \time 4/4 c'1 | d' | f' }
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
+    \context Staff=trackC \trackA
+    \context Staff=trackC \trackC
   >>
+  \layout {}
+  \midi {}
 }
--- /home/yotam/pub/lilypond/input/regression/midi/music-after-fine-warning.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/music-after-fine-warning-midi.ly	2024-07-19 18:27:48.478311576 +0300
@@ -1,13 +1,64 @@
-\version "2.23.12"
+% Lily was here -- automatically converted by midi2ly.py from out-test/music-after-fine-warning.midi
+\version "2.14.0"
 
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/music-after-fine-warning.header
 \header {
-  texidoc="A warning is expected when music follows @code{\\fine}."
+texidoc=""
+options=""
+}
+% end
+
+trackAchannelA = {
+  
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 60 
+  \skip 2 
+}
+
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
+
+
+trackBchannelA = {
+  
+  \set Staff.instrumentName = "\new:"
+  \skip 2 
+}
+
+trackBchannelB = \relative c {
+  c4 d 
 }
 
-#(ly:set-option 'warning-as-error #t)
-#(ly:expect-warning (G_ "found music after \\fine"))
+trackB = <<
+
+  \clef bass
+  
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+>>
+
 
 \score {
-   { c4 \fine d4 }
-   \midi {}
+  <<
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
+  >>
+  \layout {}
+  \midi {}
 }
--- /home/yotam/pub/lilypond/input/regression/midi/pan-position.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/pan-position-midi.ly	2024-07-19 18:27:48.818309577 +0300
@@ -1,44 +1,104 @@
-\version "2.24.1"
+% Lily was here -- automatically converted by midi2ly.py from out-test/pan-position.midi
+\version "2.14.0"
 
-#(ly:set-option 'warning-as-error #t)
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
 
+% included from ./out-test/pan-position.header
 \header {
-  texidoc = "This tests @code{Staff.midiPanPosition}.  The sound
-should begin with an oboe at left and a bassoon at right.  They should
-switch positions in three steps at intervals of two seconds."
+texidoc=""
+options=""
+}
+% end
+
+trackAchannelA = {
+  
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 30 
+  \skip 1 
+  | % 2
+  
+}
+
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
+
+
+trackBchannelA = {
+  
+  \set Staff.instrumentName = "\new:"
+  
+}
+
+trackB = <<
+  \context Voice = voiceA \trackBchannelA
+>>
+
+
+trackCchannelA = {
+  
+  % [INSTRUMENT_NAME] oboe
+  \skip 1 
+  | % 2
+  
+}
+
+trackCchannelB = \relative c {
+  a''1 
+  | % 2
+  
+}
+
+trackC = <<
+  \context Voice = voiceA \trackCchannelA
+  \context Voice = voiceB \trackCchannelB
+>>
+
+
+trackDchannelA = {
+  
+  % [INSTRUMENT_NAME] bassoon
+  \skip 1 
+  | % 2
+  
+}
+
+trackDchannelB = \relative c {
+  f, 
+  | % 2
+  
 }
 
+trackD = <<
+
+  \clef bass
+  
+  \context Voice = voiceA \trackDchannelA
+  \context Voice = voiceB \trackDchannelB
+>>
+
+
 \score {
-  \midi {}
   <<
-    \tempo 4 = 30
-    \new Staff \with {
-      midiInstrument = "oboe" % something sustained for audio verification
-      midiPanPosition = -1.0
-    } <<
-      a'1
-      {
-        \skip 4
-        \set Staff.midiPanPosition = -0.33
-        \skip 4
-        \set Staff.midiPanPosition = 0.33
-        \skip 4
-        \set Staff.midiPanPosition = 1.0
-      }
-    >>
-    \new Staff \with {
-      midiInstrument = "bassoon" % something sustained for audio verification
-      midiPanPosition = 1.0
-    } <<
-      f,1
-      {
-        \skip 4
-        \set Staff.midiPanPosition = 0.33
-        \skip 4
-        \set Staff.midiPanPosition = -0.33
-        \skip 4
-        \set Staff.midiPanPosition = -1.0
-      }
-    >>
+    \context Staff \trackA
+    \context Staff=trackC \trackA
+    \context Staff=trackC \trackC
+    \context Staff=trackD \trackA
+    \context Staff=trackD \trackD
   >>
+  \layout {}
+  \midi {}
 }
--- /home/yotam/pub/lilypond/input/regression/midi/partcombine.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/partcombine-midi.ly	2024-07-19 18:27:49.358306400 +0300
@@ -1,18 +1,13 @@
-\version "2.21.0"
+% Lily was here -- automatically converted by midi2ly.py from out-test/partcombine.midi
+\version "2.14.0"
 
 \layout {
   \context {
     \Voice
-    \remove "Note_heads_engraver"
-    \consists "Completion_heads_engraver"
-    \remove "Rest_engraver"
-    \consists "Completion_rest_engraver"
-  }
-}
-\midi {
-  \context {
-    \Score
-    midiChannelMapping = #'instrument
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
   }
 }
 
@@ -16,30 +11,100 @@
   }
 }
 
+% included from ./out-test/partcombine.header
 \header {
 texidoc="PartCombined music is preserved"
 options="--skip"
 }
+% end
+
+trackAchannelA = {
+  
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 60 
+  \skip 1*5 
+}
+
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
+
+
+trackBchannelA = {
+  
+  \set Staff.instrumentName = "staff:one"
+  \skip 1*5 
+}
+
+trackBchannelB = {
+  
+  \set Staff.instrumentName = "staff:solo"
+  \skip 1*5 
+}
 
-instrumentOne = \relative {
-  c'4 d e f
-  R1
-  d'4 c b a
-  b4 g2 f4
-  e1
+trackBchannelC = {
+  
+  \set Staff.instrumentName = "staff:shared"
+  \skip 1*5 
 }
 
-instrumentTwo = \relative {
-  R1
-  g'4 a b c
-  d c b a
-  g f( e) d
-  e1
+trackBchannelD = {
+  
+  \set Staff.instrumentName = "staff:two"
+  \skip 1*5 
 }
 
+trackBchannelE = \relative c {
+  \voiceThree
+  s4*13 g''2 
+}
+
+trackBchannelF = \relative c {
+  \voiceTwo
+  s4*13 f'4 e 
+}
+
+trackBchannelG = \relative c {
+  \voiceOne
+  s1*2 d''4 c b a 
+  | % 4
+  <b g > s2 <f d >4 
+  | % 5
+  e1 
+  | % 6
+  
+}
+
+trackBchannelH = \relative c {
+  \voiceFour
+  c'4 d e f 
+  | % 2
+  g a b c 
+  | % 3
+  
+}
+
+trackB = <<
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+  \context Voice = voiceC \trackBchannelC
+  \context Voice = voiceD \trackBchannelD
+  \context Voice = voiceE \trackBchannelE
+  \context Voice = voiceF \trackBchannelF
+  \context Voice = voiceG \trackBchannelG
+  \context Voice = voiceH \trackBchannelH
+>>
+
+
 \score {
-<<
-  \new Staff = "staff" \partCombine \instrumentOne \instrumentTwo
+  <<
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
   >>
   \layout {}
   \midi {}
--- /home/yotam/pub/lilypond/input/regression/midi/performance-name-utf8.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/performance-name-utf8-midi.ly	2024-07-19 18:27:49.698304400 +0300
@@ -1,17 +1,66 @@
-\version "2.23.11"
+% Lily was here -- automatically converted by midi2ly.py from out-test/performance-name-utf8.midi
+\version "2.14.0"
 
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/performance-name-utf8.header
 \header {
-  texidoc = "The performance name of the MIDI file can contain
-Unicode characters.  They are encoded in UTF8.
+texidoc=""
+options=""
+}
+% end
+
+trackAchannelA = {
+  
+  \set Staff.instrumentName = "~~~~~~~~~~~~~~~~"
+  
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 60 
+  \skip 4 
+}
+
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
 
-Note: due to the way the regression test infrastructure currently
-works, any breakage in this test will not be noticed automatically."
+
+trackBchannelA = {
+  
+  \set Staff.instrumentName = "\new:"
+  \skip 4 
+}
+
+trackBchannelB = \relative c {
+  c'4 
 }
 
+trackB = <<
+
+  \clef bass
+  
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+>>
+
+
 \score {
-  \header {
-    title = "Łïłîρøñđ"
-  }
-  { c' }
-  \midi { }
+  <<
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
+  >>
+  \layout {}
+  \midi {}
 }
--- /home/yotam/pub/lilypond/input/regression/midi/quantize-duration-2.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/quantize-duration-2-midi.ly	2024-07-19 18:27:50.210301389 +0300
@@ -1,19 +1,13 @@
-% Lily was here -- automatically converted by out/bin/midi2ly from /home/janneke/testmidi.mid
-\version "2.19.21"
+% Lily was here -- automatically converted by midi2ly.py from out-test/quantize-duration-2.midi
+\version "2.14.0"
 
 \layout {
   \context {
     \Voice
-    \remove "Note_heads_engraver"
-    \consists "Completion_heads_engraver"
-    \remove "Rest_engraver"
-    \consists "Completion_rest_engraver"
-  }
-}
-\midi {
-  \context {
-    \Score
-    midiChannelMapping = #'instrument
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
   }
 }
 
@@ -17,25 +11,26 @@
   }
 }
 
+% included from ./out-test/quantize-duration-2.header
 \header {
 texidoc="@code{midi2ly}'s option @option{--duration-quant} preserves first note length (16)."
 options="--duration-quant=16"
 }
+% end
 
 trackAchannelA = {
 
 
   \key a \major
     
-  % [TEXT_EVENT] Nokia Tune
+  % [TEXT_EVENT] creator: 
   
-  \tempo 4 = 120 
-  
-
-  \key a \major
+  % [TEXT_EVENT] LilyPond 2.25.18              
   
   \time 3/8 
   
+  \tempo 4 = 120 
+  \skip 4*1138/384 
 }
 
 trackA = <<
@@ -43,15 +38,33 @@ trackA = <<
 >>
 
 
-trackBchannelB = \relative {
-  e'''4*41/192 r4*7/192 d4*41/192 r4*7/192 fis,4*89/192 r4*7/192 gis4*89/192 
-  r4*7/192 
+trackBchannelA = {
+  
+  \set Staff.instrumentName = "trackB:voiceA"
+  \skip 4*1138/384 
+}
+
+trackBchannelB = {
+  
+  \set Staff.instrumentName = "trackB:"
+  
+
+  \key a \major
+  \skip 4*1138/384 
+}
+
+trackBchannelC = \relative c {
+  e'''16 d fis,8 gis 
   | % 2
-  cis4*41/192 r4*7/192 b4*41/192 r4*7/192 d,4*89/192 r4*7/192 e4*89/192 
+  cis16 b d,8 e 
+  | % 3
+  
 }
 
 trackB = <<
-  \context Voice = voiceA \trackBchannelB
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+  \context Voice = voiceC \trackBchannelC
 >>
 
 
--- /home/yotam/pub/lilypond/input/regression/midi/quantize-duration.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/quantize-duration-midi.ly	2024-07-19 18:27:50.682298613 +0300
@@ -1,19 +1,13 @@
-% Lily was here -- automatically converted by ../../../scripts/midi2ly.py from out-www/voice-2.midi
-\version "2.16.0"
+% Lily was here -- automatically converted by midi2ly.py from out-test/quantize-duration.midi
+\version "2.14.0"
 
 \layout {
   \context {
     \Voice
-    \remove "Note_heads_engraver"
-    \consists "Completion_heads_engraver"
-    \remove "Rest_engraver"
-    \consists "Completion_rest_engraver"
-  }
-}
-\midi {
-  \context {
-    \Score
-    midiChannelMapping = #'instrument
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
   }
 }
 
@@ -26,16 +20,14 @@ options="--duration-quant=4"
 
 trackAchannelA = {
   
-  % [SEQUENCE_TRACK_NAME] control track
-  
   % [TEXT_EVENT] creator: 
   
-  % [TEXT_EVENT] GNU LilyPond 2.13.54          
+  % [TEXT_EVENT] LilyPond 2.25.18              
   
   \time 4/4 
   
   \tempo 4 = 60 
-  
+  \skip 2 
 }
 
 trackA = <<
@@ -43,18 +35,35 @@ trackA = <<
 >>
 
 
-trackBchannelA = \relative c {
-  r4*1/8 c4*7/8 
-  c4*7/8 r4*1/8
+trackBchannelA = {
+  
+  \set Staff.instrumentName = "trackB:voiceA"
+  \skip 2 
+}
+
+trackBchannelB = \relative c {
+  \voiceOne
+  r32 c4 
+}
+
+trackBchannelBvoiceB = \relative c {
+  \voiceTwo
+  r4 c 
 }
 
 trackB = <<
+
+  \clef bass
+  
   \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+  \context Voice = voiceC \trackBchannelBvoiceB
 >>
 
 
 \score {
   <<
+    \context Staff=trackB \trackA
     \context Staff=trackB \trackB
   >>
   \layout {}
--- /home/yotam/pub/lilypond/input/regression/midi/quantize-start.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/quantize-start-midi.ly	2024-07-19 18:27:51.138295931 +0300
@@ -1,19 +1,13 @@
-% Lily was here -- automatically converted by ../../../scripts/midi2ly.py from out-www/voice-2.midi
-\version "2.16.0"
+% Lily was here -- automatically converted by midi2ly.py from out-test/quantize-start.midi
+\version "2.14.0"
 
 \layout {
   \context {
     \Voice
-    \remove "Note_heads_engraver"
-    \consists "Completion_heads_engraver"
-    \remove "Rest_engraver"
-    \consists "Completion_rest_engraver"
-  }
-}
-\midi {
-  \context {
-    \Score
-    midiChannelMapping = #'instrument
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
   }
 }
 
@@ -26,16 +20,14 @@ options="--start-quant=4"
 
 trackAchannelA = {
   
-  % [SEQUENCE_TRACK_NAME] control track
-  
   % [TEXT_EVENT] creator: 
   
-  % [TEXT_EVENT] GNU LilyPond 2.13.54          
+  % [TEXT_EVENT] LilyPond 2.25.18              
   
   \time 4/4 
   
   \tempo 4 = 60 
-  
+  \skip 2 
 }
 
 trackA = <<
@@ -43,18 +35,28 @@ trackA = <<
 >>
 
 
-trackBchannelA = \relative c {
-  r4*1/8 c4*7/8 
-  c4*7/8 r4*1/8
+trackBchannelA = {
+  
+  \set Staff.instrumentName = "trackB:voiceA"
+  \skip 2 
+}
+
+trackBchannelB = \relative c {
+  c4 c 
 }
 
 trackB = <<
+
+  \clef bass
+  
   \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
 >>
 
 
 \score {
   <<
+    \context Staff=trackB \trackA
     \context Staff=trackB \trackB
   >>
   \layout {}
--- /home/yotam/pub/lilypond/input/regression/midi/rehearsal-mark.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/rehearsal-mark-midi.ly	2024-07-19 18:27:51.470293979 +0300
@@ -1,12 +1,83 @@
-\version "2.23.6"
+% Lily was here -- automatically converted by midi2ly.py from out-test/rehearsal-mark.midi
+\version "2.14.0"
 
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/rehearsal-mark.header
 \header {
-  texidoc="Rehearsal marks create MIDI Marker events."
+texidoc=""
+options=""
+}
+% end
+
+trackAchannelA = {
+  
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 60 
+  \skip 1 
+  | % 2
+  
+  % [MARKER] A
+  \skip 1 
+  | % 3
+  
+  % [MARKER] B
+  \skip 1 
+  | % 4
+  
+  % [MARKER] X
+  \skip 1 
+  | % 5
+  
+}
+
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
+
+
+trackBchannelA = {
+  
+  \set Staff.instrumentName = "\new:"
+  \skip 1*4 
+}
+
+trackBchannelB = \relative c {
+  c'1 
+  | % 2
+  d 
+  | % 3
+  e 
+  | % 4
+  f 
+  | % 5
+  
 }
 
+trackB = <<
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+>>
+
+
 \score {
-  \fixed c' {
-    c1 \mark \default d1 \mark \default e1 \mark 23 f1
-  }
+  <<
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
+  >>
+  \layout {}
   \midi {}
 }
--- /home/yotam/pub/lilypond/input/regression/midi/repeat-segno.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/repeat-segno-midi.ly	2024-07-19 18:27:51.810291980 +0300
@@ -1,22 +1,105 @@
-\version "2.23.6"
+% Lily was here -- automatically converted by midi2ly.py from out-test/repeat-segno.midi
+\version "2.14.0"
 
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/repeat-segno.header
 \header {
-  texidoc="Segni and coda signs create MIDI Marker events."
+texidoc=""
+options=""
+}
+% end
+
+trackAchannelA = {
+  
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 60 
+  \skip 1 
+  | % 2
+  
+  % [MARKER] ~~~~
+  \skip 1 
+  | % 3
+  
+  % [MARKER] ~~~~
+  \skip 1*3 
+  % [MARKER] ~~~~~~~~
+  \skip 1 
+  | % 7
+  
+  % [MARKER] ~~~~~~~~
+  \skip 1*3 
+  % [MARKER] ~~~~~~~3~~~~~~~
+  \skip 1 
+  | % 11
+  
+  % [MARKER] ~~~~~~~3~~~~~~~
+  \skip 1*2 
+}
+
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
+
+
+trackBchannelA = {
+  
+  \set Staff.instrumentName = "\new:"
+  \skip 1*12 
+}
+
+trackBchannelB = \relative c {
+  c'1 
+  | % 2
+  d 
+  | % 3
+  e 
+  | % 4
+  f 
+  | % 5
+  c 
+  | % 6
+  d 
+  | % 7
+  e 
+  | % 8
+  f 
+  | % 9
+  c 
+  | % 10
+  d 
+  | % 11
+  e 
+  | % 12
+  f 
+  | % 13
+  
 }
 
+trackB = <<
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+>>
+
+
 \score {
-  \fixed c' {
-    \repeat unfold 3 {
-      c1
-      \repeat segno 2 {
-        d1
-        \alternative {
-          e1
-          <>
-        }
-      }
-      f1
-    }
-  }
+  <<
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
+  >>
+  \layout {}
   \midi {}
 }
--- /home/yotam/pub/lilypond/input/regression/midi/rest-dynamic.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/rest-dynamic-midi.ly	2024-07-19 18:27:52.282289205 +0300
@@ -1,13 +1,63 @@
-\version "2.19.21"
+% Lily was here -- automatically converted by midi2ly.py from out-test/rest-dynamic.midi
+\version "2.14.0"
+
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/rest-dynamic.header
 \header {
 texidoc="LilyPond respects rests, also when there are dynamics"
 options=""
 }
+% end
+
+trackAchannelA = {
+  
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 2/4 
+  
+  \tempo 4 = 60 
+  \skip 1 
+}
+
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
+
+
+trackBchannelA = {
+  \skip 1 
+}
+
+trackBchannelB = \relative c {
+  e'4 e 
+  | % 2
+  r4 e 
+  | % 3
+  
+}
+
+trackB = <<
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+>>
+
 
 \score {
-  \new Staff \relative {
-    \time 2/4 e'4 e r4 e\f |
-  }
+  <<
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
+  >>
   \layout {}
   \midi {}
 }
--- /home/yotam/pub/lilypond/input/regression/midi/rest.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/rest-midi.ly	2024-07-19 18:27:52.770286335 +0300
@@ -1,19 +1,13 @@
-% Lily was here -- automatically converted by ../../../scripts/midi2ly.py from out-www/rest.midi
-\version "2.19.21"
+% Lily was here -- automatically converted by midi2ly.py from out-test/rest.midi
+\version "2.14.0"
 
 \layout {
   \context {
     \Voice
-    \remove "Note_heads_engraver"
-    \consists "Completion_heads_engraver"
-    \remove "Rest_engraver"
-    \consists "Completion_rest_engraver"
-  }
-}
-\midi {
-  \context {
-    \Score
-    midiChannelMapping = #'instrument
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
   }
 }
 
@@ -26,15 +20,15 @@ options=""
 
 trackAchannelA = {
   
-  % [SEQUENCE_TRACK_NAME] control track
-  
   % [TEXT_EVENT] creator: 
   
-  % [TEXT_EVENT] GNU LilyPond 2.13.54          
+  % [TEXT_EVENT] LilyPond 2.25.18              
   
   \time 4/4 
   
   \tempo 4 = 60 
+  \skip 1 
+  | % 2
   
 }
 
@@ -43,10 +37,16 @@ trackA = <<
 >>
 
 
-trackBchannelA = \relative {
-  r4 
+trackBchannelA = {
+  
   \set Staff.instrumentName = "trackB:voiceA"
-  a, r4 a 
+  \skip 1 
+  | % 2
+  
+}
+
+trackBchannelB = \relative c {
+  r4 a r4 a 
   | % 2
   
 }
@@ -56,6 +56,7 @@ trackB = <<
   \clef bass
   
   \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
 >>
 
 
--- /home/yotam/pub/lilypond/input/regression/midi/section-label.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/section-label-midi.ly	2024-07-19 18:27:53.102284384 +0300
@@ -1,13 +1,73 @@
-\version "2.23.6"
+% Lily was here -- automatically converted by midi2ly.py from out-test/section-label.midi
+\version "2.14.0"
 
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/section-label.header
 \header {
-  texidoc="Section labels create MIDI Marker events."
+texidoc=""
+options=""
+}
+% end
+
+trackAchannelA = {
+  
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  % [MARKER] Intro
+  
+  \tempo 4 = 60 
+  \skip 1 
+  | % 2
+  
+  % [MARKER] Coda
+  \skip 1 
+  | % 3
+  
+}
+
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
+
+
+trackBchannelA = {
+  
+  \set Staff.instrumentName = "\new:"
+  \skip 1*2 
+}
+
+trackBchannelB = \relative c {
+  c'1 
+  | % 2
+  d 
+  | % 3
+  
 }
 
+trackB = <<
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+>>
+
+
 \score {
-  \fixed c' {
-    \sectionLabel \markup \smallCaps "Intro" c1
-    \sectionLabel "Coda" d1
-  }
+  <<
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
+  >>
+  \layout {}
   \midi {}
 }
--- /home/yotam/pub/lilypond/input/regression/midi/sequence-name-scoping.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/sequence-name-scoping-midi.ly	2024-07-19 18:27:54.146278246 +0300
@@ -1,7 +1,19 @@
-\version "2.19.25"
+% Lily was here -- automatically converted by midi2ly.py from out-test/sequence-name-scoping.midi
+\version "2.14.0"
 
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/sequence-name-scoping.header
 \header {
-  texidoc="If a score has a @code{\header} block which defines a title,
+texidoc="If a score has a @code{\header} block which defines a title,
   this title should override any title defined in a @code{\header} block
   of the score's enclosing @code{\bookpart} or @code{\book} (or a title
   defined in a top-level @code{\header} block) when naming the MIDI
@@ -10,73 +22,57 @@
   using the title defined in the @code{\header} block of the nearest
   enclosing @code{\bookpart}, @code{\book}, or top-level scope that
   contains a title definition."
-  title = "Top-level title"
+options=""
 }
+% end
 
-music = \new Staff { c1 }
+trackAchannelA = {
+  
+  \set Staff.instrumentName = "Book"
+  
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 60 
+  \skip 1 
+  | % 2
+  
+}
 
-% Book with a title defined in a \header block, and book parts
-\book {
-  \header { title = "Book" }
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
 
-  % score without a \header block outside of any book part -- the MIDI
-  % sequence should get the title of the book as its name
-  \score {
-    \music
-    \layout { }
-    \midi { }
-  }
 
-  % score with a \header and a title outside of any book part -- the MIDI
-  % sequence should be named with the title from this \header block
-  \score {
-    \music
-    \header { title = "Score in a \book" }
-    \layout { }
-    \midi { }
-  }
-
-  % Book part with a \header block and a title
-  \bookpart {
-    \header { title = "Book part" }
-
-    % score without a \header block -- the MIDI sequence should get its name
-    % from the title of the enclosing book part
-    \score {
-      \music
-      \layout { }
-      \midi { }
-    }
-
-    % score with a \header (and a title) of its own -- the MIDI sequence
-    % should get its name from the title in this \header block
-    \score {
-      \music
-      \header { title = "Score in a book part (w/ a title) of a book" }
-      \layout { }
-      \midi { }
-    }
-  }
+trackBchannelA = {
+  \skip 1 
+  | % 2
+  
+}
 
-  % Book part without a \header block
-  \bookpart {
+trackBchannelB = \relative c {
+  c1 
+  | % 2
+  
+}
 
-    % score without a \header block -- the MIDI sequence should be named
-    % using the title from the enclosing book
-    \score {
-      \music
-      \layout { }
-      \midi { }
-    }
-
-    % score with a \header block and title -- the MIDI sequence should get
-    % its name from the title in this \header block
-    \score {
-      \music
-      \header { title = "Score in a book part (w/o a title) of a book" }
-      \layout { }
-      \midi { }
-    }
-  }
+trackB = <<
 
+  \clef bass
+  
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+>>
+
+
+\score {
+  <<
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
+  >>
+  \layout {}
+  \midi {}
 }
--- /home/yotam/pub/lilypond/input/regression/midi/sequence-name.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/sequence-name-midi.ly	2024-07-19 18:27:53.598281468 +0300
@@ -1,7 +1,19 @@
-\version "2.19.25"
+% Lily was here -- automatically converted by midi2ly.py from out-test/sequence-name.midi
+\version "2.14.0"
 
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/sequence-name.header
 \header {
-  texidoc="The MIDI sequence generated from a score should get its name
+texidoc="The MIDI sequence generated from a score should get its name
   from the title defined in the score's @code{\header} block (if any).
   The title used for layout can be overridden for MIDI output by
   specifying a separate @code{midititle} in the @code{\header} block.
@@ -9,55 +21,57 @@
   @code{\bookpart}, @code{\book}, or top-level scope with a @code{\header}
   block that defines a title, either, the MIDI sequence should get the
   default name."
+options=""
 }
+% end
 
-music = \new Staff { c1 }
+trackAchannelA = {
+  
+  \set Staff.instrumentName = "Title shared between layout and MIDI"
+  
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 60 
+  \skip 1 
+  | % 2
+  
+}
 
-% Book without a \header block
-\book {
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
 
-  % score with a \header block including a title -- the MIDI sequence
-  % should get its name from this \header block
-  \score {
-    \music
-    \header { title = "Title shared between layout and MIDI" }
-    \layout { }
-    \midi { }
-  }
 
-  % score with no title, but a midititle defined in a \header block --
-  % the MIDI sequence should be named using the midititle in this \header
-  % block
-  \score {
-    \music
-    \header { midititle = "No title for layout, but a title for MIDI" }
-    \layout { }
-    \midi { }
-  }
-
-  % score with a title and a midititle defined in a \header block -- the
-  % MIDI sequence should get the midititle in this \header block as its
-  % name
-  \score {
-    \music
-    \header {
-      title = "Title for layout"
-      midititle = "Title for MIDI"
-    }
-    \layout { }
-    \midi { }
-  }
+trackBchannelA = {
+  \skip 1 
+  | % 2
+  
+}
 
-  % Book part with no \header block
-  \bookpart {
+trackBchannelB = \relative c {
+  c1 
+  | % 2
+  
+}
 
-    % score without a \header -- the MIDI sequence should get the default
-    % name
-    \score {
-      \music
-      \layout { }
-      \midi { }
-    }
-  }
+trackB = <<
 
+  \clef bass
+  
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+>>
+
+
+\score {
+  <<
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
+  >>
+  \layout {}
+  \midi {}
 }
--- /home/yotam/pub/lilypond/input/regression/midi/spacer-no-notes.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/spacer-no-notes-midi.ly	2024-07-19 18:27:54.478276295 +0300
@@ -1,16 +1,61 @@
-\version "2.21.7"
+% Lily was here -- automatically converted by midi2ly.py from out-test/spacer-no-notes.midi
+\version "2.14.0"
 
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/spacer-no-notes.header
 \header {
-  texidoc = "This test should produce a two-measure MIDI file in the
-key of G minor, but with no notes."
+texidoc=""
+options=""
+}
+% end
+
+trackAchannelA = {
+
+
+  \key g \minor
+    
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 60 
+  \skip 1*2 
 }
 
-#(ly:set-option 'warning-as-error #t)
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
+
+
+trackBchannelA = {
+  
+  \set Staff.instrumentName = "\new:"
+  
+
+  \key g \minor
+  \skip 1*2 
+}
+
+trackB = <<
+  \context Voice = voiceA \trackBchannelA
+>>
+
 
 \score {
-  {
-    \key g \minor
-    s1*2
-  }
+  <<
+    \context Staff \trackA
+  >>
+  \layout {}
   \midi {}
 }
--- /home/yotam/pub/lilypond/input/regression/midi/staff-map-instrument.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/staff-map-instrument-midi.ly	2024-07-19 18:27:55.038273003 +0300
@@ -1,44 +1,130 @@
-\version "2.19.21"
+% Lily was here -- automatically converted by midi2ly.py from out-test/staff-map-instrument.midi
+\version "2.14.0"
 
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/staff-map-instrument.header
 \header {
 texidoc="Midi2ly remaps voices correctly to staves in MIDI-files that use instrument<->channel mapping when combined with voice<->track mapping.  TODO: pianostaff"
 options=""
 }
+% end
+
+trackAchannelA = {
+
+
+  \key c \minor
+    
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 60 
+  \skip 1*2 
+}
+
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
+
+
+trackBchannelA = {
+  
+  \set Staff.instrumentName = "treble:one"
+  \skip 1*2 
+}
+
+trackBchannelB = {
+  
+  \set Staff.instrumentName = "treble:"
+  
+
+  \key c \minor
+  \skip 1*2 
+}
+
+trackBchannelC = {
+  
+  \set Staff.instrumentName = "treble:two"
+  \skip 1*2 
+}
+
+trackBchannelD = \relative c {
+  \voiceOne
+  f''8 ees16 d c bes aes g f8 aes' g f 
+  | % 2
+  f,16 g aes4 g16 f e2 
+  | % 3
+  
+}
+
+trackBchannelE = \relative c {
+  \voiceTwo
+  c''4 r4. f8 ees d 
+  | % 2
+  <b, d > r8 <b d > r8 <g c >2 
+  | % 3
+  
+}
+
+trackB = <<
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+  \context Voice = voiceC \trackBchannelC
+  \context Voice = voiceD \trackBchannelD
+  \context Voice = voiceE \trackBchannelE
+>>
+
+
+trackCchannelA = {
+  
+  \set Staff.instrumentName = "bass:three"
+  \skip 1*2 
+}
+
+trackCchannelB = {
+  
+  \set Staff.instrumentName = "bass:"
+  
+
+  \key c \minor
+  \skip 1*2 
+}
+
+trackCchannelC = \relative c {
+  r8 c'16 b c8 g aes c16 b c8 d 
+  | % 2
+  <c,, c' >1 
+  | % 3
+  
+}
+
+trackC = <<
+
+  \clef bass
+  
+  \context Voice = voiceA \trackCchannelA
+  \context Voice = voiceB \trackCchannelB
+  \context Voice = voiceC \trackCchannelC
+>>
+
 
 \score {
-%% TODO:PIANOSTAFF  \context PianoStaff <<
   <<
-    \context Staff = "treble" <<
-      \set Score.midiChannelMapping = #'instrument
-      \context Voice="one" \relative {
-	\time 4/4
-	\key c \minor
-	\voiceOne
-%comes
-%7
-	f''8 es16 d c16 bes ! as g f8 as' g f 
-%8	es8 d es f b, c d b |
-	f,16 g as4 g16 f e2 |
-      }
-      \context Voice="two" \relative {
-	\voiceTwo
-%dux
-%7
-	c''4 r4 r8 f es d |
-%8	r8 as g f g f16 es f8 d | 
-	<b, d>8 r <b d> r <g c>2 |
-      }
-    >>
-    \context Staff = "bass" <<
-      \context Voice="three" \relative {
-	\key c \minor
-	\clef bass
-%7
-	r8 c'16 b c8 g as c16 b c8 d |
-%8	g8 c16 b c8 d f,16 g as4 g16 f | 
-	<c,, c'>1
-      }
-    >>
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
+    \context Staff=trackC \trackA
+    \context Staff=trackC \trackC
   >>
   \layout {}
   \midi {}
--- /home/yotam/pub/lilypond/input/regression/midi/staff-map-voice.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/staff-map-voice-midi.ly	2024-07-19 18:27:55.590269759 +0300
@@ -1,44 +1,109 @@
-\version "2.19.21"
+% Lily was here -- automatically converted by midi2ly.py from out-test/staff-map-voice.midi
+\version "2.14.0"
 
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/staff-map-voice.header
 \header {
 texidoc="Midi2ly remaps voices correctly to staves in MIDI-files that use voice<->channel mapping when combined with staff<->track mapping.  TODO: pianostaff"
 options=""
 }
+% end
+
+trackAchannelA = {
+
+
+  \key c \minor
+    
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 60 
+  \skip 1*2 
+}
+
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
+
+
+trackBchannelA = {
+  
+  \set Staff.instrumentName = "treble:one"
+  
+
+  \key c \minor
+  \skip 1*2 
+}
+
+trackBchannelB = \relative c {
+  \voiceOne
+  f''8 ees16 d c bes aes g f8 aes' g f 
+  | % 2
+  f,16 g aes4 g16 f e2 
+  | % 3
+  
+}
+
+trackBchannelC = \relative c {
+  \voiceTwo
+  c''4 r4. f8 ees d 
+  | % 2
+  <b, d > r8 <b d > r8 <g c >2 
+  | % 3
+  
+}
+
+trackB = <<
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+  \context Voice = voiceC \trackBchannelC
+>>
+
+
+trackCchannelA = {
+  
+  \set Staff.instrumentName = "bass:three"
+  
+
+  \key c \minor
+  \skip 1*2 
+}
+
+trackCchannelB = \relative c {
+  r8 c'16 b c8 g aes c16 b c8 d 
+  | % 2
+  <c,, c' >1 
+  | % 3
+  
+}
+
+trackC = <<
+
+  \clef bass
+  
+  \context Voice = voiceA \trackCchannelA
+  \context Voice = voiceB \trackCchannelB
+>>
+
 
 \score {
-%% TODO:PIANOSTAFF  \context PianoStaff <<
   <<
-    \context Staff = "treble" <<
-      \set Score.midiChannelMapping = #'voice
-      \context Voice="one" \relative {
-	\time 4/4
-	\key c \minor
-	\voiceOne
-%comes
-%7
-	f''8 es16 d c16 bes ! as g f8 as' g f 
-%8	es8 d es f b, c d b |
-	f,16 g as4 g16 f e2 |
-      }
-      \context Voice="two" \relative {
-	\voiceTwo
-%dux
-%7
-	c''4 r4 r8 f es d |
-%8	r8 as g f g f16 es f8 d | 
-	<b, d>8 r <b d> r <g c>2 |
-      }
-    >>
-    \context Staff = "bass" <<
-      \context Voice="three" \relative {
-	\key c \minor
-	\clef bass
-%7
-	r8 c'16 b c8 g as c16 b c8 d |
-%8	g8 c16 b c8 d f,16 g as4 g16 f | 
-	<c,, c'>1
-      }
-    >>
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
+    \context Staff=trackC \trackA
+    \context Staff=trackC \trackC
   >>
   \layout {}
   \midi {}
--- /home/yotam/pub/lilypond/input/regression/midi/varcoda.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/varcoda-midi.ly	2024-07-19 18:27:55.926267785 +0300
@@ -1,15 +1,71 @@
-\version "2.23.6"
+% Lily was here -- automatically converted by midi2ly.py from out-test/varcoda.midi
+\version "2.14.0"
 
+\layout {
+  \context {
+    \Voice
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
+  }
+}
+
+% included from ./out-test/varcoda.header
 \header {
-  texidoc="Variant coda signs create MIDI Marker events."
+texidoc=""
+options=""
+}
+% end
+
+trackAchannelA = {
+  
+  % [TEXT_EVENT] creator: 
+  
+  % [TEXT_EVENT] LilyPond 2.25.18              
+  
+  \time 4/4 
+  
+  \tempo 4 = 60 
+  \skip 1 
+  | % 2
+  
+  % [MARKER] ~~~~
+  \skip 1 
+  | % 3
+  
+}
+
+trackA = <<
+  \context Voice = voiceA \trackAchannelA
+>>
+
+
+trackBchannelA = {
+  
+  \set Staff.instrumentName = "\new:"
+  \skip 1*2 
+}
+
+trackBchannelB = \relative c {
+  c'1 
+  | % 2
+  d 
+  | % 3
+  
 }
 
+trackB = <<
+  \context Voice = voiceA \trackBchannelA
+  \context Voice = voiceB \trackBchannelB
+>>
+
+
 \score {
-  \fixed c' {
-    \set Score.codaMarkFormatter = #format-varcoda-mark
-    c1
-    \codaMark \default
-    d1
-  }
+  <<
+    \context Staff=trackB \trackA
+    \context Staff=trackB \trackB
+  >>
+  \layout {}
   \midi {}
 }
--- /home/yotam/pub/lilypond/input/regression/midi/voice-2.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/voice-2-midi.ly	2024-07-19 18:27:56.418264893 +0300
@@ -1,19 +1,13 @@
-% Lily was here -- automatically converted by ../../../scripts/midi2ly.py from out-www/voice-2.midi
-\version "2.19.21"
+% Lily was here -- automatically converted by midi2ly.py from out-test/voice-2.midi
+\version "2.14.0"
 
 \layout {
   \context {
     \Voice
-    \remove "Note_heads_engraver"
-    \consists "Completion_heads_engraver"
-    \remove "Rest_engraver"
-    \consists "Completion_rest_engraver"
-  }
-}
-\midi {
-  \context {
-    \Score
-    midiChannelMapping = #'instrument
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
   }
 }
 
@@ -26,15 +20,15 @@ options=""
 
 trackAchannelA = {
   
-  % [SEQUENCE_TRACK_NAME] control track
-  
   % [TEXT_EVENT] creator: 
   
-  % [TEXT_EVENT] GNU LilyPond 2.13.54          
+  % [TEXT_EVENT] LilyPond 2.25.18              
   
   \time 4/4 
   
   \tempo 4 = 60 
+  \skip 1 
+  | % 2
   
 }
 
@@ -43,16 +37,30 @@ trackA = <<
 >>
 
 
-trackBchannelA = \relative {
-  \voiceOne
+trackBchannelA = {
+  
+  \set Staff.instrumentName = "trackB:voiceA"
+  \skip 1 
+  | % 2
   
-  \set Staff.instrumentName = ":1"
+}
+
+trackBchannelB = {
+  
+  \set Staff.instrumentName = "trackB:voiceB"
+  \skip 1 
+  | % 2
+  
+}
+
+trackBchannelC = \relative c {
+  \voiceOne
   e''4 e e e 
   | % 2
   
 }
 
-trackBchannelB = \relative {
+trackBchannelD = \relative c {
   \voiceTwo
   f' f f f 
   | % 2
@@ -62,6 +70,8 @@ trackBchannelB = \relative {
 trackB = <<
   \context Voice = voiceA \trackBchannelA
   \context Voice = voiceB \trackBchannelB
+  \context Voice = voiceC \trackBchannelC
+  \context Voice = voiceD \trackBchannelD
 >>
 
 
--- /home/yotam/pub/lilypond/input/regression/midi/voice-4.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/voice-4-midi.ly	2024-07-19 18:27:56.926261908 +0300
@@ -1,19 +1,13 @@
-% Lily was here -- automatically converted by ../../../scripts/midi2ly.py from out-www/voice-4.midi
-\version "2.19.21"
+% Lily was here -- automatically converted by midi2ly.py from out-test/voice-4.midi
+\version "2.14.0"
 
 \layout {
   \context {
     \Voice
-    \remove "Note_heads_engraver"
-    \consists "Completion_heads_engraver"
-    \remove "Rest_engraver"
-    \consists "Completion_rest_engraver"
-  }
-}
-\midi {
-  \context {
-    \Score
-    midiChannelMapping = #'instrument
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
   }
 }
 
@@ -26,15 +20,15 @@ options=""
 
 trackAchannelA = {
   
-  % [SEQUENCE_TRACK_NAME] control track
-  
   % [TEXT_EVENT] creator: 
   
-  % [TEXT_EVENT] GNU LilyPond 2.13.54          
+  % [TEXT_EVENT] LilyPond 2.25.18              
   
   \time 4/4 
   
   \tempo 4 = 60 
+  \skip 1 
+  | % 2
   
 }
 
@@ -43,30 +37,60 @@ trackA = <<
 >>
 
 
-trackBchannelA = \relative {
-  \voiceOne
+trackBchannelA = {
+  
+  \set Staff.instrumentName = "trackB:voiceA"
+  \skip 1 
+  | % 2
+  
+}
+
+trackBchannelB = {
+  
+  \set Staff.instrumentName = "trackB:voiceD"
+  \skip 1 
+  | % 2
   
-  \set Staff.instrumentName = ":1"
+}
+
+trackBchannelC = {
+  
+  \set Staff.instrumentName = "trackB:voiceC"
+  \skip 1 
+  | % 2
+  
+}
+
+trackBchannelD = {
+  
+  \set Staff.instrumentName = "trackB:voiceB"
+  \skip 1 
+  | % 2
+  
+}
+
+trackBchannelE = \relative c {
+  \voiceOne
   <c''' a >2 b 
   | % 2
   
 }
 
-trackBchannelB = \relative {
+trackBchannelF = \relative c {
   \voiceThree
   c''4. d8 e4 f 
   | % 2
   
 }
 
-trackBchannelC = \relative {
+trackBchannelG = \relative c {
   \voiceFour
   d'1 
   | % 2
   
 }
 
-trackBchannelD = \relative {
+trackBchannelH = \relative c {
   \voiceTwo
   c'4 c2 c4 
   | % 2
@@ -78,6 +102,10 @@ trackB = <<
   \context Voice = voiceB \trackBchannelB
   \context Voice = voiceC \trackBchannelC
   \context Voice = voiceD \trackBchannelD
+  \context Voice = voiceE \trackBchannelE
+  \context Voice = voiceF \trackBchannelF
+  \context Voice = voiceG \trackBchannelG
+  \context Voice = voiceH \trackBchannelH
 >>
 
 
--- /home/yotam/pub/lilypond/input/regression/midi/voice-5.ly	2023-10-18 21:06:35.366721450 +0300
+++ out-test/recovered/voice-5-midi.ly	2024-07-19 18:27:57.454258805 +0300
@@ -1,19 +1,13 @@
-% Lily was here -- automatically converted by ../../../scripts/midi2ly.py from out-www/voice-4.midi
-\version "2.19.21"
+% Lily was here -- automatically converted by midi2ly.py from out-test/voice-5.midi
+\version "2.14.0"
 
 \layout {
   \context {
     \Voice
-    \remove "Note_heads_engraver"
-    \consists "Completion_heads_engraver"
-    \remove "Rest_engraver"
-    \consists "Completion_rest_engraver"
-  }
-}
-\midi {
-  \context {
-    \Score
-    midiChannelMapping = #'instrument
+    \remove Note_heads_engraver
+    \consists Completion_heads_engraver
+    \remove Rest_engraver
+    \consists Completion_rest_engraver
   }
 }
 
@@ -26,16 +20,14 @@ options=""
 
 trackAchannelA = {
   
-  % [SEQUENCE_TRACK_NAME] control track
-  
   % [TEXT_EVENT] creator: 
   
-  % [TEXT_EVENT] GNU LilyPond 2.13.54          
+  % [TEXT_EVENT] LilyPond 2.25.18              
   
   \time 4/4 
   
   \tempo 4 = 60 
-  
+  \skip 1*2 
 }
 
 trackA = <<
@@ -43,39 +35,69 @@ trackA = <<
 >>
 
 
-trackBchannelA = \relative {
-  \voiceOne
+trackBchannelA = {
   
-  \set Staff.instrumentName = ":1"
+  \set Staff.instrumentName = "trackB:voiceA"
+  \skip 1*2 
+}
+
+trackBchannelB = {
+  
+  \set Staff.instrumentName = "trackB:voiceE"
+  \skip 1*2 
+}
+
+trackBchannelC = {
+  
+  \set Staff.instrumentName = "trackB:voiceD"
+  \skip 1*2 
+}
+
+trackBchannelD = {
+  
+  \set Staff.instrumentName = "trackB:voiceC"
+  \skip 1*2 
+}
+
+trackBchannelE = {
+  
+  \set Staff.instrumentName = "trackB:voiceB"
+  \skip 1*2 
+}
+
+trackBchannelF = \relative c {
+  \voiceOne
   <c''' a >2 b 
   | % 2
   
 }
 
-trackBchannelB = \relative {
+trackBchannelG = \relative c {
   \voiceThree
   c''4. d8 e4 f 
   | % 2
   
 }
 
-trackBchannelC = \relative {
+trackBchannelH = \relative c {
   \voiceFour
   d'1 
   | % 2
   
 }
 
-trackBchannelD = \relative {
+trackBchannelI = \relative c {
   \voiceTwo
   c'4 c2 c4 
   | % 2
   
 }
 
-trackBchannelE = \relative {
-  s1 d1
+trackBchannelJ = \relative c {
+  r1 
   | % 2
+  d 
+  | % 3
   
 }
 
@@ -85,6 +107,11 @@ trackB = <<
   \context Voice = voiceC \trackBchannelC
   \context Voice = voiceD \trackBchannelD
   \context Voice = voiceE \trackBchannelE
+  \context Voice = voiceF \trackBchannelF
+  \context Voice = voiceG \trackBchannelG
+  \context Voice = voiceH \trackBchannelH
+  \context Voice = voiceI \trackBchannelI
+  \context Voice = voiceJ \trackBchannelJ
 >>