Andre-- You did reproduce the behavior in the 3rd line, and that's really the only one that matters. My point is that when I use the same methods within the PMX FORTRAN to handle codas as shakes, naturally I get the same octave shifts (up one for the ornament, then down one for the next), but this doesn't work because the musixtex definition of \coda somehow shields the following command from the octave shift, where as the definition of \shake lets the following command see the octave shift.
Without any quick answers from the list, I just worked out a different musixtex definition to replace \coda, so now TeX does treat the octave changes the same as \shake, and therefore I can use PMX to calculate the octave changes the same way for codas as for shakes. --Don >-----Original Message----- >From: TeX-Music [mailto:[email protected]] On Behalf Of Andre Van >Ryckeghem >Sent: Friday, April 05, 2013 10:35 PM >To: Werner Icking Music Archive >Subject: Re: [Tex-music] Adventures with \coda > >Dear Don, > >I cannot reproduce the behaviour, i rather see the error at \shake and this >could be solved by putting it between {..} (for doing this i use the \relax >command). > >This are the tex lines from the PMX run > >\pnotes{2.83}\qu c\Coda{'g}\qu d\qu e\qu f\en% >\pnotes{2.83}\qu c\coda{'g}\qu d\qu e\qu f\en% >\pnotes{2.83}\qu c\shake{'g}\qu{`d}\qu e\qu f\en% >\pnotes{2.83}\qu c\shake{'g}\qu d\qu e\qu f\en% ---> this one runs away >\pnotes{2.83}\qu c\relax{\shake{'g}}\qu d\qu e\qu f\en% --> this solves it >\pnotes{2.83}\qu c\relax{\coda{'g}}\qu d\qu e\qu f\en% > >I used pmx270 > >Andre > >============== >1 1 4 4 0 6 0 -1 >1 1 20 0 > >t >./ >w180m >Ap >% >c44 \Coda{'g}\ d e f / >c44 \coda{'g}\ d e f / >c44 d ot e f / >c44 \shake{'g}\ d e f / >c44 \relax{\shake{'g}}\ d e f / >c44 \relax{\coda{'g}}\ d e f / >=================== > > >-----Oorspronkelijk bericht----- >From: Don Simons >Sent: Friday, April 05, 2013 5:31 PM >To: 'Werner Icking Music Archive' >Subject: [Tex-music] Adventures with \coda > >I'm trying to respond to the request to get PMX to do coda symbols. My plan >is to do it more like an ordinary ornament than I did segno, so each >instance of "oC" in PMX will only lead to a single coda symbol rather than >one in every staff. That would let you tweak the position of each symbol >independently. > >In the process I've come across some weirdness in the behavior of \coda as >defined in musixtex.tex. My first test of my new PMX coding, with input "c44 >d oC e f /", produced > >\pnotes{2.83}\qu c\coda{'g}\qu{`d}\qu e\qu f\en% > >as expected. Compare this to what I got with just "ot" instead of "oC": > >\pnotes{2.83}\qu c\shake{'g}\qu{`d}\qu e\qu f\en% > >Looks the same except for the ornament name, right? Well, it doesn't have >the same effect, because in the one with \coda, the octave change character >on g is not respected on the following notes, and they come out an octave >too low. I've looked at the definitions in musixtex.tex, and see that \coda >is defined differently so as to give an automatic horizontal offset from >where it is entered. But somehow that more complicated definition shields >the rest of the world from the octave change symbol in the argument of >\coda. > >Since I refuse to revisit and revise PMX's tried and true pitch level >tracking just to accommodate this odd behavior of \coda, my tentative plan >is to define a new coda macro \pmxcoda (in pmx.tex) that handles pitch >levels the same as \shake. I would probably do away with musixtex's >automatic re-positioning, but you could still manually shift it vertically >or horizontally. Any objections? And if there are objections, i.e., if you >want to preserve musixtex's automatic repositioning of the coda symbol, then >please suggest an alternate way of writing musixtex's definition of \coda so >it doesn't hide the octave shifts from later commands. > >PS: This complication validates my decision to postpone implementing this >until after releasing version 2.7. > >--Don > > >------------------------------- >[email protected] mailing list >If you want to unsubscribe or look at the archives, go to >http://tug.org/mailman/listinfo/tex-music > >------------------------------- >[email protected] mailing list >If you want to unsubscribe or look at the archives, go to >http://tug.org/mailman/listinfo/tex-music ------------------------------- [email protected] mailing list If you want to unsubscribe or look at the archives, go to http://tug.org/mailman/listinfo/tex-music
