Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!samsung!uunet!ogicse!milton!allyn
From: allyn@milton.u.washington.edu (Allyn Weaks)
Newsgroups: comp.music
Subject: Re: Notation programs for composers, not publishers
Message-ID: <14344@milton.u.washington.edu>
Date: 14 Jan 91 05:23:53 GMT
References: <9101101019.AA05528@oda.icl.stc.co.uk>
Organization: University of Washington, Seattle
Lines: 104

andy@oda.icl.stc.co.uk (Andy Spiceley) writes:
>Most reviews and comments I have seen relate to the ability of the various
>programs to typeset music, that is, to act as a DTP system for music. Few
>have addressed the problem of using these programs as music processors in the
>sense of (analagous to) _word_ processors: that is, programs which help the
>composer sketch out, edit, amend, expand etc, and _then_ format for printing.

Caveat: My experience is all on the Mac, so I can't speak for the messy-dos or
any other world.  Also, I'm not a composer, but I've been helping one
computerize, so I've heard a fair amount lately of wishes and 'what do you
mean it can't...!'.  

On the Mac, all of the 'professional' level notation programs are aimed at the
publication market. But the toy programs sell on 'easy to get something that
looks kind of like real music' and have adequate user interfaces to think
into.  The problem, of course, is lack of features - no double dots, 32nd is
the smallest note, triplets are the only tuplet, maybe a limited number of
voices, either altogether, or on one staff.  Then there's Encore, which gives
you all the editing pain of the typesetters, with the feature list of the
toys...  Music Prose (a subset of Finale) is one I haven't seen yet - I hear
that the interface is much better, but I suspect it has many of the same
underlying problems, since it's based on the same engine.

The two toy programs I'm thinking of are Deluxe Music Construction Set (DMCS)
and Concertware + MIDI 5.0 (CW).  They're about the same price (roughly $90 at
the discounters last time I looked).  Each is better at different things.  A
brief list of the features that are likely to matter for 'composing into':


Both:

)  cut, copy, paste contiguous regions, within a file or between files,
   either a single voice, or several voices
)  change duration for a selected region to half, double, 2/3, 3/2
)  transpose a selection up n half steps
)  play back through the Mac speaker (4 voices at a time) or through MIDI
)  bugs and quirks  (but mostly not fatal)
)  lyrics of a sort

DMCS:  
)  measure based.  Doesn't rebar automatically, though you can merge and
   split measures to get things to come out right.
)  You can only change play-back tempo (and lots of other things) at the
   beginning of a bar, so you can't put in accelerandos or ritards.
)  Good mouse note entry (click on a duration, click where you want the
   note to go) but bad MIDI keyboard note entry (no real-time, and the
   step-time is really slow and clunky.)
)  Can't write MIDI files to transfer data to another program.  (There's a
   companion sequencer that will read and write DMCS files, and the sequencer
   will read and write MIDI files.)
)  You can transpose a selection either by number of half steps, or
   diatonically.  
)  Only two voices per staff.  makes it almost impossible to 
   do a lot of keyboard music.
)  up to 40 staves.
)  decent print control - you could probably get an adequate performance
   score from it, if what you write isn't complicated, and if you touch
   it up (add ornaments, etc.) by hand.

Concertware:

)  Not measure based - rebars automagically as you enter stuff, or when
   you change time signatures, or durations.
)  Poor mouse note entry (click on duration, click on a note pallete
   that's almost impossible to read)  Good Mac keyboard entry, excellent
   MIDI keyboard entry - both real time (one voice at a time, limited but
   probably adequate quantization options) and efficient step-time.
)  Only eight voices total.  Any number can be on one staff.
)  reads and writes MIDI files (at least 5.1.4 does...)
)  limited automatic beaming
)  rotten print control - there's no way to cram more measures on a line
   than it wants to, which means you end up with lots of white space and
   many page turns.  (About 3 times as many pages as is reasonable.)


>Of course, there remains the possibility that a usable composers music
>processor has yet to be written.  Any volunteers?  I'd be very happy to
supply >a list of "must haves"!

From time to time I work on a specification for a decent program for composing
into and printing out, and it's fun because it does have such nice thorny
problems.  But I'll probably never write it, since it's the design I enjoy,
not the programming (and I couldn't afford to support myself for the 3-5 years
it would take).  

It might help if more people would speak up and prominently refuse to buy
things - i.e., spend lots of time talking with a company like Finale or
whoever, and explain just why it is you aren't going to buy their program,
rather than quietly deciding not to.  If they _know_ that they've lost 50
sales because the editing is rotten, they might be more inclined to fix it.

>It occurs to me - as it must have to many others - that we badly need
>a standard for the internal representation of music events that would
>enable the interchange of scores between different programs for editing,
>layout/printing, etc. 

There is a standard in the works.  Every now and then someone posts
information and the address of the committee.  With luck it will be approved
in a year or so, and  start being used shortly thereafter.  Most of the music
publishing software houses are keeping an eye on it and plan to incorporate it
when it's stable.

Allyn Weaks
allyn@milton.u.washington.edu