home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Music Workshop 2
/
Music_Workshop_Second_Edition_Walnut_Creek_January_1996.iso
/
extras
/
patches
/
tx81z
/
format.doc
< prev
next >
Wrap
Text File
|
1994-02-17
|
6KB
|
142 lines
Yamaha TX81Z Archive Format
---------------------------
Overall Structure
-----------------
The archive is a single ASCII text file, containing voice descriptions
and the ASCII encoded binary which implements the voice. If the file
gets too big, it can be split into several smaller files (if you don't
split up a voice), but they can still logically be thought of as a
"single archive file".
Each "voice" in the archive is separated from other voices by one or
more blank lines. There are no blank lines within a voice.
Individual Voice Structure
--------------------------
A voice consists of one or more comments, followed by the voice data.
A comment begins with an asterisk ('*') in column 1. Forgiving reader
programs would treat any characters after an asterisk and before a
newline as comment material.
Some comments have special formats, to allow automated querys of the
archive. These comments begin with two asterisks ("**"), and are
described in "Database Comments" below. Comments which are not intended
as database comments must only start with a single asterisk.
The voice data consists of 128 bytes, in the order and format of a
system bulk voice dump. These 128 bytes have been encoded in hex ASCII,
for ease in manipulation and transport. Each byte is separated from
it's neighbors by white space (space, tab, or newline). By convention,
there are 5 lines of 25 bytes, and 1 line of 3 bytes, but forgiving
reader programs should accept any layout arrangement.
Database Comments
-----------------
Certain comments in the database are in a particular format, to allow
automated queries of the voice database.
The format of the first part of these comments, from the first asterisk
to the terminating colon must be exactly as specified. All have two
asterisks, one space, a field name, and a following colon.
The database comments could occur in any order, but the order they are
defined here is suggested. In addition, experience has shown that it is
very useful for the "** Name:" comment to occur first. The following
of this convention makes writing tools to manipulate the archive much
easier.
The following are defined:
** Name:
This is the voice name, as it would appear on the display of the
synthesizer. Each voice must have one of these, since it is also in
the binary data. This comment must fit on one line. It is strongly
recommended that this comment be the first line of a new voice.
** Long_name:
If a person wants a longer, more descriptive name, this is where it
goes. Whenever available, reader software should use this name when
interacting with the user. This comment is optional, and must fit
on one line.
** Description:
This is a wordy description of the sound of the voice. It might
also include what it's good for. This comment is optional. It can
continue over more than one line. (See comment continuation below.)
** Controls:
This lists the external controls which affect the sound of the voice.
The controls and their meanings are:
KV Key velocity changes the gain of the operator
envelope
MW (LFO pitch) Modulation wheel changes the effect of the LFO
on the voice pitch
MW (LFO amplitude) Modulation wheel changes the effect of the LFO
on the voice amplitude
BC (LFO pitch) Breath controller changes the effect of the LFO
on the voice pitch
BC (LFO amplitude) Breath controller changes the effect of the LFO
on the voice amplitude
BC (EG amplitude) Breath controller changes the gain of the
operator envelope
BC (voice pitch) Breath controller changes the pitch of the voice
FC (LFO pitch) Foot controller changes the effect of the LFO on
the voice pitch
FC (LFO amplitude) Foot controller changes the effect of the LFO on
the voice amplitude
The overall volume and pitch wheel settings will also affect the
sound of the voice, but are not noted in this comment.
This comment is optional. It can continue over more than one line.
(See comment continuation below.)
** Usage:
This is helpful hints on how to use the voice, not so much in terms
of musical arrangement, but more from the standpoint of "breath
controller will make the voice sound more whatever". It is helpful
to know how the voice reacts to the various control sources, without
having to dig into its internals to see how it is wired. Another
good thing to include here is reasonable voice range. Many voices
only sound good over a limited range of notes. This comment is
optional. It can continue over more than one line. (See comment
continuation below.)
** Source:
This is where the voice came from. It's an opportunity to give
credit where credit is due. If this voice is a modification of
another, it would also be nice to credit or at least reference the
original source. This comment is optional. It can continue over
more than one line. (See comment continuation below.)
The above database comments are all that are defined at the present.
Reader programs should ignore (or do something reasonable) with database
comments which they don't understand. Obviously, since almost all are
optional, they shouldn't depend on any voice having these fields.
Some database comments can extend over more than one line. The
extending lines must be successive, must begin with two asterisks, and
must have something other than a single space character after the
asterisks. This allows reader programs to distinguish between a
continuation (more than one space) and a new database comment (exactly
one space). An example is:
** Usage: This voice works only if you get the breath controller max'ed
** out, and are playing note C5. Otherwise it causes your
** speakers to blow up.