home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Club Amiga de Montreal - CAM
/
CAM_CD_1.iso
/
files
/
497a.lha
/
ComSMUS_v2.2
/
voicebank.prog
/
README
< prev
next >
Wrap
Text File
|
1991-04-07
|
4KB
|
121 lines
IFF VOCE format
---------------
The VOCE format was created by Hackercorp as a means of addressing certain
problems with the existing IFF 8SVX sample format which prevented the
Amiga from achieving the maximum fidelity of which it was capable.
One goal in creating the VOCE format was to maintain compatibility with
existing samplers and sampling software. The result is that the VOCE
format uses standard IFF 8SVX files for audio samples; data contained
within the VOCE structure assigns 8SVX files to arbitrary ranges of
notes and note velocities within a VOCE instrument definition.
Currently, the Hackercorp Commercial SMUS Player supports this format.
Code in development at Hackercorp is expected to provide VOCE format
support within a MIDI environment.
This document describes the VOCE file format, plus it describes the
operation of the 'voiceiff' program, a currently rather user-unfriendly
(but functional) program that can be used to create VOCE files.
VOCE format description
-----------------------
The VOCE format has many similarities to other IFF files. A VOCE file
starts with a 'FORM VOCE'
FORM VOCE
Next come some optional text chunks. These behave exactly like text
chunks in the other formats. The currently supported text chunks are:
NAME
(C)
ANNO
The chunk that actually does all the work is the VDAT chunk.
VDAT first_node last_note reference_note low_velocity high_velocity samplename
Setting up an voiceiff configuration file
-----------------------------------------
We're in a hurry, so let's learn by doing.
Here's the config file for creating the piano voicebank. It's called
piano.input, and the real thing should reside in the same directory in
which you found this documentation.
30 51 43 1 127 piano.g1
52 59 52 1 127 piano.e2
60 68 60 1 127 piano.c3
69 76 69 1 127 piano.a3
76 83 79 1 127 piano.g4
84 90 84 1 127 piano.c5
91 99 91 1 127 piano.g6
100 110 103 1 127 piano.g7
piano.g1 through piano.g7 are the names of the eight IFF 8SVX files that
are used by the piano voicebank to create the piano sound.
The piano.g1 voicebank covers MIDI note numbers 30 through 51 (inclusive),
the base sound of the note (pitch the note plays when played at the sample
rate defined within the piano.g1 IFF 8SVX file is MIDI note 43) Note that
the same MIDI note ranges are used within SMUS files.
The 1 and 127 specify the range of MIDI velocity for which piano.g1 is to
sound. 1 and 127 represent the full range. A velocity split can be done
by having samples overlap the same key regions but cover different velocity
ranges. SMUS volume commands (ppp,pp,p,f,ff,fff,etc) are scaled to MIDI
velocity ranges, and are in effect for all notes on the track on which the
volume command was issued until a new volume command is issued.
Here's the config file for the madonna bass voicebank. It's called
mdonabass.input.
20 44 36 1 127 mdona.bass.lo
45 56 48 1 127 mdona.bass.me
57 68 60 1 127 mdona.bass.hi
In this voicebank, there are just three 8SVX sample files. Again, there
is no velocity split. This voicebank is a punchy alternative to some
of the existing bass voices out there.
Creating the voicebank file
---------------------------
To create the piano voicebank, execute:
voiceiff <piano.input piano
This will create an IFF VOCE file, piano, that the SMUS server can recognize
and load, resulting in the loading of a number of 8SVX files and attaching
them to one "virtual keyboard."
Epilogue
--------
OK, well, not much of a manual -- it was done in a rush. I hope you can
use it if you need to, and if you have a problem that you can't seem to
figure out, give me a call.
-karl
Postscript
----------
Feel free to alter the voiceiff source to change the output of the voiceiff
copyright and author chunks to be correct for the voicebanks you create.
I will streamline this through the config file in a subsequent version.