MidiSyn history
This page includes the program history,
describing the major modifications introduced in each release and the bugs
corrected.
New features
An automatic
installation procedure has been added.
There is
now a new effects unit, Delay, which can be used to implement not only Delay
itself but some other popular effects like Echo, Phasing, Flanging and even
Chorus.
Added
Karaoke file processing. Now .KAR files can be processed in the same way as
.MID files.
Added
compressed SoundFont processing. This permits the use of huge SoundFonts.
Improvements
Dramatic
improvements in overall sound quality.
Major
overhaul of the Chorus unit.
Sliders are
now used in the Options Dialog, instead of numerical values to set some values.
This allows a more intuitive use of the features affected.
Some speed
improvements.
Major
revision of the documentation.
Bugs corrected
The home
directory was incorrectly set: If the working directory specified in the
shortcut pointed to anything other than the directory where the executable was,
MidiSyn would loose the location of the Initialisation and Help files.
MidiSyn has
a feature that automatically remaps a bank/preset pair to a valid one, when the
bank/preset requested by the MIDI file doesn't exist in the current SoundFont.
This feature had a bug that in some cases would cause a remap when it was not
necessary.
In batch
mode do not show the message: "To obtain better dynamics in your WAVE file
you should go to OPTIONS, increase the Master Volume by n dB and try
again" as this interrupts the normal flow of processing.
Disable
bank change, Output change and Midi load while converting, Output change while
playing. This would cause the program to generate an exception and terminate.
Corrected a
minor bug in the error processing while reading a Sound Bank.
Corrected a
bug in the parse of a sample header while reading a Sound Bank. This was not
serious as it corrected itself.
Corrected
potentially serious in the sound generating procedure, in relation to the
addressing of sample data.
Corrected
bug - Off limits addressing of sample data. This was due to sloppy programming
of SoundFont banks. Strangely enough, this was found to afflict several SoundFonts
circulating around the Net.
Corrected a
bug in the processing of stereo samples. The effect of this bug was an increase
in the rendering time.
Corrected a
bug that would cause the program to hang while reading a MIDI file with errors.
Corrected a
bug in the parse of a MIDI file that would erroneously reject a file as
unsound.
New features
MidiSyn
now processes drag-and-drop of files. This can be applied to all sound file
types processed by the application: Sound Font, WAVE and MIDI files.
Added
automatic rescaling of the output file for better dynamics. After the
conversion process, the volume is adjusted to the maximum value.
Improvements
Better
sound quality due to changes in the interpolation algorithm and in the volume
envelope.
New features
MidiSyn
includes now a MIDI mixer, where you can adjust the relative volume of each
instrument in a MIDI file. You can also completely remove an instrument or
single out an instrument in the WAVE file.
Improvements
There
is a new implementation of the digital filter. Although the filter used in
version 1.5 worked well, the filter specified in the SoundFont manual has a
12dB/Oct slope. The filter in the 1.5 version had 24dB/Oct slope.
Several
changes have been made to improve speed. The program is now 60% faster, without
adversely affecting the sound quality.
New features
Pitch-bend
sensitivity processing has been included. MidiSyn is now fully compatible with
the General MIDI specification. Pitch-bend sensitivity is changed via the
standard RPN parameters.
The
output-processing block has undertaken a major revision. It is now possible to
control the overall gain externally (in the Options dialog). There is now a
compression facility, which is also controlled by a parameter in the Options
dialog.
Bugs corrected
Corrected
an ugly bug in pitch-bend processing.
Corrected
a bug that prevented in some cases the synthesis process to conclude.
Corrected
a couple of minor bugs in the batch facility.
This version of the program includes a
reworking of all the major sound processing blocks. This will hopefully
translate into a great improvement in sound quality. This version also corrects
a minor bug found in the previous version. It includes also a couple of new
features that you may find useful.
Improvements
Includes a
more precise linear interpolator, with the possibility of using a band-limited
interpolator.
Includes a
sound sample cache mechanism, which permits the use of SoundFont files that do
not fit in memory. The only restriction now is that a single sample must fit in
memory.
Includes
a new reverberator, which produces a much better sound. This version includes a
new configuration group in the OPTIONS window where the user can change some of
the reverberator's parameters (See below).
Has a new
filter, a digital implementation of the famous Moog VCF filter: 24 dB/oct
resonant lowpass.
Includes a
series of modifications that implement several of the "GM Level 1
Developer Guidelines" published by the MMA. These affect mainly the
processing of Volume/Expression and the response to Reset All Controllers.
New features
It
is possible now to change the default file/path names for the SoundBank and the
output files. This information is stored as a .INI file in the same folder as
the executable.
Includes a
transposition facility, whereby the user can shift the pitch of the notes in a
MIDI file. This applies to all the channels except channel 10, used for
percussion. The transposition can be up- or downward and can consist of any
number of semitones
There is a
new batch facility, whereby the user can convert a set of MIDI files without
having to manually set the parameters for each conversion. The instructions for
the batch go into a text file that the user can edit with a text editor like
Notepad. A line in the batch file describes each job in the batch. Each line
has three fields corresponding to: 1) The SoundFont to be used; 2) The output
file where to put the result of the conversion and 3) The MIDI file to convert.
The fields, which besides the file name can include the partial or full path,
are separated by one or more spaces. If the file name or path used in a field
have embedded spaces, enclose the file name/path between .
There is a
new configuration group in the OPTIONS window where the user can change the
amount of chorus applied. This field has the caption Wet adjust and allows you to specify a value to be added to the
amount of chorus specified by a CC93 MIDI command and to the value specified in
the SoundFont for the samples you are using. The value can be negative, so that
you can actually subtract from the value specified by the CC93 command or
SoundFont. The drawback is that the value you specify in the OPTIONS window
affects all the channels. The total amount of chorus (the value you specify in
the OPTIONS window plus the value in the CC93 MIDI command plus the value
specified in the SoundFont) must remain in the 0-100% range. If the result is
outside this range the sum is limited automatically by the program. For
instance to completely remove any chorus effect that is set in your MIDI or
SoundFont files, you put a value of 100 in the OPTIONS window.
There is a
new configuration group in the OPTIONS window where the user can change some of
the reverberator's parameters. These are:
Ļ
Room size This controls the simulation of
the room ambience. It can range from 0 for a small room to 100 for a cathedral.
The program automatically truncates values outside this range.
Ļ
Damping factor This value determines how fast a
given sound will die away. A value of 0 corresponds to the maximum time the
sound will last.
Ļ
Wet adjust This works in the same way as the
Wet adjust parameter for the chorus
effect, described above. Here we are talking about the amount of sound that
passes through the reverberation effects processor. The MIDI controller for
reverberation is CC91.
Bugs corrected
Corrects a
minor bug that occurs while reading SoundFonts generated with AWAVE Studio.
MidiSyn was trying to access the last record in the preset list. Vienna sets
the Bank/Preset to 0xFF, AWStudio sets them to 0.
This is the first released version of the
program.
Comments, suggestions and bug reports are welcome and should be sent to fadevelop@clix.pt
This page last modified 2002-12-14 - Copyright
Đ 2000-2003 ACE