home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Hall of Fame
/
HallofFameCDROM.cdr
/
mus1
/
mimic11.lzh
/
MIMIC.DOC
< prev
next >
Wrap
Text File
|
1990-03-18
|
6KB
|
106 lines
MIMIC (tm) - a Digicorder (tm) utility program
(c) 1990 Eletech Electronics, Inc. All rights UNreserved.
MIMIC and Digicorder are trademarks of Eletech Electronics, Inc.
Yes! MIMIC, although copyrighted, is devoted to public domain. You can do
anything you want with MIMIC, free of charge. Source codes are included so
rebuild MIMIC if you are afraid of virus attack.
MIMIC is a utility program that reads Digicorder recorded voice file and
outputs voice data to an IBM PC/XT/AT compatible PC's internal speaker to
reproduce the recording. Due to the simpilicity of PC's speaker circuitry the
reproduced sound quality is not nearly as good as if played from Digicorder.
None the less this capability makes Digicorder an instant standard since now
every PC can play these recordings without any extra hardware!
This package contains the following files:
MIMIC.DOC - the file you are reading now
MIMIC.EXE - MIMIC executable
MIMIC.C - MIMIC source code, the Microsoft C (5.0) part
MINT.ASM - MIMIC source code, the Microsoft MASM (5.1) part
*.16 - sample recordings recorded at 16K bits/second sampling rate
*.32 - sample recordings recorded at 32K bits/second sampling rate
To play a voice file, just type
MIMIC sampling_rate filename
Here filename is the recording's file name (optionally with a search path),
and sampling_rate has to be an integer between 12 and 64. For example,
MIMIC 32 news.32
will play news.32 at 32K Hz (or bits/second.) You can play a voice file at a
sampling rate other than the one it's recorded at. The effect will be like
playing a 33 RPM album at 15 or 45 RPM. Pressing any key during the playback
will terminate the execution of this program.
MIMIC reads 32K bytes of data from the recording file at a time. If the
recording file is read from a floppy disk, a short playback distortion may
be heard whenever MIMIC reads from the floppy disk. The faster the CPU, the
less significant this distortion becomes.
MIMIC achieves CPU speed independent playback rate by re-programming and
intercepting timer 0 (clock) interrupt. It also uses (and restores) software
interrupt #255, although this interrupt can be easily replaced with another
one should conflict with TSR programs results. Note that MIMIC is not a TSR.
Although the playback rate is independent of the CPU speed, higher CPU speed
is required for high playback rate. Usually an 8MHz XT can playback at 16KHz
while playback at 32KHz requires at least an 8MHz AT. If the system is a
little bit too slow for the sampling rate, the playback tone will be slightly
lower. If the system is way too slow for the sampling rate, the playback tone
will suddenly be a lot lower due to much lower playback speed. Also, the DOS
clock will run a little bit off beat during the execution of MIMIC but the
hardware clock will not be affected at all. See MIMIC.C for more details.
The operation theory is really simple. When Digicorder records, it uses the
Continuously Variable Slope Delta modulation (CVSD) technique. This CVSD
technique, simply put, records the direction of the waveform bit by bit. For
example, if the wave goes up from the previous sampling point, a bit of "1" is
recorded, otherwise a bit of "0" is recorded. Absolute silence is recorded as
"01010101...". Basically, the higher the sampling rate, the better the sound
quality, but also the more data to store. Digicorder's sampling rate is
adjustable from 12K to 64K Hz (or bits/second.)
When Digicorder plays, the process is reversed and waveform reconstructed and
output to its own speaker, bit by bit in real time. When MIMIC plays, voice
data is sent to the PC's speaker bit by bit. These bits either charge ("1") or
discharge ("0") the capacitor in parallel with the speaker, simulating the
ups and downs of the original waveform. Thus the sound wave is reconstructed
in a crude but legible form.
The relatively high background noise is an unfortunate result of not exactly
reproducing sounds the CVSD way. Consider recording silence at 16 KHz sampling
rate, an 8 KHz tone should be heard when the recording is played (by sending
"01010101..." to speaker at 16 KHz.) Saturation/Non-linearity of the capacitor
also adds to the background noise. XT's in general have more noise than AT's,
probably due to poorer motherboard grounding (nowadays XT's motherboard
usually has only two layers. Some noise reduction is possible by pre-
processing the recording, but a way to significantly reduce the noise has not
been found. Note that these noises do not occur if Digicorder is used to play.
Digicorder, a half-length 8-bit PC/XT/AT compatible voice card complete with
microphone, speaker, tape/cassette recorder adaptor cable and utility/device
driver programs, can be purchased for $99 (Californians add 6.25% sales tax)
from:
Eletech Electronics, Inc.
1262 E. Katella Ave.
Anaheim, CA 92805
(714) 385-1707
Dealer inquiry welcomed. Eletech Electronics, Inc. is a manufacturer sales
office and also carries speech processing chips, stand alone mini speech
modules, voice telecommunications equipments and voice-mail products.
MIMIC package revisions:
1.0 - original release.
1.1 - a mistake in MIMIC.DOC about MIMIC usage is corrected.
- ANTHEM.16 is excluded to reduce file size.
- except for MIMIC.DOC, all other files remain version 1.0