The IBM Multimedia collaboration
UZnal
In 1972 ... Dr. Martin Alpert starts Tecmar Scientific Solutions in
Cleveland Ohio,
the name "Tecmar" being derived from MARty's TEChnology....In the history of
data
acquisition, Tecmar and Scientific Solutions are actually the same company.
Tecmar and Scientific Solutions - IBM MultiMedia collaboration
In 1986, IBM approached Tecmar to design and manufacture a number of
multimedia
products. These products were to be designed and manufactured by Tecmar and
marketed by IBM. Scientific Solutions provided the engineering and design
talent, and Tecmar provided the manufacturing and test talent. The result
of this partnership was the release in 1988 of IBM's Audio Visual Connection
(AVC). This breakthru product consisted of a sound record and playback card
(IBM Audio Capture and Playback Adapter - ACPA), a video digitizing card
(The IBM Video Capture Adapter /A - VCA/A) and the Audio Visual Connection
authoring software.
The ACPA featured CD quality 16-bit digital audio recording at 44.1Khz
stereo or 88.2Khz mono, 16-bit stereo playback with 2x oversampling,
and real-time DSP hardware based compression/decompression (if desired).
The card also features the ability to download algorithms for the on
board 10 MIPS digital signal processor (TMS320C25). Typical downloadable
algorithms provide for interpolation or decimation filters to effectively
provide different sample rates and MPEG/JPEG hardware assisted image
decompression. Incidentally, the 16-bit stereo ACPA sound card with the
10 MIPS C25 DSP was released several months before the 8-bit mono
SoundBlaster product which utilized an 8051 microcontroller instead
of a true DSP. The Audio Visual Connection product was awarded the
winner in the hardware category at the 6th Annual PC Magazine Technical
Excellence awards held in 1989.
A sound card uses Analog-to-Digital converters, Digital-to-Analog
converters, amplifiers and other technologies that are common place
in data acquisition products. So it was natural to utilize the
experience and expertise of the Scientific Solutions engineers
to design the hardware which was a breakthru in quality digital
audio for the PC and a forerunner to the digital audio MP3 craze.
The same team approach was used to develop the IBM M-Motion Video Adapter/A,
(March 1990), which allowed for the viewing of full-motion video on the
computer screen, graphic overlay and full video scaling. Like the Audio
Visual Connection, this product was marketed by IBM under the IBM brand
name.
And again, the collaboration resulted in the PS/2 TV (March 1992),
PS/1 TV (1993) and PCTV (1994) products each which are external units
containing an analog TV/Cable tuner that allows for display, scaling
and freeze-frame capture of video on a computer screen. The PS/2 TV
was marketed by IBM. The PS1 and PCTV were marketed directly by Tecmar.
........
In 1987 Scientific Solutions introduced the world's fist data acquisition
products for the then new IBM MicroChannel architecture. The MCDAS family
of products (MCDAS 1612, MCDAS 1614, MCDAS 1616) provided 12, 14 or 16-bit
resolution and high speed BusMastering data transfer. The MC-IEEE provided
a fully compliant IEEE-488 GPIB interface to the IBM PS/2 MicroChannel
architecture.
http://www.scientificsolutions.com/products/ieee488/ieee488_index.html
http://www.scientificsolutions.com/products/ieee488/ieee488_spec.html
exwisdem
I have read this many times. They are really good people. They helped
me
finish the MCDAS-12/16 driver for NT, and they were helpful with the
GPIB
driver I am still working on.
exwisdem
William R. Walsh
> The card also features the ability to download algorithms for the on
> board 10 MIPS digital signal processor (TMS320C25).
Well, that *is* interesting. I had no idea it was that flexible. This does
suggest that Mwave and Audiovation were not quite the "big leaps forward"
that I thought. (Then again, I never have seen the M-ACPA complain that its
DSP was "out of memory".)
> Typical downloadable algorithms provide for interpolation or decimation
> filters to effectively provide different sample rates and MPEG/JPEG
> hardware assisted image decompression.
Wow...MPEG/JPEG decompression done on the TMS320C25...again, no idea that it
was so flexible.
The Audiovation and Mwave do have software included in the driver package to
let the onboard DSP decompress JPEG images at least. I've never seen
anything similar for the TMS320C25. Maybe I should take another look at that
book...
> Incidentally, the 16-bit stereo ACPA sound card with the 10 MIPS C25 DSP
> was released several months before the 8-bit mono SoundBlaster product
> which utilized an 8051 microcontroller instead of a true DSP.
Although it was the Sound Blaster that put sound into computers (and defined
a standard in the process), the M-ACPA card looks to have been a lot more
elegantly designed than the MCA Sound Blaster. I do suppose the Sound
Blaster was much cheaper...
I've heard of another sound product from IBM--it's mentioned in the Ardent
Tools--that was known as the Audio Recording and Playback Adapter/A. What
was that, and how did it compare (if at all) to M-ACPA and later
Audiovation?
William
brad.p...@gmail.com
<newsgrou...@idontwantjunqueemail.walshcomptech.com> wrote:
> Hi!
> > The card also features the ability to download algorithms for the on
> > board 10 MIPS digital signal processor (TMS320C25).
> Well, that *is* interesting. I had no idea it was that flexible. This does
> suggest that Mwave and Audiovation were not quite the "big leaps forward"
> that I thought. (Then again, I never have seen the M-ACPA complain that its
> DSP was "out of memory".)
Yes, I think I said something to that effect a while back. The thing
to note is that Creative Labs was years behind IBM in the use of
general purpose DSPs. SoundBlaster Live! is more like Mwave than the
orignal SoundBlasters in overall architecture.
Nothing unusual for IBM, M-ACPA to Mwave was an evolutionary step, not
a revolutionary one which is the usual IBM path.
> > Typical downloadable algorithms provide for interpolation or decimation
> > filters to effectively provide different sample rates and MPEG/JPEG
> > hardware assisted image decompression.
> Wow...MPEG/JPEG decompression done on the TMS320C25...again, no idea that it
> was so flexible.
That's what DSPs do :-)
> The Audiovation and Mwave do have software included in the driver package to
> let the onboard DSP decompress JPEG images at least. I've never seen
> anything similar for the TMS320C25. Maybe I should take another look at that
> book...
> > Incidentally, the 16-bit stereo ACPA sound card with the 10 MIPS C25 DSP
> > was released several months before the 8-bit mono SoundBlaster product
> > which utilized an 8051 microcontroller instead of a true DSP.
> Although it was the Sound Blaster that put sound into computers (and defined
> a standard in the process), the M-ACPA card looks to have been a lot more
> elegantly designed than the MCA Sound Blaster. I do suppose the Sound
> Blaster was much cheaper...
What was it, divide by 2.5? Somebody help me out here, I don't have
MSRPs for them at hand.
> I've heard of another sound product from IBM--it's mentioned in the Ardent
> Tools--that was known as the Audio Recording and Playback Adapter/A. What
> was that, and how did it compare (if at all) to M-ACPA and later
> Audiovation?
Functionally it was virtually identical to the M-ACPA. The main
differences were that the form factor was full length and the audio
stages were single supply powered. The M-ACPA audio stages have +/-
voltages on them which yields more headroom and probably better slew
rates.
The full source code for the original ACPA driver is available.
Looking through it, there is an incestuous relationship w/ the M-
Motion video capture card.
Dave Weis modified the ACPA code to run under the Linux 2.0 kernel. It
compiles and works as a playback only SB emulation driver.
I think something like the M-Wave hack for Linux would work with the
ACPA family as well. Boot DOS and load the original DOS ACPA drivers
to initialize the card as a SB Pro, then use a bootloader from DOS to
launch another O/S like Linux (Winnt?) and the M-ACPA should still
look like a SB Pro to the booting O/S. Use the SB Pro drivers from the
second O/S to get the card to work.
Brad
> William
brad.p...@gmail.com
I actually talked to Bridget Ritthaler about something when I was
building Dave Weis' ACPA Linux driver. I saw her name in the code and
looked her up when I needed some info about the driver. I didn't know
at the time that she was listed as an inventor on the patent! I said
something about this to Louis and I think he wanted her number so he
could ask her out on a date (thinly disguised intelligence
operation?). I have since then, 1998, lost her email address and
number.
Brad
On Sep 20, 12:16 am, "William R. Walsh"
<newsgrou...@idontwantjunqueemail.walshcomptech.com> wrote:
UZnal
> Looking through it, there is an incestuous relationship w/ the M-
> Motion video capture card.
Indeed, there are some video definitions lines and I assumed them as being a
part of the standard driver definitions. Without the ACPA technical
reference it will not be possible at all to understand, and hence, to modify
and adapt the code.
> Dave Weis modified the ACPA code to run under the Linux 2.0 kernel. It
> compiles and works as a playback only SB emulation driver.
The code is well documented and the general procedure not complicated at
all.