Other Stuff
In September 1990 I made my first exploration into the world of commercial
software. I wrote a drum sequencer application (called Percussaman) for the
Atari ST, and it was published in Atari ST User (for which I received the then
princely sum of £150..!). I wrote it in pure 68000 assembly language whilst I
was doing my GCSEs.
Following percussaman, during my A levels, I developed the application into a
full music sequencer called Melodiman (very similar to a modtracker program).
For this version, I replaced the Atari's OS completely - the first thing that
Melodiman did when loaded was to download itself over the OS workspace. The
program was completely stand alone, having it's own windowing system and disk
operating system, and recovered quite a bit of memory this way. I built a
hardware cartridge for Melodiman to improve the sound quality (and reduce the
CPU overhead). This cartridge output the four channels separately (for external
mixing), and also a premixed stereo signal.
You can get the schematic for the Melodiman cartridge by clicking on the image
to the left. You can download the entire source for Melodiman from the
downloads page. It's pure 68000 ASM, and pretty big. The main editor only makes
two calls to the OS before replacing it, one to switch to supervisor mode, and
the other to print a string if you have a mono monitor.
Then I moved onto PCs at University. I designed and built a four channel sound
card in my second year at university for my Amstrad 2086 (8 Mhz 8086). This was
very similar in design to the sound cartridge for the Atari, but interfaced to
the ISA slot in the PC instead.
Next I moved onto video. I designed and built a video digitiser which would
capture an image in one frame to some onboard memory and then allow a download
over a period of time to the slow host PC. Examples of the images captured are
below - it did quite a good job, but was brute force, with a large amount of
static RAM on board to hold the image. All of the control logic was 74 series
TTL. The empty sockets are from where I was planning to extend the digitiser to
full colour. I had previously experimented with systems which captured an image
over a number of frames, capturing vertical stripes which moved across the
image each frame. This meant I could capture directly to the computer, because
the data rate was lower, but the smearing when capturing from live moving video
was not desirable.
In September 1990 I made my first exploration into the world of commercial
software. I wrote a drum sequencer application (called Percussaman) for the
Atari ST, and it was published in Atari ST User (for which I received the then
princely sum of £150..!). I wrote it in pure 68000 assembly language whilst I
was doing my GCSEs.
Following percussaman, during my A levels, I developed the application into a
full music sequencer called Melodiman (very similar to a modtracker program).
For this version, I replaced the Atari's OS completely - the first thing that
Melodiman did when loaded was to download itself over the OS workspace. The
program was completely stand alone, having it's own windowing system and disk
operating system, and recovered quite a bit of memory this way. I built a
hardware cartridge for Melodiman to improve the sound quality (and reduce the
CPU overhead). This cartridge output the four channels separately (for external
mixing), and also a premixed stereo signal.
You can get the schematic for the Melodiman cartridge by clicking on the image
to the left. You can download the entire source for Melodiman from the
downloads page. It's pure 68000 ASM, and pretty big. The main editor only makes
two calls to the OS before replacing it, one to switch to supervisor mode, and
the other to print a string if you have a mono monitor.
Then I moved onto PCs at University. I designed and built a four channel sound
card in my second year at university for my Amstrad 2086 (8 Mhz 8086). This was
very similar in design to the sound cartridge for the Atari, but interfaced to
the ISA slot in the PC instead.
Next I moved onto video. I designed and built a video digitiser which would
capture an image in one frame to some onboard memory and then allow a download
over a period of time to the slow host PC. Examples of the images captured are
below - it did quite a good job, but was brute force, with a large amount of
static RAM on board to hold the image. All of the control logic was 74 series
TTL. The empty sockets are from where I was planning to extend the digitiser to
full colour. I had previously experimented with systems which captured an image
over a number of frames, capturing vertical stripes which moved across the
image each frame. This meant I could capture directly to the computer, because
the data rate was lower, but the smearing when capturing from live moving video
was not desirable.
