Re: pixelvision concept

["Alexander the Great jr." <kanaris@bode.usc.edu>]

> 

> As has been pointed out, you'd need a deck that could play and record

> at high speed---or could be modified to do so, which is how I'm sure



You can modify a cassette recorder to run at higher speeds by changing

the "gear" ratio for the capstan motor to the capstan, which is almost

always belt driven. So you need to change the diameter of the gears 

involved.  Make sure the take up spool can keep up, although

this is probably taken care automatically if it is driven through a

clutch and gear from the capstan.



> the inventor built his prototype.  You WOULD ALSO need the

> "pixelizing" electronics that cut the signal bandwith down and create

> the unique look of the pixelvision.

>

> 

> >From the way people seem to use their pixelvisions (with a decent

> videocassette recorder), a black box "pixelizer" that would "pixelize"

> any video source would be a much more desirable device, and could

> probably be built as a homebrew project from the

> patents... hmm... does anybody have the patent #(s) for the

> pixelvision?

 

I don't think there exists a VALID patent for a concept so trivial. 

Who cares about patent numbers? I bet the patents wouldn't add anything

we don't already know, and they wouldn't hold up in court because

the idea is so trivial. Thus, in theory, it cannot be patented.



I claim the "pixelvision" look is trivial to implement digitally:

An 128x96 pixels (TV aspect ratio) frame buffer requires only

12288 bytes of memory.  We'll have to buy a 16K chip because they

like to make them in "powers of two" capacities. (3 if you want color)



An A to D converter samples the video signal at NTSC speed and stores

the signal in this frame buffer. A serializer reads from the frame

buffer at a lower rate and sends to the tape during the recording.



During playback, the data from the tape enters the same buffer at a 

low data rate, but the serializer reads out the data at a faster

(i.e. NTSC) rate and sends to the D to A converter. If you get 

a 16 KB dual ported SRAM chip, the A/D and the D/A, you can build

this at home on a breadboard over a weekend or two.



I hope people with time in their hands will implement something

based on the above guidelines and post the results here.



I am sorry but I cannot offer schematics, parts selection and

design advice because it would take me some time to put these things

together and right now I have none. 









    <<alex kanaris>>  



==============================================================================

Alexander Kanaris                           ___   ___     kanaris@bode.usc.edu

Electrical Engineering --- Systems   /  /  /__   /      kanaris@thales.usc.edu

University of  Southern California  /__/  ___/  /__     kanaris@aludra.usc.edu

Los Angeles, California 90089-2562                  kanaris@alumni.caltech.edu

==============================================================================