Robert Rolf ualberta.ca> writes: > OK, I looked. I didn't see any URLs for this 'scanner' based software. > Am I missing a post? I haven't posted it yet. I saw it somewhere years ago (1990s). I'll find it and post it later. > A lot of the paper tape in my junk pile is BLACK, so black on black > gives lousy contrast. (PDP 11 & PDP 8 diagnostics tapes). > It also would be difficult to correctly light aluminum (metal) tape > (used for bootloaders and other high repeated use stuff). So one uses a white background for black tapes. > Backlighting works well if you set the camera gain manually. > With correct lighting intensity you get nearly 100% modulation > of the video signal, making it easy to discriminate. There ar no controls on the cheap kind of webcam I have in mind for this. > And you wouldn't need a PC to do it. > Just pull the tape horizontally, and have the PIC A/D sample a few > microseconds into the scan line for whatever scan lines have > the holes. Use the sprocket hole as your clock for 'data valid'. > Or pull the tape vertically and dissect a single scan line into > the bits. Could even be done with hardware shift register and > comparator + PLL. Just need to have a stable optical setup. You know the golden rule: cheap/affordable, good/works, now, pick any one. > How about a 'Rube Goldberg' style virtual design contest to come > up with the most unusual way to read a paper tape? Doh. Xylophone driven by holes and software running on soundcard (hey, one could use the little springs from musi boxes). Anyway Rube Goldberg and crc-less records of any length (longer than a few lines) are a no-no. Just for laughs compute the required error rate to read just 1 kilobyte of data from a paper tape. That's 1024*3.5 holes to get right. Or the error rate must be < 2^-13 or so to read it right 50% of the time, and it gets worse as the length increases. Does not look good for 'Rube Goldberg' systems. Peter P. -- http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist