| Hawes Mechanical Television Archive | by James T. Hawes, AA9DT |
|
Homebrew TV Color Corrector |
||||||
|---|---|---|---|---|---|---|
Video pinkeyeDid your TV set ever get pinkeye? This is a video nightmare where television becomes magenta-vision. In this odd condition, all color values turn pinkish-purple. Fortunately, the malady is usually short-lived. Better yet, the local TV set is usually okay. The program material, an old color film, is at fault.
On late-night TV, we've all seen color films that are past their prime. You can tell because they lose their greens and blues. What remains is one-third of a movie, all in magenta. Otherwise entertaining Vincent Price and Christopher Lee horror films come to mind. Ever see the pinkeye version of The Raven or The Brides of Dracula? Nobody seems to care about restoring the images. Meanwhile, high-profile films such as The Wizard of Oz undergo complete digital refurbishment. For years, I've wondered if a simple, analog process could inexpensively restore missing colors. I've worked with analog proc amps, but never thought about their film-restoration possibilities. Until now. Enter Cliff Benham, a video engineer with much color TV experience. Cliff told me that back before the day of digital enhancement, he actually built an analog color corrector. Below is an example of what Cliff's colnverter can achieve. Under the video is Cliff's story of his converter project...
Before & after videoCliff: The video on this page is a short segment from a faded, Florida tourist-attraction film. In the 1980s, I used my homemade color corrector to transfer a group of such films to video. The video includes 30 seconds of "before" and 30 seconds of "after" shots. I made the transfer and the color correction in 1982. James: The breadboard nature of your project actually makes it appealing. That's an example of what I'd call techno-heroism. Do you have a photo of the corrector unit? Cliff: No, I don't have a photo. The corrector was a one-of-a-kind project. I needed to fix some faded public service films for a client. I threw the project together. The object was an easy project that was as cheap as possible. By the way, this page is the datasheet from a National chip LM1202. The datasheet provides a schematic for a proc like the one that I used. The problemJames: You started from film, but how did you transfer the film to video? Cliff: The film system was a standard film telecine. All TV stations of that era had such telecines. The telecine consisted of a 16mm film projector with a fast pulldown and a five-blade shutter. The projector shot into the lens of a three-tube, color camera. The camera design favored film transfer. James: I remember those old telecines. Cliff: Our telecine camera had a very limited range of color and gain adjustments. After transfer through the camera, films only looked as good as the original film. Faded films looked very magenta. The solutionJames: Didn't you say that you used a proc amp from a TV camera? Cliff: The proc that I used was a board from a studio TV color camera. The proc was a three-channel device. The user could adjust red, green and blue video levels from the camera tubes. For each color, the proc had a black level control and a gain control. That's six controls in all. The proc board had no sync or burst processing. You'd find sync and burst controls on procs that clean up NTSC composite video. James: Since you don't have a picture, would you please describe the project for me? Cliff: Sure. This wasn't a nice, finished project with a case and a proper inside layout. Just stuff that I hooked up on a tabletop for about a month. The setup consisted of the proc board with wires that I tack-soldered to it. For the project, I cobbled together a power supply. I used the board's screwdriver adjustment pots directly. I never bothered to install external pots with knobs. Luckily, the circuit board held up for the duration of the project. James: You mean that you achieved this improvement by adjusting the RGB levels? Cliff: Yes. With my color corrector, I rebalanced the film chain video so that it looked better. For each color, I could adjust the gain or white level and the black or blanking level. Input signalJames: Was the input video RGB? Cliff: Yes. RGB right from the filmchain camera. James: Did you ever have to decode composite video? Cliff: No, for this project, I never tried an NTSC decoder to make RGB. In the late 1970s, the quality of NTSC bandpass decoders was pretty bad. At that time, there were no practical, low cost, three delay line comb filters. Even expensive decoders looked fuzzy. Output signalJames: What did your corrector put out to the recorder? Y and C? Or, since these are 1980s pictures, maybe composite video? Cliff: The converter output RGB that fed an NTSC encoder. Then on to the VTR and a monitor.
|
| |||||