[julie_m]

Correcting for ion burn by boosting the signal within a central circular area should not actually be too difficult a task, especially compared to resampling between line standards and colour encodings. It would need one output frame's worth of memory containing a table of values from 0 - 255 (or pairs of values 0 - 15, for half the memory requirement but then needing a bit more processing; the eternal battle of time versus space, and all that .....) representing the gain to be applied to that pixel.

The awkward bit is going to be programming it to the exact area of the damage on the picture tube, and the amount of gain required ..... Because if you got that wrong, it would look even worse! Anywhere the gain was too high, the image on the tube would be excessively bright, and vice-versa. There might also be a bright ring (if the boosted area was oversized, so increasing the signal where the phosphors already produced plenty of light) or a dark ring (if undersized, so leaving some of the dim phosphor with an un-boosted signal) in the middle of the picture.

I suppose it's also still technically over-running the tube, just using the set's video amplifier as an accomplice in the crime. Star that shines twice as bright, and all that .....