Early digi in analog?
 

I have a copy of the 1936 RCA Blue Book on the developments by all the RCA engineers in the day. There is something for everyone in the 452 pages. But page 266 had a paragraph that jumped out at me as a latter day director and engineer well versed in analog and digital video manipulation that we all know now. They discovered a trick to electrically manipulate the Iconoscope to move the video around like our early digital tricks.

I will paraphrase a bit and include quotes as needed. What they discovered is that the target plate of the Iconoscope was so good, the electronics that followed could not reproduce the full resolution of the mosaic. So they started playing around and did a few tricks which never made it to production that I know of.

"...it is possible to scan an area considerably smaller than the full size of the plate before the resolution of the plate becomes the limiting factor".

"By adjusting the position of the scanning pattern to various sections of the mosaic, the effect of turning the camera may also be obtained."

In other words, reduce the scan area to the point of compatibility with the circuts. The trick was to introduce DC at H/V rates in to the sawtooth to shrink the scanned area. This gives you a cutout scan inside the full plate which is still getting the "wide" shot but only showing your cutout. You could "zoom" the camera to a center spot. You could "pan" the camera to a side image closeup or a pan across the stage and then revert to the original stationary wide shot. A camera op or video engineers nightmare.

It is underscan anyway you look at. So is an 8k digi replay with a box showing the wide receiver fumble. RCA...been there...done that.

Wow - never heard of this, but it's great that they thought outside the box in terms of "what can we do with this great new piece of gear."

Very fascinating, and thank you for sharing this, Dave.

If I am correct, they sort of figured out "electronic image stabilization" 65 years before camcorders had it.

Not stabilization, and definitely not digital. So they had a full image on the plate and they could select sub-image from it.

Yes, this is how crappy cameras do HD on a 4K or 6K sensor, because they do not have power to resample 6K into HD in realtime with decent quality, so they scan a smaller section, which also effectively changes focal range. Or, you can shoot stationary 4k or 6K or 8K and then do pan or tilt in post, although it may look fake, which it is.

Anyway, they did not mean it as a stabilization method, although they probably could do it in analog domain.

Maybe we just call it Iconoscope Pan and Scan like our 4x3 movie VHS conversions.

If the iconoscope image plate resolution was better than 4x necessary I can think of a useful application. If you needed multiple title slides or test patterns for a station or show that could be switched between rapidly then you could probably put all 4 images on the iconoscope plate then pan around to each...