Can a prism be used to average all the light waves falling on one end?

This may be outside the scope of this site, but it’s worth a try.

Years ago, some time between 1989-1992, I was living in Bridgeport CT and working as a photographer. Since it was pre-Craigslist etc. I responded to a newspaper ad that was selling some sheet film equipment. The seller was a gentleman in his 50’s or 60’s and showed me a piece of equipment in his photography studio that he claimed was an analog version of JPEG compression. I don’t think he used the term JPEG since it was most likely pre-jpeg, but the concept was one that I already understood from early use of computers. At the time I believe I had a Macintosh LCIII. He said that it was used to create a "pixel dissolve" effect for film before that was done more simply with digital equipment. It was about a 3′ square frame holding many 1"x1"x3" (approx from memory) clear rectangular prisms all clamped together in a grid. I think they were held that way since gluing them in any way would change how they acted as prisms. This was a long time ago, and I was 22-24 years old at the time. I had a decent background in computers, chemistry (AP chemistry in HS) but physics was the weakest of my sciences.

He explained that if he projected a color image on one end, it would take all of the colors/frequencies that fell in each square and "average" them, passing that averaged color evenly out the other end. He demonstrated it by projecting a color slide image of a house on one side, and on the other side was a thin sheet of frosted film. Sure enough every square was a perfect solid color, not just a blurry interpretation of what was falling on the other end. He went on to say that for years experts said it couldn’t be done, but he figured out how. He also claimed he visited Edward Land of Polaroid fame and asked his advice about patents. Supposedly Land asked him if anyone else knew how to do it, to which he replied that it was still a secret. Land then told him to just keep it a secret and charge for it as long as he could. The gentleman claimed it was used for movies before the advent of simple computer effects replaced it and made it obsolete. I don’t remember any dates he might have given, but I would swear he said it was used in The Lathe of Heaven (1980), which could make sense as it was produced by WNET in New Jersey, and on a shoe-string budget where special effects were very minimal due to cost. He also claimed that for years experts still couldn’t figure out how he’d done what he did, and they would want to inspect it. He wanted to keep his secrets as long as he could so he made outlandish hints about how he did it, and would hide really strong magnets around the device so that if they snuck in a magnetometer it would give crazy readings. He claimed that one scientist insisted for years that he was doing it with magnets.

So is it possible to take all of the light frequencies falling on one end of a prism and average it all evenly out the other end? I would love to do more research about it, and wish I could even figure out who he was. But it was all so long ago and I don’t even know what vocabulary to use to even do a decent internet search.