I'm responding to posts by Adam_S, as well as that just previous, by Mediagy:
First to Adam's comments:
You're words regarding prints, theater surroundings, lamp houses, color temperature, etc. are very accurate. I'll add to that the huge difference between carbon arc (and the type of carbon rods) and non-carbon -- pick a bulb flavor, xenon, etc. Watch any cinematographer today walk into a screening and he or she will pull out meters, and check illumination and color temperature in both machines. Add another layer, the age, condition and "color" of the reflector, the age of the lens, ie. glass, etc. These all make an impact on the final projected image.
Want to make it even more obtuse?
While smoking is no longer permitted in movie houses, it was once the norm, and allowed us one of those great memories -- shafts of moving light being emitted from the booth and traveling magically to the screen.
Those who love film generally agree that some of the finest color work came from the Powell / Pressburger films -- The Red Shoes, A Matter of Life and Death, Black Narcissus, etc.
The late Jack Cardiff, cinematographer extraordinaire, was in final color timing of Black Narcissus at a major theater in London. As he told it to archivist Scott MacQueen, he and the team of Technicolor were all seated viewing a final dye transfer print after weeks of work, and had finall hit perfection. Everything was as he wanted it to be.
The film ended, main curtains closed, the lights came, congratulations and thanks all around.
And then the curtains parted, and a crew walked on stage...
with ladders.
They didn't think much of the scene until men began climbing the ladders with cans of paint...
and began painting the screen a beautiful new white.
Cardiff turned to the house manager and queried what was occurring. The answer?
Every five years or so, mostly because of the constant smoking in the house, they needed to re-paint the screen white. And as they looked on the painters began to brush pure white over what they then realized was a
very faded yellowed surface. And they'd just approved final color timing, based upon all of the various parameters of that theater.
A bit more information on prints. As dye transfer prints were created in decades past, generally in 1000 foot, 11 minute reels, they were assessed by people whose job it was to identify all of those reels that came out a point or two or three magenta, cyan or yellow. These reels were assembled into final prints. Those prints which correctly represented and matched the final answer print were shipped to the major cities, and down to those less perfect, which generally ended up in locations such as Horse's Breath, MT.
To Mediagy's comments:
Gone with the Wind has changed mightily over the decades. The original nitrates from 1939-41 were color timed much like titles such as Nothing Sacred and Little Princess. With the additional black & white record in place these early "four-strip" prints had a tendency toward dense, flat and heavily lowered color saturation. Unlike Oz, which was a fantasy film, or the later Fox musicals, which literally bounced off the screen with color, GWTW was relatively muted and tended toward an overall sepia tone reducing color. The last prints to look anything close to this, and which were still quite a distance away, were the 1954 safety re-issue prints. I have one of these in my office at the moment, which will shortly be on its way to the Academy Archive.
I'm personally aware of only two original prints or sections thereof. One, which I'm told is in an archive in China. The other reels protected by the George Eastman House in Rochester. I've examined those in Rochester.
I've had several discussions over the years, probably the earliest with Turner's Dick May, who created the protection IP in the '80s, and they all center around what an audience recalls, and what a modern audience will accept. The bottom line is that while the color and density of an image can be gently pushed toward the original, it still must appeal to a modern audience. And what we're seeing in the Blu-ray of GWTW, as scanned and finalized at Warner's MPI via Ned Price, is not only a modern magnificent image in quality, but concurrently pays proper respect and tribute to Mr. Selznick and his team, who created the film. Mr. Price has access to all available color reference going back to 1939, has taken that information and molded it into something that I firmly believe David O. Selznick would not only recognize, but as a consummate showman, would approve for
his audience of 2009.
RAH
Quote:
Originally Posted by
Adam_S 
I've seen original prints of 3-strip films on nitrate, and I've seen prints of 3strip films that are color balanced for today and color balanced according to the original look. That said, even though the most recent viewing of a three strip film was only a few months ago, I would never claim to have an accurate memory of what those films look like. Any memory I would have would be hopelessly biased and would fixate on one aspect, there's no way to win when you're comparing against memory, because the memory will always triumph, that's why comparison's like dvd beaver's are so invaluable for making us look at all of them together rather than simply relying on appeals to authority (an authority like memory). I've seen a lot of different films at different movie houses, The decor of a theatre, the color temperature, strength and age of a bulb in the projection booth, or the color of the curtains can have a tremendous effect on your color perception of a movie. Was the movie you saw in 1947 in a theater with red curtains? then it probably affected your perception of the color of the whole movie. Not to mention the warming effect that a projection bulb probably had at that time (making everything more yellow). How well do you remember the color/density and quality jumps that would occur during reel changes then? It's something I always notice in older movies projected at houses that make reel changes, on older prints you can almost always see a difference. But your brain is likely to smooth over this difference within moments and a reel that is in different condition from a previous reel when soon appear to your eye to be the same as the reel before (unless it's a really radical change). And if you then change to a third reel that is identical to the first reel, your brain will see it as being substantially off from both of the previous reels, and then it will perceive it as the new normal within moments. And this isn't even getting into how a home environment can provide totally different optical illusions about the color look and color temperature perception of a film.
This is one of the reasons why you shouldn't try to calibrate your set for a particular movie, but calibrate it to a default standard (akin to what professional equipment uses, though commercial equipment is not nearly so reliable), and the fact that you spent four hours trying to do so should illustrate the futility of it.
Perhaps RAH can provide real information on how the color team A-B-ed Gone with the Wind against other references, but the fact is that any one person's memory is a totally unreliable reference, it's mutable and completely variable. If you were a colorist you may get stuck in the first five shots trying to make each shot match your memory only to go back and change it again as though your memory of the original had changed (which it had) when you looked at the new correction you'd made. The colorists at WB were able to use hard, real-world references as to what a vintage print of Gone with the Wind looks like, and they may have very well tried to match the final corrected master for this DVD to that of the original prints, or to one of the reissues, or maybe they color timed it without reference. But it's pretty likely that the colorist on Gone with the Wind has a much clearer picture of what a 1940s GWTW print looks like than any one person who last saw a 1940s GWTW print in the 1940s.