A Few Words About A few words about...™ The Myth of Dye Transfer Printing (1 Viewer)

[Peter Apruzzese]

You mean the rorschach images?

I see a couple of gray candleholders stacked upon each other.

 

[bigshot]

The squiggly line to the left of the picture is the soundtrack. That's where the audio comes from when the print is projected.

Yes, I know. But I don't know if the film scanner they used to get that image had the latitude to reproduce a tint in a completely clear part of the film. It's easy to get blown out by the scanner. If I had the piece of film in my hand, I could tell if the stock was yellowed from the leaders and soundtrack. But not from a scan.

With cels, a cel might look basically clear to the eye, but if you back paint a color on it, the tint will have a significant affect on it. That is especially true of rosy pink colors. As an example... Mickey Mouse's flesh color is a pretty bright pink, but every nitrate cel of Mickey that I've ever seen has yellowed to the point where it looks like an orange-ish cream color. You can see the exact same thing on Snow White's flesh color in the artwork I posted earlier. Nitrate yellowing can have a big impact on color.

It may be that you folks just haven't had to deal with film prints that have gotten to the state of decomposition I'm talking about. When the yellowing starts, it's usually accompanied by buckling. And by the time the stock turns that amber color, it is usually brittle and cracking. It certainly wouldn't be projectable by that point.

 

[Dave Moritz]

This is one of the things I love about this forum is having access to the information and experience that Robert Harris brings to Home Theater Forum. I very much appreciate and respect the knowledge and input that Robert Harris give us. Thank you Robert for all you do here at HTF :- )

 

[Bob Furmanek]

It may be that you folks just haven't had to deal with film prints that have gotten to the state of decomposition I'm talking about.

This folk has handled nitrate prints since the late 1970's and worked with one of the top nitrate preservation labs in the country, John E. Allen. I've seen all stages of decomposition. In fact, I just donated an early 1920's travelogue that was getting powdery to the Library of Congress.

One of my jobs at the lab was handling nitrate stock footage with literally thousands of examples dating back to the turn of the century. I've worked with 35mm prints, lavender fine grains and camera negatives, including THE WEREWOLF OF LONDON and MEET JOHN DOE. Come to think of it, I'm the person who found that one!

We've just preserved the earliest known nitrate footage from a 3-D film. http://www.3dfilmarchive.com/home/holygrail

There isn't any "nitrate yellowing" on the SNOW WHITE scan. The image is accurate to what audiences saw theatrically in 1937. I saw the footage when Jack Theakston scanned it and the image that I have posted is an accurate representation of the print.

Let's get back to my original point for sharing it. Should the new masters be timed to match a year of release Technicolor nitrate print?

 

[ThadK]

Let's get back to my original point for sharing it. Should the new masters be timed to match a year of release Technicolor nitrate print?

It should at least look like some kind of projected IB Tech, which can't be said for any home video release of Snow White.

 

[ThadK]

This is one of the things I love about this forum is having access to the information and experience that Robert Harris brings to Home Theater Forum. I very much appreciate and respect the knowledge and input that Robert Harris give us. Thank you Robert for all you do here at HTF :- )

I do too! It makes the misinformation presented even more laughable when it's being used against historians and preservationists who restored so many classics of cinema and take the time to post here.

 

[bigshot]

Well perhaps the print I saw at UCLA and at the LACMA screenings was of a later vintage. It looked a lot closer to the artwork than that scan does.

 

[bigshot]

Back in the early 80s, Disney put out books that consisted of a couple hundred frame blow ups from every sequence in one of their animated features. I collected them because they were a handy instant reference to find which scenes had which color palettes and to identify where in a film a particular piece of artwork came from. The color balance of the printing was very close to the print I saw a couple of times at LACMA, and also the infamous improper aspect ratio theatrical release. The frame grabs appear to be taken right off a release print- I'm guessing from around the time of the 1967 or 1975 re-release. If you're interested, I'd be happy to scan a couple of pages around Les Clark's dance sequence and show you how I have always seen the colors. It's totally unlike either the pumpkin orange version, or the green walled blu-ray version.

 

[ROclockCK]

Once again, I love it when you guys talk technical.

SnowWhite_zps4653e204.jpg

Question for the experts: "My 'squiggly line' has snow flakes floating through it...is this normal?"

 

[Robert Harris]

The Snow White scan appears remarkably like, and accurate to some frames that I scanned from another 1937 production from which I scanned original frames, as well as a nitrate of Nothing Sacred that I donated to an archive. And then there's the 1939 print of GWTW. All the same look, with no yellowing of the base.I'll post frames early next week.RAH

 

[notmicro]

Technicolor London always seemed to be the best of the best.

I did some testing in Beijing with Duel in the Sun. I believe the plant is now a strip mall.

RAH

Speaking of which -- why on EARTH hasn't there been any activity (so far as I'm aware) on a release of Duel in the Sun??? Talk about a Technicolor classic begging for HD...

 

[ahollis]

Speaking of which -- why on EARTH hasn't there been any activity (so far as I'm aware) on a release of Duel in the Sun??? Talk about a Technicolor classic begging for HD...

MGM controls the property through there agreement with Disney/ABC. Duel and several other Selznick titles just languish on the shelves waiting for the agreement to end.

 

[Bob Furmanek]

The frame grabs appear to be taken right off a release print- I'm guessing from around the time of the 1967 or 1975 re-release. If you're interested, I'd be happy to scan a couple of pages around Les Clark's dance sequence and show you how I have always seen the colors.

No, thank you. Re-issue Technicolor printings are usually inaccurate and would be of no value in determining how the film was seen when first released. The fact that they are being used as reference for new color timing is odd.

 

[bigshot]

The frame grabs in this book are actually more representative of the color palette used by the artists who created the color harmonies for the paintings that were photographed. Could they have been rushed in producing the first release prints and later went back and retimed to improve things? Or maybe someone at Technicolor was worried that Disney's palette was too subdued and suggested goosing it. It's odd that Frank Thomas seemed to think that the print we saw at LACMA was an accurate color balance. I talked with him about that a bit and he indicated that it was.

It's interesting if the original prints had thick hues wallpapered over the top of the colors like that. Was it all through the Dwarf's Cottage scenes, or was it just the dance scene the only one to be shot through a strong filter like that? It pretty much kills the work of the color stylists.

 

[bigshot]

At what point did they stop doing the gross corrections in timing for effects like that? I know Gepetto's Workshop is pushed hard to the orange side. And Night on Bald Mountain artwork doesn't always look the same as the film. But other than pushing some night scenes a bit to the blue, I can't think of any later films where the colors of the artwork vary so much from the color of the print. From Cinderella on, the colors in the film are exactly the same as the colors on the artwork. I've spoken with ink & paint ladies who did wedge tests in the color lab, and they said the chemist was working to make the results on the artwork as close as possible to the results on the screen.

 

[ahollis]

I just wonder how true reproductions in a printed book are to the actual frame or cell. There seems there could be so many areas before the final printing of the book that colors could change, even slightly?

 

[ThadK]

Why would a scan of a book scan of a film frame from decades ago be more accurate than an actual film frame scan today?

 

[bigshot]

I had the book back when I saw Snow White projected, ahollis. At the time, I found it useful because it represented the way the color palettes in the film looked on the screen. It isn't perfect... it's a bit duller in color, but you can easily tell that the dance sequence is nowhere near the pumpkin orange frame scan we've seen here. It is in the same neighborhood as the original artwork I posted.

Obviously, I can't speak for the way the film looked in 1937 because I am too young for that, but every time I saw it projected, it had this particular color balance. The clips I've seen on the Disney TV show have always had this palette too. When I spoke to Frank Thomas about the colors in Snow White, he indicated that the prints I had seen with him at LACMA were the way the artists intended it to look... a subdued palette that favored earth tones. His guardedly diplomatic comment about the colors in the first video release was, "Nice colors... not the ones we created for the film in 1937, but nice." It's pretty clear that the 1937 frame, the release print Frank Thomas and I saw at the LACMA screenings and the blu-ray are three completely different color schemes.

My question is, how long did that orange filtering remain on the dwarf's cottage scenes? If I remember correctly, Snow White was re-released something like 8 or 9 times. All of the releases I ever saw had the subtle earth tone palette. What year did they stop putting the orange filter over the top?

Again, I'd be happy to scan a couple of pages from this book and post them if someone is interested to see the colors I'm talking about. It's a pretty good representation of what the film has always looked like in my lifetime.

 

[Stephen_J_H]

Let's get back to my original point for sharing it. Should the new masters be timed to match a year of release Technicolor nitrate print?

I'm gonna say that it depends. If a film was timed for carbon arc projection, it ought to be adjusted for that. If it's timed for xenon projection, no adjustment should be necessary. Of course, all harvests should be adjusted to correspond with a reference print.

 

[bigshot]

Carbon arc projection wouldn't make that much of a difference. We're talking about completely different approaches to the color timing here. Are there other blu-ray releases of three strip features that are this different? For instance, Robin Hood. How accurate is that blu-ray to the original release?

I know from working in animation production for TV that first run is often very flawed. We get rounds of "second air retakes" to correct mistakes that we don't have time enough to fix for first release, and often things are tweaked to improve the film as a last pass before putting it down. Could such a thing have been the case here? Is it possible that Technicolor did some quick and dirty color shifts for the first batch of prints, and then the artists from the studio had them go back and make corrections for subsequent color timings? I can easily see a color stylist getting very angry at that orange filter. They tried their darnedest to create all of the color and lighting effects in the artwork so they wouldn't have to depend on gross color corrections like day for night in live action.