Question for colorists

[Neil Cronin]

Because there are quite a few colorists here: Why did the tennis ball in The Shining change to pink?

-n

 

[David Mullen ASC]

Just to chime in as a cinematographer... many of the hallways in "The Shining" were lit with Warm White fluorescents (maybe not Danny's though, but the kitchen set was for example), which have a mild green cast that would have to be timed out of prints and video transfers, by adding magenta. The main room with Nicholson was not lit with Warm White tubes however. So perhaps the ball shifted when they tried to time the green out of the fleshtones in the hallway lit by Warm Whites, and perhaps the color problem was exaggerated by the aging of the negative. The final shot on the blu-ray looks a touch too magenta however, maybe it was over-corrected.

Another possibility is just that the ball was changed in color by Kubrick, though it seems unlikely since pink is not a common tennis ball color. But in the documentary "Room 237", they point out that the typewriter changes from a light beige to a medium grey at some point.

 

[Neil Cronin]

Another possibility is just that the ball was changed in color by Kubrick, though it seems unlikely since pink is not a common tennis ball color. But in the documentary "Room 237", they point out that the typewriter changes from a light beige to a medium grey at some point.

Every prior "print" I've seen has Danny's ball as green/yellow. The Blu-ray, made in 2007, which was 7-8 years after Kubrick's death, has it as pink. Would this mean that there was no DI for the Shining's 2007 release?

TIA!
-n

 

[David Mullen ASC]

What do you mean by a D.I.? That term is usually used to describe something color-corrected digitally for cinema release, though now it's a bit vague since digital is not only an intermediate stage, it may also be the delivery format for cinema release. For home video release, you'd probably describe what happened in 2007 as being "digital mastering" or an "HDTV transfer/telecine" for blu-ray -- doesn't really matter, all the technology is more or less the same, I just don't know if they started with a film scan or a telecine transfer, that's all.

As I said, to time out the slight green cast, if they had shot under Warm White fluorescents, would have required adding some magenta, whether to a print using photochemical methods or in a digital color-correction session. Maybe the aging of the negative, which causes uneven dye fading and loss of color saturation, has exaggerated the problem so that even more magenta was needed to correct out the green cast, and in the process, with an image on the negative that has faded somewhat, so less chroma information to work with, the tennis ball lost its yellow-green color.

But clearly from the two frame grabs, the colorist of the first version also kept the image on the greenish side and the colorist of the second version kept things on the magenta side -- somewhere in between would probably be closer to neutral.

As for the ball, even if the color had faded a bit over the years in the negative, the colorist could have employed a mask in the color-correction software and tracked the ball and colored it separately from the rest of the frame, to retain its yellow-green color while correcting Danny's flesh tones away from green. However, generally colorists working on mastering for older titles don't like to employ such tricks as correcting individual objects in a scene because they could be accused of creatively tampering with the original, unless they are trying to fix something very obviously wrong, like a red stop sign or the American flag, something everyone knows the color of. In general, though, a colorist's priority is flesh tones.

So my guess is that the colorist who did the 2007 transfer balanced the colors in the scene for Danny's skin, and if the image had a green cast, maybe from fluorescent lighting, that would mean adding magenta, and if the negative had faded somewhat, you're adding magenta to a desaturated image which could tend to give the image a somewhat magenta cast. So the faded tennis ball in the shot shifted to magenta and the colorist did not correct it separately, for whatever reasons - either he was told that he couldn't use masks and windows to change or alter the color of selective parts of the frame, only do overall color adjustments, or he felt he shouldn't take that step, or he simply missed that the color of the tennis ball had shifted, or maybe he even thought that was the correct color.

Keep in mind that the color elements for "The Shining" are fading slowly over time, you won't get the same colors in a print made today off of the original negative as you would have in 1980. Over the years, if working from original 1980 film elements, someone is going to have to rebalance the colors just to get normal fleshtones, and in doing so, other colors in the scene may shift.

The color of the tennis ball could easily be fixed today with digital color-correction software, just by tracking a window on the ball, which is easy -- it's a round shape in a solid color -- but would require someone to OK such a fix.

 

[Neil Cronin]

What do you mean by a D.I.? That term is usually used to describe something color-corrected digitally for cinema release, though now it's a bit vague since digital is not only an intermediate stage, it may also be the delivery format for cinema release. For home video release, you'd probably describe what happened in 2007 as being "digital mastering" or an "HDTV transfer/telecine" for blu-ray -- doesn't really matter, all the technology is more or less the same, I just don't know if they started with a film scan or a telecine transfer, that's all.

As I said, to time out the slight green cast, if they had shot under Warm White fluorescents, would have required adding some magenta, whether to a print using photochemical methods or in a digital color-correction session. Maybe the aging of the negative, which causes uneven dye fading and loss of color saturation, has exaggerated the problem so that even more magenta was needed to correct out the green cast, and in the process, with an image on the negative that has faded somewhat, so less chroma information to work with, the tennis ball lost its yellow-green color.

But clearly from the two frame grabs, the colorist of the first version also kept the image on the greenish side and the colorist of the second version kept things on the magenta side -- somewhere in between would probably be closer to neutral.

As for the ball, even if the color had faded a bit over the years in the negative, the colorist could have employed a mask in the color-correction software and tracked the ball and colored it separately from the rest of the frame, to retain its yellow-green color while correcting Danny's flesh tones away from green. However, generally colorists working on mastering for older titles don't like to employ such tricks as correcting individual objects in a scene because they could be accused of creatively tampering with the original, unless they are trying to fix something very obviously wrong, like a red stop sign or the American flag, something everyone knows the color of. In general, though, a colorist's priority is flesh tones.

So my guess is that the colorist who did the 2007 transfer balanced the colors in the scene for Danny's skin, and if the image had a green cast, maybe from fluorescent lighting, that would mean adding magenta, and if the negative had faded somewhat, you're adding magenta to a desaturated image which could tend to give the image a somewhat magenta cast. So the faded tennis ball in the shot shifted to magenta and the colorist did not correct it separately, for whatever reasons - either he was told that he couldn't use masks and windows to change or alter the color of selective parts of the frame, only do overall color adjustments, or he felt he shouldn't take that step, or he simply missed that the color of the tennis ball had shifted, or maybe he even thought that was the correct color.

Keep in mind that the color elements for "The Shining" are fading slowly over time, you won't get the same colors in a print made today off of the original negative as you would have in 1980. Over the years, if working from original 1980 film elements, someone is going to have to rebalance the colors just to get normal fleshtones, and in doing so, other colors in the scene may shift.

The color of the tennis ball could easily be fixed today with digital color-correction software, just by tracking a window on the ball, which is easy -- it's a round shape in a solid color -- but would require someone to OK such a fix.

That makes sense, thanks for the insight David. Sorry for abusing the term "DI". I assumed that the 2007 Shining "print" to Blu-ray was managed entirely digitally, and I assumed that masks and windows would be used where needed, and I assumed that this was a case where they were needed. Your point is taken about creative liberty not being a good thing through these processes, in general.

I wonder, though, why did this global coloring process not notice this continuity error? Time? Proper communication channels/Chain of command? Technology?

I also take your point about the negative decaying. I must say that even though there are some issues with this 2007 "print", it looks great in general. It's amazing how much better it looks compared to the 1990s versions.

-n

 

[Marc Wielage]

As for the ball, even if the color had faded a bit over the years in the negative, the colorist could have employed a mask in the color-correction software and tracked the ball and colored it separately from the rest of the frame, to retain its yellow-green color while correcting Danny's flesh tones away from green. However, generally colorists working on mastering for older titles don't like to employ such tricks as correcting individual objects in a scene because they could be accused of creatively tampering with the original, unless they are trying to fix something very obviously wrong, like a red stop sign or the American flag, something everyone knows the color of. In general, though, a colorist's priority is flesh tones.

That's true. I believe the original transfer of The Shining (and quite a few other Kubrick films) were initially color-timed by Pat Miller over at Warner MPI, and then later 4K versions were done in the mid-2000s by Jan Yabrough, and supervised by longtime Kubrick editorial employee Leon Vitali.

It's easy to pull a key off something like a tennis ball, so you wouldn't necessarily have to track it -- the key itself could be so precise, nothing else in the frame would be affected (assuming there's no identically-colored object in the frame). If there is, then you'd have to use a garbage matte to avoid contaminating everything else.

Secondary color controls could be used to fix skintones without changing anything else. Usually the deal is to set a reasonable "flat" picture, then balance the blacks, mids, and whites, and then start working on skintones once everything else is OK. It's not unusual for three or four secondaries to be activated, and this can be done pretty easily (in very little time). A lot depends on the nature of the original cinematography, but a film like The Shining is most likely going to look damn near perfect right off the negative scans, assuming they're done correctly. Starting with an original balanced image helps a lot.

I'm often puzzled why little things like a tennis ball (or a white tablecloth or something obvious like that) gets contaminated with weird colors in a later transfer. But you never know: there's always the chance somebody in charge demanded that it look that way. And it's not unusual for transfers this important to use an archival print as a reference right in the room, using an Arriflex LocPro or a similar 35mm projector. Even though it's technically in the wrong colorspace, at least it can give you the general intent of the scene.