"Nunsense," or, "In the pink," or "Getting to be a habit with me," or, OK I'll stop...

So. I was getting ready to retire the other night, as I've gotten to be a bit of an early-to-bed-and-early-to-rise type of late, and I thought before that I'd check out the comments threads over at the ole blog and see if anything new had come under the transom, and I find that on the thread for "Digital Archery" there's a question from esteemed laser disc and DVD assessor Doug Pratt, and the question is, "What did you think of the pink outlines around the white nun's habits in Black Narcissus?" And my first reaction is a more, erm, Christian variant of "Oy gevalt." My second reaction is "Is that a trick question?" And if not, well, what am I supposed to think of the pink outlines? I'm guessing that I'm not supposed to approve of them. Well, I'm not going to start investigating the issue right now. It'll wait until tomorrow. 

And the next morning I got up and decided, first off, to stop being so paranoid. For heaven's sake. And I got out the standard-definition DVD of the new Criterion edition of Black Narcissus and threw it on the computer. And it did not take me long to find the pink outline of which Mr. Pratt speaks. 

Now understand, that when Pratt or other mavens discuss things such as pink outlines, they don't necessarily mean things that look like force fields that attach themselves to moving objects and so on. No, such observers are detail men. So below, a screen capture in which the discussed pink outline is at its most evident.

For those of you who might be asking "what pink outline?" see the cropped detailing at left. That pink outline. And now that we've established what the pink outline is, we have to ask, well, just what is it? Usually, shimmering halo effects around outlined figures in DVD images are accounted for by overuse or misuse of DNR, for Digital Noise Reduction, the technology that's used to "clean up" perceived film grain, cut down on actual video noise, which is something else entirely, and increase sharpness. Much has been written about how over-cleaning film grain actually reduces detail, and how you end up with a false sharpness, a big bright shiny DVD or Blu-ray image that has less picture information than what these ostensibly increased-definition formats are hyping. 

Those who follow the formats also know that Criterion is hardly a company known for slathering a lot of DNR on its releases, and that it has a "let grain be" policy that's earned the ire of...oh, hell, I'm not even gonna go there. The main point is that, first off, the anomaly is not likely to be the result of DNR because precedent strongly suggests that wouldn't be the case, but more importantly, because the anomaly doesn't look like a DNR artifact. The pink outlining—which, I should point out here, is not a constant problem, as the screen cap some grafs below demonstrates—is more solid, more constant than a DNR halo would be; it doesn't have that maddening shimmery quality. 

My suspicion was that the anomaly was a sort of holdover from the Technicolor process itself, but I could not put my finger on it. So I thought to consult my rebbe on such matters, Robert A. Harris, the film preservation and restoration expert, maestro in both theory and practice. I recall being in the Manhattan theater back in the day, at an event at which the cinematographer Jack Cardiff spoke, and Harris asking if the lavender tint that suffused the print that Criterion had used for its laser disc of Black Narcissus was supposed to be there, and Cardiff saying no, it most certainly wasn't. It seemed to me at that moment new ground was broken in terms of the director-or-cinematographer-approved home incarnation of a film, a welcome development even if it did open a whole new can of worms (see Vittorio Storaro and aspect ratios). I know historically I'm probably off on that. But I insist that the moment got the ball rolling that led to the largely excellent version of Black Narcissus we have today. 

I e-mailed Mr. Harris and asked what he made of the pink thing. And his answer made me say, "Of course." Here is some of what he wrote to me: "First, Black Narcissus, as opposed to The Red Shoes, was composited analogue; presumably going from the original negs to an interpositive...Keep in mind that three-strip Technicolor was not three fully 'direct-to-emulsion' differing exposures, but rather two, with the third exposure, what's called the cyan record (which controls red), being exposed through the base of another record—and ALWAYS SLIGHTLY OUT OF FOCUS!" (Going into even more detail about processes, he points out, "the bi-pack pair were emulsion to emulsion, with the image first exposing the yellow record, and then the cyan, with cyan—always—in slightly soft focus.This would never have been noticed via dye transfer printing, and especially with the optics in use at the time.")

This is why his observation that Black Narcissus was composited analogue is so crucial to the question. Because while digital technology can be a very hurtful thing when used incorrectly, it can also be an incredibly valuable tool to make various fixes. Harris notes: "If one is compositing the records digitally, the cyan (red information) can be slightly sharpened, re-sized, etc., toward a better fit.What we're seeing [with the pink outlines around the habits] may be either that cyan record problem, or just as possible, optical aberrations within either the camera prism system, or the taking optics themselves." Thus, Harris concludes, the new Narcissus could have done with some digital tweaking, as was used by Warner in its recent restoration of Gone With The Wind. "Or," Harris adds, "more to the case, the new Robert Gitt/UCLA/MPI restoration of The Red Shoes," the basis of the new Criterion standard def and Blu-ray editions.

Is my enjoyment of the new Black Narcissus spoiled by these fruits from the tree of knowledge? I can't say it is. What is enhanced, rather, is my appreciation of there almost always being room for improvement in these matters. 

Comments

[Assassino]

Storaro is, as Duvall put it in Network, "...intractable and adamantine..." http://mutinycompany.com/article_vittoriostoraro.html

[Mike Mazurki]

The restoration of BN was done a few years prior to RS, and as such didn't have the option of going through the same workflow RS did in restoring all three Technicolor layers separately at 4K, working instead (as Harris confirms) from a new Interpositive made the traditional (photochemical) way from the best available elements, resulting in some occasional registration fluctuations.

Registration remains the real bugbear in 3-strip restorations (provided the original elements still exist) as all three elements would have been subject through the years to unique shrinkage and other damage (warping, torn frames, etc) making them very difficult to put together seamlessly. Fortunately digital has come a long way with this, as evident from The Red Shoes, as well as the new restoration of the new African Queen.

What one needs to bear in mind is that restorations, 3-strip Technicolor or otherwise are always at the mercy of what elements still exist and in many cases, the original negatives were lost long ago.

[hamletta]

My experience is with print; I used to do press checks on four-color web (newspaper) presses. With all that motion, they jump out of register frequently, which is why the Sunday funnies look so funny some Sundays.

Doesn't surprise me that the Technicolor process has analogous problems. I remember watching "Wings Of Eagles" on TCM years ago and seeing blatant edges that would have had me running to the pressmen.

[hamletta]

Forgot to add: Anyone who'd be put off by these barely noticeable "pink edges" in such a brilliant, all-consuming work of art like "Black Narcissus" might oughta find another hobby. Like press checks.

[Doug Pratt]

For the record, I was not put off by it in the slightest. Even diamonds have flaws. I just found the oversight of mentioning it in the original article to be curious, and was attempting to elicit a minor elaboration, which, to my surprise (and pleasure), was a rather more extensive response that what I anticipated...

[hamletta]

Sorry, Doug. I was being snarky.

Actually, I love press checks. Press solvent smells like...victory.

Because it would mean I had a real job that used my skills.

[alan]

chromatic aberrations are common in photography (when capturing the image), especially with inferior or zoom lenses, and then mostly when shot wide open, mostly in the edge of the frame.

there's a bunch of 'purple fringing', as it's often called, apparent in your iphone shot (around the fire escape/patio, and most other high contrast areas):

http://somecamerunning.typepad.com/.a/6a00e5523026f588340133f209185c970b-pi

usually, it's purple but can be any color.

Black Narcissus was made in 1947 so was (probably) not shot with a zoom lens (they were very poor quality at the time). fringing also occurs with primes, though

[Joseph "Jon" Lanthier]

Jut got finished watching this blu-ray. Quite a disc, indeed. The only blemishes I found at all distracting were some wavers in color temperature throughout, particularly when the photography turned towards more opulent props and costumes (all those spangly Indo murals). I assume, however, that this is another unavoidable byproduct of digitizing three-strip technicolor.

An honest question: When Sister Ruth traipses through the jungle to visit Mr. Dean and works herself up into a cerebral hue-gasm, I could have sworn that the screen went from red to blue rather than red to black (I'm gonna dig out my original SD Criterion issue to confirm this tomorrow). Does anyone know if the blue was an inaccuracy to begin with, or if/why it was ditched in this print? I always found it rather daring, and an incredibly visceral simulation of what's it's like to faint (a coolness overcomes you after the initial white wave), so I was somewhat disappointed to see it missing.

[Fabian W.]

I don't know if you're aware, Glenn, but Jonathan Lethem linked to this post on his Facebook page, calling it "One definition of love".