We use cookies to ensure that we give you the best experience on our website. Click here for more information.

Topics/Questions/Exercises Of The Week—30 April 2010 (Special Blu-ray Preservation/Restoration Edition)

Glenn Kenny

The Warner Home Video restoration of David Lean's 1965 epic Doctor Zhivago is a thrill to behold in the high-definition Blu-ray format. So much so that watching it, one is apt to forget or at least ignore the problematic aspects of the film. And despite those, there's a real mastery at work here of the sort that's bracing to revisit today. The Blu-ray, and a standard-def DVD edition of the restoration, come out next Tuesday. I was fortunate enough to be able to have a conversation with Warner Home Video's Vice President of remastering, Ned Price, yesterday, not too long after the company announced it was putting a moratorium on two popular classic titles, Citizen Kane and Ben-Hur, in anticipation of remastered versions for standard-def and Blu-ray for some time next year. So we had more than Zhivago to discuss. Highlights of our conversation below.

ON THE ZHIVAGO RESTORATION: "It was done by many people. It was quite a collaborative effort. We worked quite a bit at Warner Brothers Motion Picture Imaging [MPI], and by working in-house we were able to do things like scan the camera negative in 8K resolution, whereas that would be rather beyond most people's price range for a film of this length [Zhivago is over three hours long]. This also allowed some research and development work when necessary, to tune up problems inherent in the negative. Every project of this sort is different. For Zhivago, the image was in relatively decent condition, minimal color fading, the problems we had stemmed from the fact that David Lean had wanted to shoot the film in 65 mm, and the studio didn't let him, and the concession made to him was that major markets, major cities would receive 65 mm blowups, directly from the camera negative. Since there wasn't decent intermediate, they wanted it to be as sharp as it could. But printing directly from the original camera negative, which has all the splices and everything on it, is a pretty heavy thing to put on such a piece. The splices cut loose, and inevitably they catch during the printing process, and tear the negative. Also, the perforations from the side had stretched, and then pulled. From the printing. So back in 2000 when we attempted to restore the film, ultimately we had to abandon that restoration, because we couldn't get a stable image from the camera negative.

For this restoration, we went back, working at the Warner MPI, we were able to do things like modifications of the scanner, take our time to stabilize the image, and work within perforations that were still secured, and then stablize the image after the fact. Scanning is costly, and with a film of this length it's costlier still. It's the data space as well, and the data management. Because when you scan, you usually scan to one server...and in our case we did a direct preservation output to film from that server, so we went from 8K directly to film."

PHILOSOPHY OF RESTORATION/PRESERVATION: "You can't do preservation without a fair amount of restoration. But I usually separate preservation from creating something for distribution purposes. Two very simple paths. First, you take the original materials, always the original sources, or closest to the originals you can get, and literally preserve them. Copy them, migrate them to a new piece of film. Film is currently our preservation element because it's a rather inexpensive medium. It's very dense, it can hold up...it's quite substantial. You can leave it up—not that we would—but you can leave it up on the shelf for seven years and come back to it and know it will be there. Data at the moment is not quite that reliable. So what I do is scan the camera negative of Doctor Zhivago and immediately output it to film, so you're oversampling out to film from 8K. Then we take the data and then we go downstream that for distribution. We start the dirt-fixing, we start doing the color correction. And at that point we work in film space. We complete our restoration in film space. And that restoration would be potentially for theatrical relese prints. And then we make another version for digital cinema. Which is a different color space. And then we make yet another version, which is for high-definition TV, and that's what's on the Blu-ray. Which is in yet a different color space."

WHY SOME HD TRANSFERS LOOK WORSE THAN OTHERS: "Usually the errors made in the over-processing of images are the result of the monitors being inadequate. I think currently the monitoring situation is our weakest link. So you'll find that over time as monitors improve you get an image revealed which has artifacts that were not visible prior. I find that if I'm working in very high resolution there's really no need to alter the original image per se. Though we do often things for distribution such as fix shortcut optical dupes, for things that are cut in where the color and the contrast doesn't match up at all. We do help things along so you're not jarred out of the narrative. Basically, now, we have a window into the camera negative. And if you all of a sudden see an optical dupe, it has twice the impact that it did in a release print in a theater; it can really just knock you out of your seat. So we help that along a little bit."

ON THE UPCOMING BLU-RAY OF CITIZEN KANE AND OTHER CHALLENGES: "Kane was mastered [in 2001] using 525 resolution monitors. So what you saw, what we were able to see when we were working on Kane, was not a true film restoration. We were basically creating a distribution version. And we did make film backups, on the surviving elements, the originals. Now that I've got increased resolution, you have to approach everything differently. We don't have the original on Kane, so you would first make your preservation materials again. I would revisit it, make film elements from a digital scan. Because we could get closer resolution to the second generation element which survives, rather than having a third generation dupe to put in the vault. Again, there would be less need for processing on that particular element because it's a decent element. There are cases where you're not. The opening of Wizard of Oz, the sepia element, the black-and-white negative for that, didn't travel with the camera negative, Technicolor three-strip sequences. It was, unfortunately, destroyed. And the best surviving backup element is very dense. It was improperly processed. And it lives at the Library of Congress. In that case, we needed to process part of the image because it was so thick and grainy. We had to push it a little to get the detail, and then we had to tone down...to compensate for our push-up. Some people believe we should take a very purist approach and leave a film as it is...but if we actually did that we'd be doing a disservice. The logic and approach we take is to render a theatrical experience, for the home electronic market. It's a different medium that has different characteristics than actual film does, and you have to keep that in mind."


ColumnsDavid LeanOrson WellesTopics|Questions|Exercises
Please sign up to add a new comment.


Notebook is a daily, international film publication. Our mission is to guide film lovers searching, lost or adrift in an overwhelming sea of content. We offer text, images, sounds and video as critical maps, passways and illuminations to the worlds of contemporary and classic film. Notebook is a MUBI publication.


If you're interested in contributing to Notebook, please see our pitching guidelines. For all other inquiries, contact the editorial team.