A PIECE OF ROBERT A. HARRIS Part 5

What follows is the transcription of a fragment of an article intitled "Universal Upbeat in Studio Fire Aftermath" by Robert Marich for Broadcasting and Cable, published on June 20th. You can read the original here.

In February, 1993, a symposium was held by the National Film Preservation Board of the Library of Congress in Washington, D.C. to discuss the problem of the decay of motion pictures. Excerpted from the transcript of the February 26 testimony is this submission by film preservationist and restorer Robert A. Harris. Few people in the world have the ability to judge the condition of our film heritage as does Bob Harris. Working with David Lean, he gave us what is arguably the best presentation of Lawrence of Arabia ever seen, even after the original negatives and sound were considered to be virtually lost. With his equally fastidious partner, James C. Katz, Bob has worked miracles on the deteriorated likes of My Fair Lady, Spartacus, Vertigo, and Rear Window, not only saving these films from loss, but providing vivid new prints that rival the originals in many ways. Before placing this transcript in the American WideScreen Museum web site I asked Bob to review and update his statements. He altered one sentence. And that, boys and girls, says a lot about the condition of the films, good and bad, that many of us grew up with.


--------------------------------------------------------------------------------

Robert A. Harris' Statement at the Film Preservation Study:
Washington, D.C. Public Hearing, February 1993
We have all heard many times that we have already lost some 50% of the films made before 1950, that our nitrate heritage is slowly turning to powder before our eyes while budgets and time are running out. This is all true. However, with little mention of our post-1950s film this perpetuates the myth that film preservation is dedicated to our remote past, something that belongs more in museums than on theater screens.


Robert Harris and David Lean during the restoration of Lawrence of Arabia
Further, it gives the impression that all of these nitrate films simply decomposed while attempts were being made to preserve them. This is untrue. Most of the early films did not survive because of wholesale junking by the studios. There was no thought of ever saving these films. They simply needed vault space and the materials were expensive to house. There also seems to be a feeling that we must save it all. Like art and the written word, there was as much junk film produced during the first half of the century as is being produced currently. I made one of those in 1987, a junk film, which we can add to the pile of junk. It simply is not all worth saving with today's limited funding.

If our greatest problem were nitrate, then my chosen work in the archival field would be incomparably simple. But it is not. In 1988, while completing work on David Lean's Lawrence of Arabia, I was asked to look into Tom Jones, the 1963 Academy Award-winning Best Picture, newer than Lawrence by only one year. Try as I did searching vaults worldwide, the best I could come up with was a single dye-transfer Technicolor print with an Italian soundtrack. It was so worn that it could not be used for duplication.

I was told that the feeling when Tom Jones was in post-production was that it would not amount to anything. A judgment call. UA never made protection separations to back up the original negative. Today, a proper print of Tom Jones cannot be produced.

Although I do not necessarily have the answers to our preservation problems, I can at least raise or help reevaluate some of the questions.

Since the early fifties, we have been dealing with the Eastmancolor negative. There is nothing inherently wrong with this material except that it fades. We have lost the original negatives to many important films of the fifties, and we are now going into the sixties; films of the seventies are now showing signs of fading.

Lawrence, a 1961-62 production photographed in Eastmancolor and processed by Technicolor London was fortunate. For some reason, the work done at that particular laboratory seems to survive years longer than film processed elsewhere, possibly the water from the Thames, I am not sure. Lawrence was still in good enough shape, although the negative was cracked and falling to pieces, that new color protection materials could be produced. We could not produce new black-and-white protection material, because the negatives would not run three times to produce separations.

The camera negative on Universal's 1960 Spartacus was totally faded, totally unusable. Nothing could be done to produce any printing material from that element. We worked from black-and-white separations and had to create the equipment to manufacture a 65mm preservation internegative on the film. We worked from the seps but those seps had been produced defectively. They had been vaulted 30 years before and never tested. I will not go into the problems that were encountered, but the lesson learned was simple and dramatic: black-and-white master separations, when produced, were routinely vaulted and forgotten, assuming they would yield beautiful results when needed. We now know that this simply is not accurate in all cases.

No one knows what materials can be produced from separation masters unless they have been printed, not selectively tested or reviewed on a Rank [film-to-tape transfer machine], but printed. This should be done before the negatives that they protect are no longer viable printing elements. If the protection is defective and the negatives have gone, nothing further can be done. Without doing so, we may have no protection for the last 40 years of color film history. Every film worth saving which has not been backed up should be looked into with immediacy.

The large format films, 70mm, Technirama, etc., are in the highest risk group. Since most 70mm prints were made directly from the camera negatives, many are extremely worn. Most large format masters are untested. They probably will not register very well and they generally are not backed up by large format color interpositives of any vintage no less recent.

My single hard and fast rule is do not rejuvenate original negatives; and this keeps happening continuously in laboratories around the world. Do not put chemicals on preservation materials. This will go against everything that you will hear if there are people coming here representing rejuvenation and scratch-removal vendors. They will tell you that the panacea for saving our film heritage is to coat it with chemicals.

There is a problem right now with the original camera negative of a very well-known 1968 film. Someone allowed the negative to be chemically treated and a lab wetgated the footage. That film is now a solid block. We are trying to ease it apart and remove the coating without the emulsion coming off.

Rejuvenation causes film to shrink, warp and shed its emulsion. Particles of dust and dirt are caught under the coatings and become a part of the picture never envisioned by the director of photography.

The overall quality of preservation work done by some vendors is a joke. What they produce is generally in the "good enough to get us paid" category. Once materials are delivered, they are generally accepted. Although much preservation material was produced before the advent of wetgate printing and therefore wear was more apparent, there were still materials produced after this process was available, which just was not done professionally.

If someone 15 or 20 years ago had made decent materials on pictures like Casablanca, they would look a lot better than they do today, and they would not have to be constantly redone. There are not adequate materials produced on hundreds of films; but there are materials. You can generalize that the more popular the film, the worse shape it is going to be in. Quality is a problem that has been with us for decades.

Nitrate is preserved once, then again and then possibly a third time, hopefully the right way. Original negatives are sometimes pulled to create non-preservation elements to be used for a video transfer. This places wear and fade on the negative without accomplishing anything.

One final point on this subject. Once materials are preserved properly, that does not then mean that the original nitrate should be junked. I have to assume that today's technology will be constantly supplanted in the future with new means of creating even higher quality preservation materials. You never want your finest surviving asset to be a dupe when you can have the luxury of going back to an original element.

If someone asked what I would do if I could selectively control all film preservation except that being done by the few studios and archives who are doing it correctly, my initial answer would be very simple: nothing. I would shut it down completely. Vendors should be checked for quality and accredited to do preservation work rather than just offering it on rate cards. Preservation work should not be synonymous with lab work. Simply producing a set of separation masters does not mean that a film is preserved. Producing a finegrain from a nitrate original does not mean that it is preserved. All these materials have to be produced correctly not just produced, shipped and billed.

There are too many situations in which the wrong material is produced from the wrong material. This does nothing more than spend preservation dollars for masturbatory or "voodoo" preservation. All this junk has to be stored, placed on databases and occasionally checked but it will always be junk. Years from now, someone will come along and wonder why it was produced but the good materials will already be gone. Once all these problems were solved, then I would start the wheels turning again.

Some people who work in preservation do not know what a preserved or restored film should look and sound like. This is exacerbated by the fact that few titles have an original print. Without a reference print, you have no idea what the intentions of the filmmakers were regarding color, density, contrast or even major points like "day for night" scenes. When a reference print is available, it may not be an approved print. It could well be left over from a reissue or have survived as a lab reject.

Work is accepted which should be rejected because some people are either too lazy or just not knowledgeable enough to know what to do. People with a background in business as well as film history, film elements and lab techniques should be running motion picture asset protection programs.

Studios and rental vaults are now placing inventory on computer. Sometimes, as in the case of Universal, people actually open cans, inspect and listen to material. However, all too often these inventories are simply perpetuated error. An element is incorrectly listed on a label, insecurely attached to a can, then transferred to a card, and then years later, from the card to computer by someone who cannot quite decipher the original handwriting. All of this with never a look back at the actual materials, especially if they are in another country or vaulted underground.

There have been too many occurrences in which I have called someone to see if they have protection on a long version or stereo tracks, only to be told that the film was monaural or that there was no long version. It is simply bad recordkeeping.

If we are going to really start taking all this seriously, now is probably a good time to begin. If we do not, here is what we can do. Make a list of films produced since 1953, then draw a line around 1965. That is the date before which it is safe to assume we will not be able to protect anything much longer unless it is already protected.

Take a look at the titles and then dismiss every great film that you would like to share with your children or grandchildren or possibly just see again. They are not going to be there when we want them. It is all as simple as that. Either we do something now and do it right or let's forget it all. It will soon be just so much junk.

Last Saturday I got to live one of my dreams... (almost).

I got to watch "Vertigo" on the big screen. I say "almost" because it was only a crappy DVD being projected. My dream involves it being shown on 35mm... To watch the DTS-70mm print that blossomed after the Katz-Harris restoration back in 1996 would be more like a wet dream..... ;)

Anyway, the soiree took place in this gorgeous theatre in downtown Mendoza. The image, of course, sucked. Don`t worry! I asked the box-office dude before going in to watch it. It`s not that I´d realised that it was a DVD while sitting on this beautiful(ly uncomfortable) leather-bound seats. I decided it was worth the $8, despite the crude quality.

Counting down the days until I get to watch "Psycho" on the big screen... whenever that may be.

Some other time I´ll tell you about the time I projected a rather nice copy of "Dial M for Murder" myself, back on Xmas 1998. Until then...

A PIECE OF ROBERT A. HARRIS Part 3

A PIECE OF ROBERT A. HARRIS Part 2

What follows is the transcription of an article intitled "Film Restoration: More than ´Rear Window´ dressing" written by Bill Desowitz for cnn.com, published on December 11th, 1997. You can read the original here

First "Vertigo" was brought back from oblivion; now the ever-popular "Rear Window" is the latest Alfred Hitchcock classic starring Jimmy Stewart being saved from the ravages of time.

To enhance its image quality, the restored version will be among the first films printed in Technicolor's new dye-transfer process (so far used only experimentally with a few prints of Warner Bros.' "Giant" reissue and "Batman & Robin").

This marks the third Universal Pictures restoration by the team of Robert Harris and James Katz, following "Spartacus" and "Vertigo." The "Rear Window" fix-up is less expensive the other large-format projects -- "Vertigo," for example, cost $1.3 million -- but no less intensive.

"Like 'Vertigo,' we can make 'Rear Window' look like it's never looked before, with rich colors to show off the new process," says Katz, who produces the restorations. "This serves the future without forgetting the past."

Indeed, as Technicolor gears up for next year's high-tech return of dye transfer (abandoned since the mid-1970s) -- which has suddenly become cost-effective in the current "event" film market -- the "Rear Window" announcement proves that the process can also accommodate special reissues as well.

According to Frank Ricotta, Technicolor's senior vice president of worldwide technical and engineering operations, the new process yields sharper, less grainy prints than conventional color positive prints, as well as "an extended tone scale resulting in blacker blacks, whiter whites and improved color rendition."

The restorers concur that the results are impressive.

"'Rear Window' is a dark-looking movie," Katz says. "And in the tests, we've been getting a lot of detail in the blacks you don't ordinarily see. There's also increased sharpness. You can see it in the bricks on the buildings (in the courtyard) and in the patterns on the dresses."

Actually, "Rear Window" has never looked as good as it could have, according to Harris, even during its initial release in 1954. That's because the first dye-transfer prints weren't made until the 1962 reissue, when they were badly timed and came out beige. "So this will be the first time we see the film's full-color spectrum," he says.

But first the restorers must clean up an extremely dirty negative, abused from the very beginning. "It's a mess," Harris says. "There were 400 runs off the camera negative before the end of 1954. We don't know why they didn't do it dye-transfer, which would have saved printing off the negative. This was reasonably unusual for the period. And we're missing 1,000 feet of negative."

It is ironic that "Rear Window" has suffered this fate, considering it was one of Hitchcock's personal favorites and his most successful film of the '50s. It was also the first movie from his peak Paramount period, which he eventually owned outright and later withdrew for a decade. When last reissued, in 1983, "Rear Window" grossed another $9.1 million for Universal, which acquired it from Hitchcock's estate.

Inspired by Cornell Woolrich's cunning short story about murder, voyeurism and confined spaces, the director was obviously at his crafty best. He had all the technical toys he desired at his disposal, plus two of his favorite stars, Jimmy Stewart and Grace Kelly. "Rear Window" was Stewart's favorite of the four films he made with the director.

Breakthrough made restoration possible
Yet were it not for a recent technological breakthrough conceived by the restorers and developed by Phil Feiner, executive vice president of Pacific Title film lab, a complete restoration of "Rear Window" would not have been achievable.

Besides the accumulation of dirt, pieces of the original Eastman color negative that were dupes (titles and optical effects) have faded in the yellow layer of emulsion due to aging and improper storage -- a malady that sooner or later strikes all color negatives from the early '50s through the '80s, when a more stable stock was introduced. This first- layer deterioration includes loss of contrast, blacks and shadows going blue and facial highlights turning, in the words of the restorers, "a lovely shade of crustacean."

Compounding "Rear Window's" problem is an additional reel of negative that had its yellow layer stripped off the emulsion when a lacquer that was applied and reapplied over the years to cover scratches was eventually removed.

"No one was malicious here," Katz says. "The film was a victim of its environment over the years. It doesn't stand alone. There are thousands of films out there like this. No one knew enough about the properties of emulsion or temperature control for color negatives."

The restorers were first stymied by the yellow-layer problem last year when trying to correct the revelatory flashback in "Vertigo." Frustrated by their inability to restore the sequence, they later approached Feiner about devising a solution for subsequent restorations.

Using test footage from "Rear Window" and the 1956 remake of "The Man Who Knew Too Much" (to be restored at a later date), Feiner spent six months implementing a cost-effective photo- chemical solution.

To bring "Rear Window's" colorful brilliance back to life, an interpositive, an extremely fine-grain, low-contrast color- duplicating element, will be made from what survives on the original negative and combined with what survives on the yellow separation master.

"This is something that's going to benefit all of these pictures," Katz adds. "All the damage seems to be in the yellow layer."

'That shot is going to be breathtaking'
The most important moment in "Rear Window" to benefit not only from the yellow-layer breakthrough but from the new dye- transfer process is Kelly's seductive entrance -- which over the years has developed a yellow-green tint. Stewart, the aloof photographer confined to a wheelchair because of a broken leg, is awakened in his apartment by her tender kiss. To prolong the kiss and make it appear more sensual, Hitchcock shot it in close-up and then double-printed each frame in slow-motion.

"That shot is going to be breathtaking in Technicolor," Katz predicts. "When people first see her, they'll understand why everyone who came into contact with her fell in love with her."

The main title sequence, another casualty of yellow-layer fading, has already been repaired. "We used five different formulas to get the precise color for the main titles," Harris says. "There's a kind of Chinese red drop shadow that was difficult to reproduce. We finally got one that was acceptable using the yellow-layer method."

The original Paramount logo, incidentally, will return to the film's beginning and end, linked, of course, with the rising and falling bamboo shades in Stewart's apartment.

While "Rear Window" lacks the stunning visual reputation of "Vertigo," its technical virtuosity is legendary, boasting Hitchcock's most elaborate set to re-create the Greenwich Village apartment complex where Stewart spies on his neighbors with his telephoto lens. The effect, as Francois Truffaut first observed, is like viewing ambiguous scenarios of love and marriage from tiny movie screens or TV sets.

It was all part of Hitchcock's famous "pure cinema" scheme: creating tension between Stewart observing and being observed -- and involving us in the suspense of waiting for him to get caught. In fact, the way he introduces Stewart and his neighbors through a series of serpentine camera movements is as succinct an example of visual exposition ever filmed.

"People don't ordinarily think of 'Rear Window' as a stylish movie," Katz says. "But what we're discovering is that there is a style in the way Hitchcock used color. With Miss Lonelyhearts, we can now see the mauve in her apartment and the green in her dress. Each apartment has its own color style."

Harris and Katz expect the restoration to be completed next spring. However, Universal's distribution plans could be complicated by a blackout in place until 1999, coincidentally, the centenary of Hitchcock's birth. Sheldon Abend, who owns the copyright on the Woolrich short story from which the film is based, won a landmark U.S. Supreme Court decision in 1990 declaring his copyright renewal was violated.

"The important thing is, we're saving another Hitchcock," Katz adds. "And we have a whole new audience right now for this."

A PIECE OF ROBERT A. HARRIS Part 1

What follows is the transcription of an article by Stephanie Argy and Rachael Bosley for the ASC Magazine. The original can be found here

BEFORE:


AFTER:

Widely regarded as an American classic and a landmark achievement in cinematography, Paramount Pictures’ The Godfather (1972) is identical to most films of its era in one respect: it was not properly preserved. Paramount, like most Hollywood studios, did not create a preservation program — “asset protection” in industry parlance — until the home-video boom of the 1980s proved film libraries could have indefinite, lucrative lives. Before that awareness took hold, original negatives were typically used as printing negatives, which meant the original negatives for popular pictures took a lot of abuse. The Godfather was not only popular, it was Hollywood’s first blockbuster, and over the years, “the neck of the golden goose was certainly wrung out,” says the film’s cinematographer, Gordon Willis, ASC, with typical candor.

Now, a new batch of golden eggs is in the offing. At director Francis Ford Coppola’s request, film-preservation expert Robert A. Harris and associate archivist Joanne Lawson of The Film Preserve recently spent a year digitally restoring The Godfather and The Godfather Part II (1974) with Warner Bros. Motion Picture Imaging (MPI). Harris’ key collaborators in the endeavor were ASC associate member Jan Yarbrough, MPI’s technical director/senior colorist; Daphne Dentz, MPI’s vice president of digital services; Willis, who shot the pictures and gave advice by phone; and Allen Daviau, ASC, who paid many visits to MPI to lend the effort a cinematographer’s eye.

After working through The Godfather and Part II, the team turned to the much more recent Part III (1990), which required no restoration at all; they did a digital intermediate (DI) to create the director’s cut, a version that previously existed only on home video. For all three films, there are now new 4K preservation negatives, separation masters from which new printing elements can be derived, and backup data tapes. Shepherding the project to completion was Martin Cohen, Paramount’s executive vice president of feature postproduction. “Marty’s support was vital,” says Harris. “He is a postproduction executive who truly understands film restoration — he gets it.”

Best known for meticulous restorations of Vertigo (1958) and Lawrence of Arabia (1962), The Film Preserve began work on The Godfather in the fall of 2006. Paramount delivered the film’s surviving elements — original camera negative, YCM separation masters, intermediate separation masters and thousands of feet of miscellaneous elements — to Pro-Tek Preservation Services in Burbank, where an inspection confirmed that radical surgery was required. Held together with tape, the original negative was filthy and riddled with scratches, rips and tears, some of which broke into the image area; in some sections, parts of the image had actually been torn away. An entire reel (1B) had at some point been removed and replaced with a dupe. Scenes were even missing from the final separation masters because they had been made before the cut was final. (Today, studios routinely combine separation masters or at least portions of them before tucking them into the vaults, but “until the mid-1980s, separation masters were typically made and never looked at or not made at all,” notes Harris.)

In consultation with Rick Utley, Pro-Tek’s vice president of preservation services, Harris determined that a photochemical restoration was out of the question. “Not only that,” he says, “but we determined that the original negative of The Godfather should never be run through a pin-registered mechanism. It could crack up.”

As he got into the project, Harris discovered that the negatives for the first two Godfather films had sustained additional damage in the 1980s, when Paramount sent them to an optical house to make new prints. The original rolls were disassembled and then reassembled incorrectly, a cheaper but chemically damaging fill was used, and the films’ lyrical 12' and 16' dissolves were replaced with dissolves of generic length for ease of printing. He recalls, “I locked a current print into a synchronizer with an original print, which is what I always do when I begin a restoration, and they were not tracking at all. Paramount knew nothing of this [damage].”

Harris believes it’s critical for a cinematographer to be part of the restoration process, and because Willis lives in Massachusetts and could not be in Los Angeles for the many months the restoration would require, Harris asked Daviau to consult on the project. “Allen standing in for Gordon was one master standing in for another,” says Harris. “Allen has the best eyes in the business —he’ll see a quarter-point difference shot to shot. The first thing I asked him to help with was figuring out exactly what ‘black’ is in these films; that was our biggest challenge in terms of Gordon’s work. Allen donated his time, and without him and Gordon, we would have been lost.”

In his conversations with the restoration team, Willis emphasized that the most important visual aspect of The Godfather was its color structure. “It’s yellow-red in much of the lighting as well as the lab work, and that ties all three films together,” says Willis. “So my main concern was to get the color right.” Harris was able to track down original dye-transfer prints of The Godfather, and one of them, from the Academy of Motion Picture Arts & Sciences’ Technicolor collection, was screened for Willis, who determined it to be an accurate color reference.

“Gordon kept saying all three pictures look the same, that they have the same rhyme and rhythm and use of contrast, and it’s true,” says Daviau. “A lot of people wouldn’t notice that, but when you’ve dealt with the color correction, you see he’s right.” Willis notes, “The repeatability of the visual structure really has to do with making the right choices. The initial choice is taste, and maintaining the look is craft. There’s great elegance in simplicity. My choices in lighting and the overall color were designed to create a mythic, retrospective feel, [one] without clutter.

“On every movie I shot,” continues Willis, “I maintained strict developing and printing control — everything was printed on one light. In fact, much of the negative on the Godfather films will only work when printed that way. I lit and exposed things at the level I wanted to be perceived on the screen; if you don’t do that, anyone can decide what your work is supposed to look like, and I never believed in giving the studios that kind of flexibility. So when making exposures, I based my exposures on the full curve of the film, shoulder to toe. The exposures are right where they should be to achieve a given look on the screen as long as they’re printed as designed. There’s no room to move things around on the printer.”

Despite the physical condition of the Godfather negative, Harris found that much of the image information was well-preserved and usable, “a testament to the viability of Kodak’s negative stocks.” The restoration began with a 4K scan of the negative, which was carried out on a Spirit Datacine. (Some pickups and title material were scanned on a Filmlight Northlight.) Harris recalls, “Every time a new piece of the negative went into the scanner, I was sweating bullets, because if we’d torn certain scenes, there was no quality backup — some shots in the first two films existed only on the negative.”

The digital files were imported into Yarbrough’s Baselight system, where they were cleaned, repaired and color-corrected scene-by-scene. There was also a team doing cleanup with MTI’s Digital Restoration System, but some details were only visible in Yarbrough’s room, which had the best viewing environment. “Most environments for doing dirt removal and scratch removal involve some sort of flat-screen video monitor, and once the dirt and scratch work is completed, I prefer to view the result on a screen similar to a standard theater screen,” says Yarbrough. “In our DI suites, we can view the projected image at 1 to 1.5 screen heights [back from the screen]. This gives us an excellent critical viewing environment in the native data resolution.”

The project stayed at 4K throughout the restoration, and Harris says this capability was key to the team’s success. “Paramount didn’t wait to do this work because they didn’t want to invest in it,” he says. “What was necessary was the ability to work in true 4K resolution; anything they might have tried on the digital side before that would have been a waste of time.” Often, the DI process is carried out with 2K proxies, but some adjustments, including noise reduction and sharpening, do not translate well from one resolution to another. In this case, occasional sharpening and noise reduction were necessary, largely to even out differences among the many elements Harris brought in to remedy problems. He explains, “We started scanning bits and pieces of separation masters, interpositive, CRI [Color Reversal Intermediate], internegative — anything from which we thought we could harvest what we thought was the finest image. One would be sharp but wouldn’t have the right color; another had the right color but was soft or damaged.”

In one instance during the digital grade of The Godfather, Yarbrough was able to draw out detail that was previously invisible because of a lab error during the production. The sequence in which Michael Corleone (Al Pacino) murders Sollozzo (Al Lettieri) and Capt. McCluskey (Sterling Hayden) in an Italian restaurant was shot over two nights, with the first night covering the first half of the scene, up to the point when Michael goes to the bathroom to retrieve a hidden gun. The first night’s work came back from the lab looking exactly as Willis intended, but the second night’s work came back too thin: the lab had neglected to push the film. “They simply pushed the wrong button and underdeveloped a day’s work,” says Willis. “When that happens, you don’t lose that much speed, but the fog level drops, and the look of the film changes. We left that lab and continued at Technicolor.”

Reshooting the scene wasn’t possible, and Harris believes Technicolor tried various photochemical processes to salvage it, creating separate masters of the daily rolls at different gammas to bring up the contrast, then mixing and matching those YCM masters to see what would look best. As a result, those shots have always been a couple of generations down, even when the negative and prints were new. Harris notes that on dye-transfer prints, it looks slightly off but not too bad, but when printed to direct positive, “the image started falling apart.”

The search for original film elements for this sequence went on throughout the restoration process — “[Pro-Tek inspection technician] Joe Caracappa spent months looking for anything from the Italian restaurant,” recalls Harris — and the team eventually recovered about 80 percent of the original negative, as well as usable intermediate separation masters for all but one or two shots. Yarbrough was able to bring out additional information, which is particularly noticeable in a close-up that tracks in on Michael just before he pulls the trigger. “Jan is like a concert pianist, a true artist,” says Harris. “Now, you can see the stress on Michael’s face. In the dupe of the dupe of the dupe, you lose that.”

This, he adds, raises ethical questions familiar to those in the preservation field. “It’s a dual-edged blade, a slippery slope. If there’s a photographic problem during production that could not be fixed at that time but can be fixed now, do we fix it?” Willis’ take: “Yes, provided that doing so has no impact on the overall restoration.”

In initial discussions with the restoration team, Coppola gave a directive: they could replace problematic frames or shots as long as the replacement was “imperceptible” except for “forensic examination of grain structure,” says Harris. In a few cases, a shot was salvaged by using pieces of camera negative that hadn’t actually been used in the movie. For example, there was a very badly damaged shot of Vito Corleone (Marlon Brando) in his hospital bed, with the camera looking up at him from the foot of the bed. Using camera reports and the editor’s line script, Harris was able to find the trims of the take that contained the footage used in the movie. He had the remainder of the take scanned, and with Yarbrough’s help, he found a portion in which Brando’s chest was rising and falling in sync with the clip that was used in the movie, and replaced the original clip with the new one. “Where you can use original negative, you do,” says Harris. “When you find a problem with a shot, you have to think about how to get around it, and the answer isn’t always to go to a dupe or to the separation masters. There are a number of viable options.”

Although restoring sound can often be as difficult as restoring picture, that wasn’t the case on The Godfather. “The sound really wasn’t a problem because everything survived, and in 1997, [sound re-recording mixer] Walter Murch went back to the elements and created wonderful stereo mixes of the first two films,” says Harris. “[Post supervisor] Jeff Cava and [vice-president of DVD mastering] Ron Smith at Paramount worked with what Walter created.” The analog tracks on the 35mm prints of the restored films are new, clean mono tracks; stereo tracks are also there for DTS and Dolby Digital presentations. “That way, if an archive wants to run the films with their original sound, it will be able to,” says Harris.

The original negative for Part II had far fewer problems than the negative of the first film. “It was filthy from continuous wet-gate printing, and the recut had created havoc, but overall, it was about 75 percent better than the negative of The Godfather,” says Harris. One reason for the disparity is that Paramount did a much larger initial print run for Part II because the first film was so popular. “The initial print run on the first Godfather was short,” explains Harris. “Many more dye-transfer prints were produced on Part II and survived, so the studio didn’t need to go back [to the original negative] to make more prints as quickly or as often. We never did find a solid color reference for Part II, however. The dye-transfer prints were all over the place, and it was a much more difficult film to work through in terms of color and densities than the first film, much more complex.

“When we turned to creating the director’s cut of Part III, which had a Super 35 negative, we had to conform to video,” he adds. While Harris and Yarbrough were finishing up Part II, colorist Ray Grabowski started the digital grade on Part III in another room at MPI, with Bob Raring, the film’s original color timer at Technicolor, providing some guidance.

Yarbrough estimates that the digital files for the Godfather trilogy required 160 terabytes of storage space. (By comparison, the final scans for a 4K DI of a 100-minute feature require about 8 terabytes.) The corrected files were recorded out to 35mm (Kodak Vision 2242) via an Arrilaser at the last possible minute because Harris kept unearthing new pieces of film. For example, well into the restoration of the first picture, he found the original negative for Reel 1B in a film can labeled “Dupe #2.” He notes, “Paramount has a superb vault system, and the staff there was incredibly cooperative. The problem was that the system was put in place in the ’90s, and two of these films were from the ’70s and had been handled by many other people. They brought in every element Paramount owned on all three films, and every element had fingerprints on it from the various incarnations of the trilogy.”

Because Part II was the last feature whose release prints were made using Technicolor’s dye-transfer process, Part III featured black levels that did not match those of the first two films. The contrast and saturation controls available in the DI suite made it possible to achieve a closer match, and new prints of all three films have been made (at Technicolor in Los Angeles) on Kodak Vision Premier 2393, making the blacks even deeper. “When we tested regular Vision [2383], the blacks were milky,” says Harris. “There are occasional shots in the first two films, maybe a dozen in 175-200 minutes, where the Premier grabs the blacks and drags them down a little too much. But I feel the sins of Premier were far outweighed by its positives.” After a pause, he adds wryly, “To be honest, I would have liked to print on Vision 2391 — which doesn’t exist.”
The restored versions of the first two Godfather films were recently screened for select audiences in Los Angeles, the first via 2K digital projection and the second via 35mm print. (New DVDs of the trilogy will be released in September.) However, at press time, the new prints were works in progress. After screening them in New York, Willis had this to say: “I think a remarkable job was done repairing all the damage done to the negative — very difficult work. The Godfather is a very good-looking picture now. They have not tried to dial anything into the film that wasn't there before.” That said, he continues, “I’m the first to admit that the darker material, especially in Part II, is difficult to reproduce properly; it was beautiful on dye-transfer, but at the moment, the positive available isn’t doing the job we all hoped for. The print stock I saw doesn’t produce a full and comfortable range from black to white, and the mid range is compressed as well. Everyone is pursuing a solution.”

Willis emphasizes that a restoration effort as disciplined as this one was essential, especially because The Godfather’s unusual look made studio heads nervous even in its day. “When people see things they don’t understand, they become frightened, and the concept of what The Godfather is — or was — still eludes some people,” he observes. “We all tend to reduce or expand things to a level we understand, and it can be fatal to a film like The Godfather if what someone understands is Petticoat Junction. The goal here was to restore, not to remake. The film has already been made.”