Preparing for Act III

If videography is a grand drama, shooting may be considered Act I (with pre-production the Prologue). Post-production editing would then be Act II. If so, what about Act III? And does being digital help or hurt?

In discussing acts, it may be useful to consider actors. Nicole Kidman is one of the most visible movie stars these days. The Internet Movie Database lists 47 titles in which she has appeared, including six movies that haven’t yet been released, TV movies, and a TV series. Of course, she’s not yet 40 years old.

Barbara Stanwyck lived into her ’80s, and the Internet Movie Database lists an even more impressive 88 titles for her. But neither of them can hold a candle to the person the database lists as having her credit on an astonishing 213 motion pictures! And that’s an incomplete list.

In 2004, alone, her name could be found in the credits of 50 First Dates, Garden State, Secret Window, Shark Tale, Shrek 2, Spider-Man 2, and the short Wednesday Afternoon, according to the Internet Move Database. The movie database of The New York Times adds The Chronicles of Riddick, A Cinderella Story, Collateral, Exorcist: The Beginning, and New York Minute. For 2003, the Internet Movie Database lists 23 features for her, with extraordinary range: from Kangaroo Jack to Mystic River, from The Last Samurai to Bad Santa, from Matchstick Men to <Freddie vs. Jason, from Holes to The Matrix: Revolutions, from Charlie’s Angels: Full Throttle to Gothika.

Who is this person who has been involved in more top-grossing movies than Steven Spielberg? Her name is Mo Henry. She’s a negative cutter. And so was her father.

A negative cutter, to use a term from the old days of videography, is a conformer. She takes precious, often irreplaceable, negatives and cuts them to match edit decisions made using film “workprints” or even computer workstations.

In 2002, when she was working on a project involving a 1962 movie, Henry told Los Angeles magazine that negative cutting hadn’t really changed at all in 40 years. It has probably been a lot longer than that.

For various hundredth-anniversary tributes to motion pictures, sequences from movies dating back to the beginning of the art were inter-cut with more-recent classics. That was possible, it is said, because there is no technical moving-image standard as enduring as that of 35-mm film. But, is that really true?

In those tributes, the motion of actors in silent movies appears exaggeratedly fast. Silent movies were shot at different speeds from the 24 frames per second used in movies today. So were Cinerama movies.

In fact, the 1962 movie that Mo Henry was working on restoring was a Cinerama movie, How the West Was Won, which had originally had its negatives cut by her father. Its speed was the least of her problems. Cinerama used three synchronized strands of film in both shooting and projection to achieve a high-resolution widescreen presentation.

Traditional Technicolor films also used three strands in shooting, this time one each for three different color ranges, like imaging chips in a color video camera. Technirama achieved a widescreen aspect ratio (ratio of picture width to height) by running 35-mm film sideways (with an eight-perforation frame width instead of the traditional four-perforation height) and added anamorphic optics to intentionally squeeze the image. Techniscope went the other way, creating the same widescreen aspect ratio by chopping 35-mm films in half vertically and using only a two-perforation height.

Obviously, a projector designed to deal only with a modern-standard 35-mm frame could not properly display either a Technirama or a Techniscope movie directly. Seventy-millimeter film strands wouldn’t even fit in the projector. There have also been changes over the years in sprocket-hole shapes, sound tracks, picture shapes, and even the width of a 35-mm film strand (sometimes shaved to get around legal restrictions specifying the measurement). But Mo Henry was right. The basic technology of negative cutting hasn’t really changed.

Similarly, the basic technology of final video editing hasn’t really changed since the days when it went from a physical cutting of the medium (more similar to negative cutting than to any form of electronic editing) to an electronic switch from one source to another (see sidebar “The Cruelest Cut”). Whether that switch operates on analog video or digital, whether it’s controlled by a microprocessor in a computer or by a finger on a stand-alone button, it’s still an electronic signal switch, not a slicing and reconnecting of physical media.

Just as many old movies can still be seen today, so, too, can many old TV shows. Some were actually recorded on film. Others were recorded on the old two-inch tape used in the first videotape recorders to be sold. A few compatible players still operate for converting archival material to newer video formats.

They may not be around forever, but, then, neither will old movies. Some old movies, recorded on nitrate-based film, have already deteriorated beyond hope of restoration. Others have shrunk so much that they cannot be projected.

When Fox began to preserve its Movietone News library in 1993, some of the film was in that shrunken state, so equipment was developed that could collect and re-size the images to their original values. As for the recording medium, Fox chose ID-1, an instrumentation (data-recording) form of the D-1 digital videotape recording format.

Questioned about the longevity of the format, Fox’s Andrew Setos noted that, as long as the signals were digital, it didn’t really matter, because the tapes could be cloned to new digital formats at any time. In fact, he said, Fox could simply have written down the ones and zeros of the digital information and had someone later enter them on a keyboard.

In theory, that’s absolutely true. But practice is something else. The Movietone film library had over 40 million feet of film, with sixteen 35-mm frames per foot, and 1024 scanning lines per frame in the camera format chosen for the restoration. That’s over seven quadrillion ones and zeroes, not counting sound. A thousand typists who could accurately enter five ones or zeroes per second, working 40-hour weeks with no coffee breaks, would take almost 200,000 years to make one clone of the library.

Clearly, Setos didn’t literally mean that writing down the ones and zeroes would be acceptable. But the task of cloning the library is daunting in any case. Assuming the ability to play back the ID-1 tapes in real time and record clones simultaneously, if the operation took place 24 hours a day, seven days a week, it would take roughly a year to clone the library to a new digital format. And would money be available for the task?

It’s difficult to think in terms of quadrillions of bits. Perhaps it’s best to consider something more manageable, like edit-decision lists (EDLs). They contain much smaller numbers of bits — simply records of time codes at which edit-points were selected and the sorts of transitions at those points.

The idea of EDLs was to make electronic video editing more similar to the workflow of film editing. The expensive, high-quality, potentially irreplaceable videotapes would not be used in edit-decision making. Instead, the video equivalent of workprints would be used for “off-line” decision making, and then the original tapes would be conformed to the decisions — roughly equivalent to negative cutting.

An early implementation was the CMX 600 and CMX 300 combination. The CMX 600 sounds like something from a recent video trade show exhibit. The video was recorded on removable hard drives, and a stylus that could sense the light coming off a video screen was used to select edit points. It’s hard to believe from that description that the system was introduced almost 35 years ago.

It’s easier to believe it when considering how EDLs were transferred from the computer of the CMX 600 to the computer of the CMX 300 that controlled the conforming tape machines. The EDLs were punched as holes into a paper tape of the sort used in teletypewriters.

The CMX 600 gave way to the CMX 50, which used relatively inexpensive videocassettes instead of wedding-cake sized hard disks in dishwasher-sized drives. And punched-paper tape EDLs eventually gave way to floppy-disk drives — drives for eight-inch floppy-disks.

Imagine being instructed to re-edit a show from that era, going to the vault, and extracting either a punched-paper tape or an eight-inch floppy disk as the EDL. Where would you even go to get the data entered in your modern computer? In that case, it probably would be easier to read the time codes off a piece of paper and enter them on a keyboard.

Today, we are in the era of tapeless camcorders. Ikegami has hard-disk packs, and JVC has hard-disk backs. Panasonic’s memory-card system is called P2, and Sony’s disks deal with laser-light blue.

All of them (and other tapeless systems) appear to be fine for Act I’s shooting. All of them excel for Act II’s editing. They all provide near-instantaneous access to any point. But how will any of them do with Act III’s archiving?

First, there’s a cost issue. The Betacam SP tape format is unquestionably superior to its Betacam predecessor in both video and audio quality, but, despite the approach of Betacam SP’s 20th year of sales, many TV stations still shoot Betacam tapes. The oxide formulation of Betacam is less expensive than the metal-particle of Betacam SP. The difference is small, but to anyone who archives all news footage, it can add up.

Then there’s the age factor. Old two-inch videotapes can be played today. Will the same be true of optical disks and flash-memory cards fifty years into their future? Will computers a half-century from now recognize today’s hard drives, or will those drives be like the punched-paper tapes and eight-inch floppy disks that no one knew what to do with after half that period?

Mo Henry told Los Angeles magazine in 2002 that she thought her position might disappear within five years. There are still three years to go, so her belief could still prove correct.

In the meantime, however, even producers of computer-generated movies like Shark Tale haven’t given up on negative cutters. Instead of using less of that old craft, they keep asking for Mo.

###

The Cruelest Cut

Early videotape editing was somewhat like negative cutting. Each tape would be played to the desired spot, which would be identified by a grease-pencil mark. If there was an audio cue point, the tape could be rocked back and forth to help identify the exact moment of the cut.

Then the tapes were removed from their playback machines and “developed” by means of a solution of iron particles that would line up over the recorded magnetic tracks, making them visible. Each developed tape would then be placed into a positioning jig, so that a cut could be made between tracks and between frames. When the two tapes were aligned, they would be spliced together.

Only then, after painstaking effort, would an editor potentially discover that 55 minutes of tape on a one-hour reel had just been spliced onto 55 minutes of tape on another one-hour reel, with no way to wind or rewind. That’s when an editor realizes that time can be on your side only if you have sufficient reel space for it.

###