In January 1988, Susan Dalton of the American Film Institute's Centre for Film and Video Preservation inquired about the possibilities of the National Archives of Canada assisting in the restoration of Maurice Tourneur's 1915 film "The Cub". As the only existing copies of this title were on 28mm prints at the Library of Congress and George Eastman House, the request was directed to the National Archives of Canada since we possess a 28mmm wet gate movement for optical printing. This gate was manufactured to copy a large collection of films distributed by the Ontario Government Motion Picture Bureau (1917-1934) that were acquired by the National Archives in 1970. After the funding arrangements and time estimates were resolved, this unique joint restoration project was approved by Dr. Jean-Pierre Wallet, the National Archivist of Canada.
The two 28mm prints arrived in separate shipments during the last week of March. Both were fairly worn projection prints, each copy comprising five reels. The condition of both copies was fair with extensive broken perforations, some tears and numerous splices in poor condition. Both prints were brittle with a variety of physical distortional problems including curvature, buckling, and edge fluting. The shrinkage on both copies was very low, averaging .4 to .5 percent. Both prints were inspected on hand rewinds, repaired as required for initial inspection and leadered. At this point, is should be noted that the National Archives possesses a complete range of 28mm ancillary equipment including split reels, cores, splicers, synchronizers, and a hand-powered viewing apparatus. All of this equipment has been modified from 35mm to 28mm format. The leader is from a precious discovery of unused 28m print stock found a few years ago.
Following this initial inspection, a more detailed comparison inspection was made using the double- gang synchronizer and viewer as necessary. This scene to scene inspection was necessary to determine the optimum quality and length of each shot in both 28mm copies. The reconstructed version was made from both 28mm prints. For reels 1,2,4 and 5, the ratio of shots taken from each copy was roughly equal (i.e. 50% of the scenes from the Library of Congress print and 50% of the scenes from the George Eastman House were used.) Reel 3 however was copied entirely from the Library of Congress print.
During the reconstruction exercise a detailed shot list for the first two reels was discovered in the Library of Congress copyright records. Although incomplete, the list was helpful in confirming the order of the shot sequence. Essentially both original versions were intact and in sequence. The only variation was reel five where the George Eastman House print was obviously in the wrong order. Neither version had a final "The End" title. Since the final fade-out shot is clearly the end of the film, this lack of an end title may have been intentional, but one can only surmise.
The reconstruction was listed out frame-by-frame for the camera operators on the optical printer. In addition to basic details such as light grading and shot length, the timing sheets indicated from which 28mm print each shot should be taken and any special restoration effects such as fade outs, iris shots and freeze frames. These techniques were utilized to restore shots where damage or incompleteness of the 28mm scenes was noticeable. For example where an iris in or iris out was incomplete (due to a splice, for example) a special adapter built to simulate the original iris effect was mounted on the printer and the shot rephotographed frame-by-frame using the remaining available frames in and around the "original" iris shot. Similarly, particularly bad splices and misframes were eliminated optically. The misframes in the 28mm prints had a streaked vertical blurring which can probably be attributed to the original 35mm to 28mm printing.
The actual printing went fairly smoothly. Aside from the fact that the operators were constantly changing the 28mm originals, the main problems encountered were the brittleness and curvature of the film. The problem of bubbles in the wet-gate solution (perchloroethylene) was aggravated by the film curvature. Often a first pass of each shot was required without the camera operating so the technician could check the registration and ensure the gate remained bubble free. Since the aspect ratio of 35mm full frame, 28mm and 35mm academy aperture (the format chosen for the restoration) are all slightly different, the framing was critical in order to avoid any further cutoff of original picture area. It was clear that some image was already lost in the original 35mm to 28mm printing. This combined with the new enlargement to 35mm probably amounts to a cutoff of 5% of the original vertical image. This is apparent in some scenes.
The new camera stock was 35mmKodak 5234 Black and White dupe negative which was processed at gamma .55 to reduce contrast inherent in optical printing. The 28mm prints contained both black and white and tinted scenes. Since the actual number of tints was limited, it was decided to shoot on black and white negative and attempt to match the tints on a colour print stock. This technique offers some cost savings on negative stock and also allows for a black and white print to be made if desired. The colour matching was somewhat more difficult, but our past experience has shown very successful results. The completed negative was taken to the Film House laboratory in Toronto where the matching of the colour tints was supervised by technicians from the National Archives and the Head Timer at the laboratory. The 28mm prints were viewed over a lightbox balanced to match the illumination of the Hazeltine colour analyzer. Each scene was corrected for colour and density to match the tinted original and ensure that the black and white scenes printed up properly on the colour print stock (Kodak 5384).
Both the negative and the new prints were inspected to ensure that the residual hypo levels were within archival tolerance.
Restoration Credits - National Archives of Canada
Project Coordinator: William O'Farrell
28MM Inspection, Timing and Reconstruction: Luc Morisset
Optical Printing: Richard Tremblay, Scott Low
35MM Quality Control and Preparation: Steve Segal, Dennis Waugh
35MM Colour Control: Dennis Waugh, Dale Gervais, Chris Severn (Film House)
Processing and Release Print: Film House, Toronto , Canada
- A Short History of the Paper Print Restoration at The Library of Congress
- Optical Printer Article from SMPE 1940