Behind The Scenes in Bologna: The Art & Science of Film Restoration
- November 22, 2016
with Joshua Ng, IT & Technical Executive, AFA.
*This is the second of a four-part blog series as part of Asian Film Archive’s Save Our Film initiative. Read part 1 here.*
During my time at Bologna, Week 3 and 4 consisted of the two-week training program and internship. The staff at the restoration lab, L’Immagine Ritrovata, walked us through the 13 different stages in a film restoration workflow. Each stage has its own department within the lab. In chronological order, the different stages are:
- Film identification
- Film repair
- Film scanning
- Film comparison
- Digital restoration 1
- Digital restoration 2
- Digital restoration 3
- Colour correction
- Sound restoration*
- Film mastering
- Digital asset management
- Subtitling
- Print and processing
* Sound restoration can happen in parallel with the rest of the stages. The restored sound would be combined with the restored pictures at the film mastering stage.
Film restoration is a long and arduous process. It is not unusual for a project to sometimes take up more than 1,000 man-hours just for the film repair process alone. As such, film restoration is a very costly endeavour and choosing a film to undergo restoration is a selective process, both physically and curatorially. Film restoration though is a crucial aspect in film preservation. This is especially so for films whose original elements have deteriorated substantially, for restoring them might be the only chance they have for continued survival. For example, the films Sultan Mahmood Mangkat di-Julang (1961) and Gado Gado (1961) were selected by AFA to be restored in 2015 because the sole surviving film prints were in bad shape. Without restoring them, the prints might decay to a point where it would be too late to do anything.
The class that I found most interesting was “Restoration workflow and lab infrastructure” by Emanuele Vissani (Ema), the supervisor of the Mastering and Quality Control department. He discussed how the lab’s computers and devices were connected, with either 10 gigabit Ethernet (10GbE), 1GbE or Fibre Channel; the kind of RAID (redundant array of independent disks) he was using; the disks he was experimenting with in anticipation of new upgrades; when to use direct attached storage (DAS) and when to use network attached storage (NAS). What to me was most significant was his opinion that the best way to store film is on film stock. Linear Tape Open (LTO) is the runner up, being more reliable than hard disks. But LTO also comes with its own complications as it is designed to be backward compatible with at most two generations read and one generation read/write. As such, while an LTO-7 tape drive can read LTO-5 and LTO-6, it can only write to LTO-6. (It has full functionalities for LTO-7). This means migrations are inevitable every 3-5 years. Technology advancement is ironically a woe to long term preservation.
Ema gave a brief introduction on how he arrived at the type and calculation of the number of devices and computers he needed. The actual calculation involves considering the needs of every department and its predicted buffer. After the class, I spoke with Ema in greater detail, piling him with specific questions. He shared his experience of building the lab from scratch when resources were scarce. It was not easy but they did their homework, took the plunge and survived to tell the tale. It was a very heartening story, and I learnt the important lesson that there is no one-size-fits-all solution. Sometimes, one just must make an informed decision and jump right in. Even now, the lab is still working hard to monitor and constantly trying to optimise and upgrade their system. In fact, to keep up with work demands, Mario, a full-time IT system administrator was hired. Mario’s main task is to ensure everything is up and running 24/7, making sure there is no network congestion. It is not unlike a traffic controller. Given the huge amount of data traversing through the different devices via the different network channels, it is no small feat. With Mario helming IT, Ema is now focusing on research, planning and quality control. Given the speed of technological changes, this is a much-needed area of work.
I thoroughly enjoyed being a film ID detective during the film identification class by Camille Blot-Wellens. The three-step toolset we were given were used to identify, date, and situate a film. Camille demonstrated how we needed to use all our senses to identify the film we have on hand. For the longest time film, could only be duplicated, so there are often clues left behind as to which generation the duplicate belongs to. It was also very important to keep all the accompanying materials that came with the film together. She explained that, “Each element is a slice of history. It’s important to understand why and how each element was made.” For example, the can holding the film reel would usually contain information as to where the film was made, the year, as well as the reel information.
Through this class, I had the opportunity to touch for the first time, nitrate film stock. Not just any random nitrate film, but a 1939 Spanish titled re-release of a 1919 Italian silent film starring Francesca Bertini. I must concur with the general description of degraded nitrate film smelling like a roomful of stinky feet.
While films may just be a form of entertainment to many people, I have realised that film is a snapshot of history and, as some would say, of time itself. A lot of effort goes into producing a film. It captures society at that moment in time through the filmmaker’s lens. For example, local folktales retold through film are often modified to suit the palate of the target audience at the time. Therefore, by studying the film and its variation from the original story, one can deduce the reasons why such modifications were made.
Restoration is new to the AFA since we started restoring films only in 2014. When I was in the SG50 film restoration project team, I realised there were a lot of gaps in my knowledge. Often, I had to research and learn aspects of film restoration from Internet sources. Attending FRSS 2016 has given me an overview of the whole process, from start to end. Equipped with this knowledge and some newly acquired practical skills from Bologna, I can do my job better.
* * *
As an IT trained person, I have always been more inclined towards digital technology. But now I have a better and perhaps, different opinion about film. One of the hot topics discussed at the FIAF Congress and Symposium was the issue of film vs digital. At a film archiving event, you would expect views to lean more towards film, but surprisingly, opinions were quite balanced. There is no doubt that digital technology has advanced the archiving field, but celluloid film at its core is a completely different technology altogether — a horse is not a camel. While film in digital form provides access more easily, it is hard to ignore the fact that film as a format has been around for close to 100 years. Most of Louis Lumière’s films from 100 years ago are still in good shape because of proper storage conditions. This cannot be said for digital formats, as without emulation or migration, some digital born films are already inaccessible. Until digital technology has proven its longevity and archives have figured out how best to preserve our digital heritage, original film elements will not be thrown out.
Original film elements are also important in film restoration. Technology will continue to improve, so every attempt at restoring the film can only get better. However, without the original film elements, there would be no restoration. Ned Price, head of preservation at Warner Brothers Entertainment, shared with us a cautionary tale. During the remastering of a Clint Eastwood film, it was discovered that the digital intermediate was corrupted. Fortunately, the original film elements had been preserved. It turned out that it was easier to work on the original film elements compared to fixing the digital files.
However, this does not mean we should ignore digital technology altogether, especially since most of the materials received from filmmakers are now born-digital. Film archives have a responsibility to keep up with and contribute to the latest digital preservation efforts. Price’s advice is, “Digital is in its adolescence, help it grow, push it to grow.” I concur.
Throughout the Congress, symposium and restoration programme, it was heartening to see how everyone were so willing to help each other out. We shared tips, documents, and promised to keep each other updated with the latest progress in digital preservation.
Digital technology has caught the film archiving community by storm. AFA is not alone in this!
* * *
About the FIAF Film Restoration Summer School (FRSS)
FRSS takes place every two years. The next one will be in 2018. You can look out for the announcement on the FIAF website. As a FIAF event, priority is given to staff working at film archives, institutions or similar organisations, as well as students in this field. For more information, please visit: About the FIAF Summer School
*This is the second of a four-part blog series as part of Asian Film Archive’s Save Our Film initiative. Read part 1 here.*
