A Call to Arms: Captain Corelli's Mandolin
By Stephen Pizzello

This article originally appeared in the April 2000 issue of American Cinematographer Magazine.

In Captain Corelli's Mandolin, World War II provides the dramatic backdrop for a touching love story set on the idyllic Greek island of Cephallonia. Adapted by screenwriter Shawn Slovo from the critically acclaimed historical novel by Louis de Bernieres, the film recounts the Italian occupation of Greece and its harrowing aftermath, during which Nazi troops were eventually deployed in Cephallonia to attack their former Axis partners.

The story's critical view of warfare is personified by its reluctant protagonist, Capt. Antonio Corelli (Nicolas Cage), a mandolin-strumming artillery officer whose exuberant spirit initially elicits the disdain of the natives -- most notably Pelagia (Penelope Cruz), the beautiful daughter of a respected village doctor (John Hurt). Educated and strong-willed, Pelagia is engaged to a local fisherman (Christian Bale), but she gradually succumbs to Corelli's charms after the officer is billeted at her father's home. When the strife on the island begins to escalate, however, the lovers are forced to weigh their feelings for each other against their ethnic allegiances.

An A-list project all the way, Captain Corelli's Mandolin teamed Oscar-nominated director John Madden (Shakespeare in Love) with two-time Academy Award-winning cinematographer John Toll, ASC (Legends of the Fall, Braveheart). Their efforts were bolstered by a top-notch team that included production designer Jim Clay, supervising art director Chris Seagers, costume designer Alexandra Byrne and makeup artist Lois Burwell.

When AC phoned Toll to request an interview, he was busy shooting six-day weeks in Los Angeles on Vanilla Sky, a project that reteamed him with director Cameron Crowe, with whom he had collaborated on the critically acclaimed Almost Famous. Despite his hectic schedule, Toll graciously spent part of a Sunday-morning respite answering questions posed by executive editor Stephen Pizzello. (Editor's note: Although Captain Corelli's Mandolin was initially scheduled to be released in April, the date was pushed to August at press time.)

American Cinematographer: Did you read the novel by Louis de Bernieres before you began working on Captain Corelli's Mandolin?

Toll: When I knew I'd be shooting the picture, I started to read the novel, but I quickly decided not to finish it – I knew right off the bat that the film would only be able to cover about 20 percent of the material in the book. On previous projects, I'd gotten very involved in the book, only to realize later on that we weren't really shooting the book. This time around, rather than be distracted by all of the great things in the book that wouldn't be part of the film, I opted to approach the work purely as a film that was simply based on the novel. Otherwise, I might have found myself second-guessing various decisions or trying to talk somebody into including a certain scene from the book. The director, John Madden, and the screenwriter, Shawn Slovo, already had their hands full doing the adaptation, and they didn't need my involvement as well.

Were you still able to draw inspiration from the novel?

Toll: The book had a tone of 'magic realism' that I found interesting, but both John and I were drawn more to the story's realistic elements. That approach just felt right to me, because I felt that if we leaned too much toward the 'magical' elements there was a danger of making the film too sweet. Those aspects of the story worked well in the book, but when making a film I think you sometimes walk a very fine line between what is perceived as ‘real' and what isn't. It is a wonderful book, however, and the island characters are fantastically drawn.

John Madden came onto the picture after the original director fell ill. How did that development affect the production?

Toll: John came aboard on relatively short notice, but the preparation of the picture was already well underway; they were locked into a shooting schedule because of Nic Cage's availability. The English production designer, Jim Clay, had already made plans to build sets on the island. When John took over, he met with Jim and together they came up with a final plan for locations and which sets would be built.

I got involved with the picture after John had signed on. I met him in February 2000 and went to Greece in March on a location scout. I then came back to LA for awhile, and we began shooting on Cephallonia at the end of May. I like John as both a person and a director; he's a very intelligent, personable guy, and he has very good taste.

What kind of collaboration did you have with him?

Toll: We'd sorted out our relationship before I even took the picture. We had a great meeting in LA, and we also spent a lot of time on the phone. He would tell me his ideas, and I would tell him mine, so we knew what we were after by the time we scouted the locations. We had a similar take on our general approach to the work, so it was just a question of going in and finding the details. John had a lot of specific ideas, and he thinks about the whole picture. He realizes that in addition to the story and the performances, you also have to factor in the production design and the overall visual context. In that sense, he's a total filmmaker, as opposed to, say, a writer who might be most focused on seeing his dialogue performed onscreen.

Cephallonia is where the events depicted in the book actually took place. How did the authenticity of the location affect the narrative?

Toll: Since John decided to shoot the story where certain aspects of it actually happened, we were automatically thrust into the more realistic aspects of the story, which validated our approach to the material. There were a lot of older Greek people around who confirmed that the book was pretty accurate; they told us that yes, Germans had taken hundreds of Italians out in the field and machine-gunned them to death.

A big part of the film's story is about how well the Greeks and the Italians began to get along, even though the Italians had invaded the country. There are many similarities in the two cultures, and Cephallonia had been occupied by the Venetians for a few hundred years. The Italians were on the island for almost three years during World War II, and once they settled in, they mingled well with the Greeks. In fact, when the Italians surrendered to the Germans, some of them tried to hide, and many of the Greeks tried to help them.

Today, Cephallonia is a very popular destination for English tourists. They even have Captain Corelli tours that visit all of the places in the book. I'd never really spent time in the Mediterranean, but I liked the idea of shooting there, and I also liked the idea of doing a love story set against a big historical event.

Given the nature of the story, I assume that you set out to exploit the island's natural beauty.

Toll: We did, but I wasn't interested in presenting it in a pictorial, 'beautiful' type of way. I was trying to go out of my way not to take that classic sort of approach. For example, the water usually has a very vibrant and saturated blue/green color that definitely gets your attention when you see it, but on film I found it to be almost surreal and unattractive in a certain way. In contrast to that, the landscape was very mountainous and had a feeling of being rugged and harsh, which I thought was quite interesting.

We wanted to develop a look that worked against the picture-postcard aspects of the settings. We felt that if we went with a simpler and more realistic approach, the audience could appreciate the predicament of the characters more than if the settings were 'magical.'

What was the quality of the light in Cephallonia?

Toll: The day exterior light was extremely contrasty. We were shooting in the summer months, and the island is pretty far south. We usually had clear, bright, blue skies, and we actually went through a period where I didn't see a cloud for six or seven weeks! It was also very hot, usually over 100 degrees, sometimes as high as 115 or 120.

Because it gets so hot and dry during the summer, the landscape becomes very stark-looking –

almost like a desert at times. The sun is very harsh, and there's this wild contrast that exists between the land and this huge expanse of intensely saturated blue water and sky.

When we worked outside, it would usually be for a whole day, and we were faced with the usual problems of lighting continuity as the light changed throughout the day. We were also shooting in anamorphic widescreen, so we usually found ourselves placing characters in big settings, which made controlling the light difficult or impossible in the wider shots. I actually suggested shooting in the 1.85:1 format, but John liked the idea of using the compositional opportunities of 2.35:1. I felt it might be interesting to work against what you might expect from a picture of this kind, and that it might be good if the picture felt smaller and didn't have a sense of being a 'big picture.' I thought the picture would have plenty of scope in any format, and I didn't think we needed the extra help of the 2.35:1 ratio.

When I suggested this, I knew that both John and production designer Jim Clay thought I was completely insane, but they very politely listened to me. Then John said that he really liked the widescreen format and thought it could work well for us. Now, having shot the picture, I find it impossible to see it any other way.

How did the anamorphic format affect your shooting and lighting?

Toll: The anamorphic format didn't change the day-exterior work much. It probably made light control more problematic, but it would have been difficult in any format. We probably had to work harder to use the frame well. When you shoot in anamorphic, I think there is a real obligation to fully use the frame, and sometimes compositions become a little more studied and formal because of the extra width. I think you also have a tendency to move the camera in different ways than you would in other formats. These are not necessarily bad things, and most of the time they're very interesting and exciting. To me, it just means that you need to be extra conscious of how the whole frame is working all of the time.

For the night work and interiors, I worked at a higher light level than I would have on a spherical picture. In lower-light situations, I find the lack of depth of the anamorphic format a bit distracting.

What did you do to try to combat the island's contrasty conditions?

Toll: I actually became attracted to the stark, realistic feeling of it. Rather than trying to minimize the contrast, I tried to use it as much as possible. After a period of seeing many cloudless days, I could pretty much depend on consistent sunlight and, like it or not, that was the reality of the island. However, I actually did think it was interesting. Seeing very intense sunlight glaring off the rocks and dusty roads seemed right for the feeling of the place. It just became a question of how to make it work for the picture, but even in direct, overhead, midday sunlight, there is usually a way to make the light work for you. We were careful about how we staged sequences and planned our angles of coverage, so we were usually shooting in a direction that was best for us. John was terrific with all of that, because it required careful planning in staging the scenes, and he worked very hard to help make it work.

Lighting continuity was also a concern, because we would often be in a sequence for a whole day or over several days. We therefore needed to plan our angles to avoid obvious mismatches from shooting at different times of day. The close-ups could also get a little tricky at times, because it meant manipulating the light on the actors' faces to maintain the feeling of harsh sunlight, without some of the unflattering characteristics of that look.

Did you use any filtration or special processes to deal with the contrast or enhance the look of the film?

Toll: I decided not to; I opted to play the realism and not try to over-romanticize the look. I did do a lot of testing, and at one point I actually contemplated exaggerating the natural contrast of Cephallonia. Before I went to Greece, I did some tests at Deluxe Hollywood with Beverly Wood. I tested the CCE process simply because I wanted to leave no stone unturned. During those tests, I used just about every emulsion that Kodak manufactures, just to see what would happen. I thought the CCE look had some possibilities; I considered really going for shadow detail and very hot highlights, and when I tested that approach it looked pretty interesting. I shot some contrasty situations in LA, and when I went back to Greece during prep I shot some additional test footage there in even more contrasty situations. I then took that footage to Deluxe London, where I experimented with different looks.

When I shot in Greece, I used some local people from the island as stand-ins for Penelope and Nic. The woman especially looked a lot like Penelope and had a very similar skin tone. When I began to see how the CCE process was working on the skin tones in close-ups, I started to think that it probably wasn't the right look for this picture. I thought the process itself was very interesting, but it appeared to be a little harsh and unflattering to the olive skin tones that most of the actors in this film have. I decided it wasn't right for us and went back to the normal print processes.

However, I did find myself wishing that I had a way to minimize the color saturation of the existing print stocks. I find that the Vision print stocks have quite a bit of contrast, so I thought that I didn't really need the CCE process to add much more contrast to the image. But I kept thinking that it would be great if there were an alternative to the existing color-saturation levels of the current Vision print stocks. I think both of the Vision print stocks are great if you're looking for very rich blacks and saturated color, but every picture doesn't necessarily want that look. I wanted to keep rich blacks, but I also wanted less color saturation, and there seemed to be no way to do this. I tried various types of filtration, but I wasn't happy with the way they changed the overall feeling and look of the picture.

Which film stocks did you end up using for Corelli's Mandolin?

Toll: I tested with Kodak's medium-speed stocks, but I wound up shooting most of the picture on Kodak's pre-Vision, slow-speed (100 ASA) 5248. To my eye, it just looked the best; the shadows looked deeper, and the highlights had more detail. I often tried to blow out the highlights and make them disappear, but when I wanted them to have detail, they did. For some of the interiors and night exteriors, I went with the Vision (500T) 5279, because we were shooting anamorphic and I needed the stop. But I found that the 5248 worked very well as a general, all-around stock. It looked great outside in all types of lighting situations. It has a great range of detail, and I found that it worked well with the contrast of our location and the black levels of the Vision stock. I know that many cinematographers have begun using the wider-latitude, medium-speed negative emulsions like 5277 and to help soften the contrast of the Vision print stocks, but I found that the 5248 worked well, especially in the highlights, and maintained a wonderful sense of richness without sacrificing any of the blacks.

Do you ever worry when you have to mix different emulsions within a picture?

Toll: I never used to, and in the past I have mixed emulsions in the middle of individual sequences. I used to hear people talk about the grain structure of different emulsions being a problem, but I never found that to be the case. These days, however, I do have more concern about mixing negative emulsions. I think this has come about because of the print stocks. It's my feeling that the change to the Vision print stocks has changed the apparent characteristics of some of the negative emulsions. In particular, I feel that the higher-speed negative emulsions look more contrasty and feel harder and sharper-edged than when they were first introduced and when we saw them printed on 5286, the previous version of print stock. In making negative-comparison tests, I see more of a difference between the current negative emulsions than I did before we began using the Vision print stocks. As a result, I'm now more sensitive to mixing negative emulsions within a sequence; in fact, I try to stay with one emulsion as much as possible while shooting a feature.

You were working with a British director, so did you use the English crew system?

Toll: I like working in the American system, but I've also done a lot of pictures out of the country and I can work both ways. My crew was actually a great mix of UK-based, American and Australian people. I took my American assistant cameraman, Christopher Toll, but the rest of the camera crew was English. The camera operator was Peter Cavaciuti, and the B-camera assistant was Graham Hall, who had worked with me on Braveheart. The grips were mostly English and headed by Kenny Atherfold; the lighting department was Australian and included gaffer Mick Morris, with whom I'd worked on Wind and The Thin Red Line. We organized the camera department in the American system, and the grips and lighting department worked in the English system. John Madden and I would work out the shots, as we would in the American system, and everything fell into place from there. It worked quite well.

Why did the production build such extensive sets on the island? Was it too difficult to create a period look with the existing architecture?

Toll: There was an earthquake on the island during the 1950s, and it completely destroyed about 90 percent of the buildings there. As a result, nearly all of the architecture that now exists was built after the quake. The buildings that were there before the quake were really beautiful, and the architecture had an Italian influence because the Venetians had been on the island from the 1600s through the early 1800s.

In that respect, the big problem with our location was that all of the elements that would have helped us recreate the historical context no longer existed. The only thing about the island that helped us, in terms of creating a sense of authenticity, was that it was the same place where these historical events had occurred. It was a fairly large island, so our choices about where to place the sets and characters became the most important consideration. The main town in the story, Argostoli, is on the southern part of the island, and it's a really busy place. Since the town no longer had any actual World War II-era architecture, we opted to find another area to shoot in.

Some of the movie takes place in a small village where the family lives, and other scenes take place in the town where the Italians set up their base of operations. We therefore needed to create the village and the town with sets that resembled the original Venetian architecture that still existed during the 1940s. We also needed a seaside port for arrival scenes of the Italians getting off their boats. We eventually found a suitable town called Sami on the northern coast of the island. Jim Clay worked out a way to re-clad the fronts of the existing buildings to create authentic period facades, and art director Chris Seager supervised all of the detailing. They eventually built three main, interconnected streets, a town square and a village, which involved an incredible amount of work. We began work on the picture in the town, and we shot there for about a month. The sets really gave us a sense of the period, and they were so complete that we could essentially shoot in 360 degrees.

Jim did a fantastic job with all of his work on the picture. In fact, the sets were so good that a lot of the tourists couldn't believe that the buildings weren't made from authentic materials. While we were there, people even started shooting postcards that featured our sets! Even after we ripped the sets down, tourists kept showing up looking for 'Old Argostoli,' because they'd heard it was up there! (Laughs.)

Did you light any major night sequences in the town set?

Toll: We did a big waterfront party sequence in the town. There was a large dance floor and a couple hundred extras. I wanted to keep the night sequences warm in color, so I used tungsten-balanced light and stayed away from a cooler night look. The warmer look seemed more appropriate for this picture. We hung electric party lights over the dance floor and used those as motivation for general ambience. We lit mostly with 20Ks, aimed directly through layers of diffusion when we had enough room, or bounced into large muslin frames when we were in tighter places.

I usually shot at T4 or 4.5 because of the number of people in the shots, and to facilitate some depth of field. There were some big shots with quite a few people, and I was trying to keep backgrounds and foregrounds as sharp as possible. There was also something about this picture that made me want to shoot with more depth of field; I think it was the period aspect and the fact that the sets were so detailed and interesting. I often used wider lenses to try and show as much of the sets as possible. Panavision's 35mm anamorphic lens was one of our workhorses, and it presented us with many compositional opportunities. It's a wide lens but it doesn't have an exaggerated perspective, and we kept using it to place