News

The Ultimate Challenge – Available Light Camera Test

Writer/director Kate Dennis and DOP John Brawley shot six shorts with six available light locations on six different cameras. Ross Mitchell, post-production manager at Deluxe helped them in this adventure. This is their story.

John BRAWLEY: Kate and I were working together on the first series of Offspring. We had some night exteriors to shoot in the backstreets of Fitzroy, and we were both amazed at how bright the street lighting looked when shooting the RED MX camera using fast prime lenses. Kate has a film in development and she was concerned about how to shoot on the streets of Paris using only existing or available light. In the context of a wider discussion I’d been having with her about the differences between both film and digital acquisition, I thought it would be really interesting to test them all.

KATE DENNIS: Being ex-camera department, I had inherited some scepticism regarding the digital formats, but here was a camera that could capture those night exteriors without a great deal of additional lighting so it made me realise I had to open my eyes to digital. I’d seen a few tests online, but what I hadn’t been able to see was something that compared film and DCP projection over a variety of cameras. For me, the test was beyond test charts, dynamic range and resolution comparisons; it became about an emotional connection to the format. John and I were ultimately both more interested in de-intellectualising the test process, revealing which format each of us had an instinctive preference for.

JB: We wanted to have the storytelling requirements inform our choices about how to shoot on the day. It would have been meaningless to go out and shoot landscapes of the city at night, so we realised we had to have actors in it.We decided to organise an available light shootout, using 35mm as a reference and the two leading digital cinema cameras that were available to us, the Red MX and the ARRI Alexa. I was also interested in looking at super 16 as well, so we threw that into the mix.

Since DSLR cameras like the 5Ds are a hot item at the moment it seemed natural to include them as well. I chose the 1D mkIV instead of the 5D as I’ve found it a smidge more sensitive in low light. These five cameras were supplied by Lemac and, at the last minute, Sony came to the party and supplied us with an early engineering sample of the PMW-F3 to round out our six cameras. It was confluence ofthe number six; we had a crew of six, six cameras and shot in six locations.

Kodak came to party and donated some film stock for us. We invited Deluxe to be involved because we wanted to take this test all the way through to the ideal viewing environment: the one you get in a cinema. In this era of Vimeo and YouTube I get sent a lot of tests or examples from cameras to look at, but they are always from online sources; you just never get to see them on a big screen and directly compared against the same photographic content.
I wanted to see them finished to DCP and to 35mm print, because digital projection is certainly more and more prevalent and within a few years I expect it will replace 35mm release prints for distribution. It was critical for us to compare film originated images and digitally-originated images on both film and digital projection systems; Deluxe took care of all the post, including the film print.

Night of the 168 Shots
JB: Kate wrote the script to suit the locations we had chosen due to their available light characteristics. We then staged the actors (Jane Harber and Mike McLeish) in that space and decided what the coverage would be for that scene. Cail Young acted as a data wrangler and assistant with all the formats.
Just like a regular shoot we went through each setup, doing six takes, one for each camera; we used the same lens (Cooke S4, from Inspiration Studios) and simply switched it between cameras. We were literally shooting with a ratio of 1:1 and rarely had enough time to do more than one take for each format. We usually shot the first take with the 1D to make sure it worked on that first; it was the only format that would get more than one shot, as we’d sometimes tweak the shot or action before locking it in.

Each short ended up with about 28 shots in it. That’s 6×28 shots, so we managed 168 takes over the night, which is pretty impressive considering the size of our crew. We were shooting without permission, and just moving around on foot.

KD: The test is really a dramatic sequence, not a short film. We had no location sound and tried to tell a simple story that would work with music. The end result is six almost identical sequences. It was all handheld, pretty mad in fact, but kind of liberating to be working at that speed and with such a small crew.

JB: There are some small differences in the edit, mainly because we wanted to make each one look the best it could in terms of story, rather than religiously matching each cut. Then the wonderful colourist Stanly Lopuszanski from Deluxe spent hours of his own time trying to get the most out of each format; we wanted each format to have its own characteristics and play to its strengths.

Changing Perceptions
ROSS MITCHELL: John is always striving for a method of making things work that little bit better, and I was sure the collaboration would produce results to benefit John, Kate, Deluxe and also the greater industry. The proposal also sounded like a great opportunity to test our own workflows and give us some valuable data on how the same job coming through on six different formats would compare in the amount of resources required to produce a final image we’d be prepared to exhibit.

Deluxe committed to offering all of its available services from processing the neg, telecine transfers, Avid Symphony for Ben Joss to cut on (FCP was used to cut the F3 footage), ARRI scanning and recording, Lustre grading, sound mixing (courtesy of Deluxe StageOne Sound in Sydney), DCP encoding to film printing. Inspiration Studios were great in supplying all of the data acquired from the digital cameras already wrangled onto a single drive meaning we could get on with ingesting into Avid for Ben to cut.

The film rolls were ingested to Avid via the well worn traditional SD telecine – digi beta path. Only the F3 footage was cut in Final Cut purely for convenience of resources that particular day. The F3 footage was never compressed further for offline editing so the cut sequence was played straight out as a DPX sequence (with a mild colour space secret sauce) and sent to Lustre for grading. All other formats went through an online/conform stage either in Lustre (35mm & 16mm) or Clipster (Alexa, Red MX, 1D) until all 6 sequences sat in Lustre as 10bit log DPX files. Stanley Lopuszanski then conducted a series of grading sessions with Kate and John in attendance to achieve the look that finally went out to film via the ARRI laser recorder.

And the winner is…
JB: I was really surprised that the differences between the DCP and the film print itself introduced. Given that DCP is likely to be the prevailing exhibition standard within the next few years, I wanted to see how the film and digital cameras compared.

When I actually watched the DCP go though, I was pleasantly surprised at how well the two high-end digital cinema cameras looked; they were very close to 35mm in terms of dynamic range and colourimetry. There were only very small differences in the image fidelity between the Aaton 35mm camera, the Red MX and the Alexa when screened from the 35mm print. Most people would have difficulty in picking which was which. It was a little easier to pick them on the DCP, mainly because the Aaton 35 had tell-tail grain. The digital cameras were much cleaner, even in low light.
I wonder if the viewers’ perception of image fidelity is changing. It seems some find grain to be an offensive trait; I however find that once you’ve seen a couple of shots, the grain seems to disappear into the image. I like the texture you get from it. Most found the super 16 to be too much, but I believe it had a great look all of its own, which would be entirely appropriate for the right kind of story material. It shouldn’t be written off at all.

I expected that the film-originated material would have not been able to compete with the digital wonder cameras. This was a low light torture test, sort of a worst case scenario with the worst kind of uncontrolled lighting conditions. Surprisingly the 35mm film seemed to have the edge, but only just… there’s so much mythology about how good these digital cameras are in low light, but there was the Aaton 35 looking just as good. The downside, of course, is the cost, and for Australian productions that’s very relevant. I costed each format out and the 35mm was nearly double the cost of the Red to shoot, and about 30 percent more than the Alexa.

It’s hard to quantify the image differences, and it’s up to each filmmaker to work out if they want to pay the extra price to get that little bit extra that film can still deliver even in these circumstances. Otherwise the Red MX and the Alexa get you very, very close for a lot less outlay; at the screening, the consensus was that most preferred the film material overall, but only just. Many were surprised at how close the Alexa and the Red MX were in a direct comparison.

There was another difference with the two smaller cameras, the Canon 1D and the Sony F3. They are smaller and lighter, so it’s actually much harder to hand-hold these cameras and you can really tell on the big screen. The 1D had a sort of jitter that was quite pronounced from being hand-held. The larger cameras, which tend to sit on your should rather than being held in front, were more steady when being hand-held as well.

In a way there was no real winner for these tests. You could envisage scenarios where all six of these cameras could be used to shoot a film for a cinema release; the decision about which format to use just depends on the story requirements, budget and resources available.

KD: I had expected the Alexa to out trump the Red, but no… I actually find it difficult to call it the clear winner. I’d have to say that film projected on film is still my favourite, but given that digital projection is our inevitable future, that is almost a moot point now.

The digital formats look so good projected digitally; it is hard to go past them. I wonder if, in fact, in a few years time, our eyes will be so accustomed to the cleanliness of a digital image, that 35mm will feel like 16mm does now, shockingly grainy. The one thing I do hope improves is the way the digital cameras respond to horribly flat daylight, because film still wins hands down in that regard.

For more information about further screenings, visit http://johnbrawley.wordpress.com

ADVERTISEMENT

Get the latest media and marketing industry news (and views) direct to your inbox.

Sign up to the free Mumbrella newsletter now.

 

SUBSCRIBE

Sign up to our free daily update to get the latest in media and marketing.