I had a great time in Lofoten earlier this month — more precisely at a surfing spot near a village called Unstad, lighting a short film for my friend Torkel Riise Svenson. The film is about a surfer who is forced to move away from the beach by the local authorities because they want to make way for German tourists, but of course, he’s not going to make it that easy for them.
I won’t spend too much time writing about HDSLR on this site, but I am interested in finding out what they’re good for. Unlike others, I am a bit skeptical to the current generation, and after testing them on a few projects, I am still convinced that this is nowhere near what you would expect from a professional digital motion picture camera. However, it is a very affordable tool with a specific aesthetic (due to sensor size) that few other solutions in the same price range can offer. This is my take on the format and how to avoid the biggest issues associated with it.
3D is coming, now also to Norway. I spent Saturday with graduating cinematographer Kristoffer Archetti, head teacher and cinematographer Kjell Vassdal and a slew of other experienced colleagues at Bislett Stadium, shooting Kristoffer’s final project at The Norwegian Film School with a couple of REDs and the STEREOTEC 3D-rig, recently brought to Norway by ULTIMAX 3D and director/cinematographer Morten Skallerud.
How do you approach the shot making process when working with a director on a film project?
When most young directors say they want it to have a very improvisational feel, generally it goes: ‘Fuck you./No, fuck you./No, fuck you./No, fuck you.’ And then someone pulls out a gun.
— Tim Roth
I’ve reflected quite a bit on this lately, and found that I don’t really have an answer. There are endless opportunities to decide what you do (and what you don’t) show the audience, and my task is to help the director narrow it down to the one choice that fits the project best. Of course there is no such thing as the one right choice, but there are plenty of awful choices available. And no two directors with similar preferences, neither in style nor in communicative preferences.
It is also difficult to articulate the subtleties in cinema, because there aren’t words or metaphors which describe many of the emotions you are attempting to evoke.
— Conrad Hall
An early experience, photographing a short film and then watching the final edit weeks later, is that none of the amazing setups I convinced my creative partner to shoot have made it into the actual film. Every book on cinematography I’ve ever read have stressed this, but I’ve still made the same mistake countless times. If the director doesn’t understand why we’re doing it, it’s a major waste of time and energy. Communicating about something as abstract as a mental vision of an image is a big challenge, but the times when I’ve been able to craft a wonderful shot in tight cooperation with the director, are the times when I’ve usually seen it on the big screen afterwards.
As one of the editing teachers of my school so wonderfully puts it: “The language of film is really quite simple”. Every image needs to drive the story onwards, or it’s just a distraction that will disappear once the director is left alone with an editor. (And rightfully so).
I never want to feel like the way that I see it is the only way. Sometimes mistakes happen and that’s better than what you thought the scene could be. You allow room for the possibilities.
— Steve Buscemi
Another realization I’ve made is that however well planned and coordinated I am with my director, nothing ever plays out the way I thought it would on set. Something unexpected happens, a better solution becomes apparent, there hasn’t been a single project where the original plan was a perfect fit. I’ve narrowed it down to two scenarios, the one where we try to force the scene to unfold the way we planned, and the one where we pick up on what is happening and adapt to it. The first one has most of the time turned out to be an awkward solution in the film. Pre production is a wonderful creative time when I try to get inside the head of my director, and shape fragments of inspiration and ideas into a sense of mood and style, but production is where the actual film unfolds in front of the lens. And as my teachers in cinematography keep stressing, you have to be perceptive and adaptable. But in my opinion also bold enough to follow through on the original vision of the film, when required.
The fine balance between what we planned to do and what the situation is telling us to do is maybe where something interesting happens. In any case my director is my best friend, and if the film is shit, I’d feel better knowing that we kept an open mind throughout the process and made the crap together, as a team.
We mystify the art of moviemaking, but it’s not a mystical science. You take a good screenplay, put a group together and you hammer it out.
— Bill Paxton
I’ve spent a few blogposts writing about the exciting advances in digital acquisition, so I think it’s time to give some much deserved attention to the beauty of physical image capture on celluloid, also known as film. Even after all these years of bits and bytes, it’s still the absolute king of the hill when it comes to simplicity, quality and pure brilliance.
It’s difficult to not fall in love with a film camera once you’ve seen it in action. Essentially a piece of metal with a hole in, this fine piece of machinery will pull the film forward, stop and hold it absolutely still, before pulling it forward again multiple times per second, exposing it to the focused light coming through the lens. No advanced electronics and software, no rolling shutter issues, no bitrates or different formats, no complicated backup schemes, just light trapped in a physical object you can hold and look at. So simple!
The marvel of this machinery is only surpassed by the absolutely gorgeous images it can produce:
The reason why I’m dedicating a post to the film medium, is because I’m currently stranded on a remote island in Norway, pulling focus for one of the exam film productions for the final year students of The Norwegian Film School. We’re burning 16 and 35mm, and I would never want to work with anything else for this particular project. When snow and rain is hitting you at the same time at 20 meters per second, there’s nothing more reassuring than an Arri chugging along beside me. My biggest worry is keeping the lens clean.
But even when the weather is nice, I’d rather be loading my mags with film. It’s a shame that the economical reality of the Norwegian film business is pushing the vast majority of productions into the digital realm. It’s not just a matter of practicality or costs, it’s a major aesthetical decision.
An example; we’re standing on an exterior location with a 100-or-something meter long chain of 15 watt bulbs in an evening landscape just after sunset with a huge miniature in the foreground of a larger construction in the background, and light is falling fast! By the time we get the shot in the can, the light is not reading on the incident meter anymore. But watching dailies afterwards, the 35mm proves it’s worth the extra pennies, as an amazing scene unfolds, with clarity and contrast a RED could only dream of capturing. The entire crew was stoked, and this was only a lousy firstlight to DV PAL.
I am very lucky to be part of a film school that still teaches this soon to be ancient technique to the next generation of Norwegian cinematographers, and will always go to great lengths to pursuade producers if I think it’s the appropriate medium for the project.
It feels like being late to a party and trying to pursuade the remaining few to stay up just a bit longer.
PS: The movie I’m referring to is titled Tuba Atlantic, directed by Hallvar Witzø and shot by Karl Erik Brøndo, both graduating from the Norwegian Film School this summer. It is their final project, a 25 minute short film, and will air on Norwegian television and at the Norwegian Short Film Festival this summer.
WARNING: POSSIBLE SPOILERS AHEAD.
Did he do that on purpose? I’m still wondering, and looking around on the internet hasn’t offered much of an answer. I’m talking about Shutter Island, and whether or not the use of greenscreen was purely artistic or more of a practical choice.
It’s not unusual for film and TV drama these days to use a lot of greenscreen, the technology has become very good, and the practical and economical benefits can be quite tempting. Just take a look at this montage on kottke.org: http://kottke.org/10/02/green-screened. But Shutter Island has an unusual amount of, dare I say, seemingly unnecessary greenscreen-shots for a big budget period drama from one of the biggest names in modern film.
The first and most obvious example is the opening scene on the boat. I can understand why Scorsese would shoot it in a studio for practical reasons (although I’ve seen this done a lot better!). Or the scene in the car after Leo has met with Solando. It’s easier to control, and the actors can concentrate on the dialogue. But why do it in a perfectly normal hallway in a location you’ve already shot lots of pictures in? If you start looking for it, you will find a lot of shots that will get you wondering. And sometimes it almost feels poorly executed on purpose. By the way, did you notice when Chuck hands the woman they are interviewing a glass of water, and they cut to a closeup of her lifting the glass, there’s not actually a glass in her hand? It’s empty! There are small glitches like this throughout.
The visual style of the film, with it’s unconventional angles and abrupt movements, combined with the sometimes disturbing use of sound and music, is all part of giving the impression that there’s something horribly wrong. Which there is, as Scorsese masterfully lies, then lies again, spinning a web so thick with doubt and suspicion that you don’t anticipate the final twist at all as it hits you in the face (unless you read the book of course).
So here’s the theory: What if the glitches and the sometimes poor, sometimes just weird greenscreen is part of Scorseses plan to deceive? I really have to see the film again soon to try and find a pattern to this, but it sort of fits. As I was thinking about it, I also started thinking how the regular audience members, most likely blind to these technical aspects, would react. And it might actually be a very subtle, but effective trick. On a conscious level you can’t really tell that there’s anything wrong with the picture (and you most likely don’t consciously question it), but your subconsciousness will have it’s doubts. And then it all starts adding up.
How did you feel when you watched the movie? Am I just rambling, or has Scorsese employed these technical and logical glitches on purpose? I’m anxious to hear if anyone else had a similar experience.
Exciting things are happening in digital film cameras, as both RED and Arri are rolling out new sensors rated at ISO 800, with relatively noise free boosting to at least ISO 1600 (looking forward to see this in full resolution in a cinema). This is more than one and a half stop higher than the currently most light sensitive film stock, and looks to be a golden number in terms of what you can actually “see”.
Yes we’re talking unlit night scenes again. The first RED ONEs are having their sensors swapped at this moment, and the first Mysterium X sensor will reach Norway and Eirik Tyrihjel this month. Footage from the Mysterium X has already surfaced on the internet, and shows:
1) Leonardo DiCaprio lighting a cigarette and himself with only a matchstick (direct download)
2) some people walking around the city at night, filming the scenery and each other (direct download)
3) people in industrial settings, cars in the dark and longboarders (direct download)
Seems like ISO 1600 with a wide open prime lens is what’s needed to look around a city at night, like countless HDSLR shooters have been showing us for the past year or so. What might be more interesting with such a high sensitivity is how you can control depth of field, especially when shooting miniatures or high-speed work (not to mention miniatures in high-speed). Also, it’s interesting to note that HDSLR started this race, and now the big players are following their lead, as digital acquisition builds a real competitive advantage over film.
Arri are still building their next generation cameras with the new ALEV III sensor, but they should also be released during 2010. Footage has yet to show up online, but needless to say, it should be a potent match to the Mysterium X. It will find it’s way into three different cameras with different features and pricing. They are not planning to upgrade the D21, but an Arri representative at Gothenburg International Film Festival last week claimed that it would still be top notch in their lineup for at least 2-3 years. Then again, productions that can afford the D21, can usually afford a few lighting units as well.
Is this a sign of things to come? Digital for sensitivity, film for quality? Or will digital catch up in latitude and color depth as well? At the very least, the DoP will have more freedom to focus on the creative process as limitations get fewer.
We are wrapping up our first proper project together in Lillehammer, a 4 minute sequence from an imagined longer story entitled “The Settlement” (attempted translation anyway). Our team consisting of one student from each profession, shot a story about a woman confronting her husband’s lover, a female tattoo artist, with the fact that they have a common love interest.
The first semester at The Norwegian Film School is filled up with a lot of watching films, workshops and lectures, and small team building exercises, and it’s a wonderful feeling to finally be making movies again. Or rather, sequences, as the school keep reminding us we’re only making one movie in the time we’re here, which is the exam film.
Projects at the school are based on the principle that creativity is better stimulated with restrictions, and we’re not training to make short films, we’re training to become feature film cinematographers (and directors, producers, sound designers, production designers, editors, and script writers). The main restrictions on this exercise was: A minimum of two locations (one required to be a waiting room), a maximum of 12 setups, two predetermined actors, and a maximum of two rolls of S16mm film (approximately 20 minutes of recording time in 25 fps) framing for 1.85:1.
Esteemed Norwegian DoP Hallvard Bræin was mentor for the cinematographers on the project, and gave some very stimulating input during pre-production and on the set. Our generation of photographers being a very digital one, he made a point of making us light by eye and meter, and not use digital cameras as a mental crutch as most of us are used to. Judging fine nuances in contrast can be quite a challenge for the eye and the mind, especially when it comes to the darkest areas which can be quite critical on a negative. One of my personal goals with the exercise was to push the shadows quite a bit, to see how far I can go before losing details and adding too much grain. In the end I found a simplified version of Ansel Adams’ zone system to be very helpful, as I’ve used it a lot when shooting 35mm stills.
We shot Kodak Vision2 200T and scanned to ProRes 4444 at Stopp in Sweden. An important aspect of the school is what they call “the right to screw up”, which means none of what we do is screened anywhere, but I thought I’d post a few screengrabs as a “teaser”. (This might not look right on your monitor, depending on calibration).
In 1989 Madonna spent more than $5 000 000 to create a music video for the song Express Yourself (directed by David Fincher). Today, bands like OK Go record their videos on cheap DV cameras in their own backyard, while C-Mon & Kypski have their fans contribute in the latest internet phenomena called One Frame of Fame.
My classmate Jørgen first showed me this interesting music video/social media experiment/viral campaign for the (at least to me) unknown band C-Mon & Kypski which features 6843 (and counting) frames of fans mirroring dance moves on their webcams to create something that looks like a mix between stop motion animation and bullet time. Basically, you play back the video, and at certain points the video cuts from the performance of the band member to a short sequence of fanframes that continue the movement. You come back tomorrow, and the video will have evolved with even more fans joining in.
In the modern music video, concept is king.
Reminds me of an Icelandic short film I saw some years ago (Þröng sýn) where the filmmakers had been sitting in the streets asking random people to draw individual frames of the film on top of the images they had shot (16000 drawings by 1350 members of the public). Only with the added benefit of the internet.GPTVG49D4YA3
Apart from maybe the reintroduction of 3D and the final breakthrough of the RED digital cinema camera in the established business, one of the big technical news of 2009 seems to be HDSLR cinematography. What happened? Everyone was waiting for Sony, Panasonic, Arri and the other camera manufacturers to answer the challenge from RED, when suddenly Canon and Nikon flicked a switch on their prosumer and professional digital stills camera systems, and created a small revolution in itself that even made RED revise their next generation of digital cinema cameras.
It’s a young format, and clearly far away from widespread use in drama productions, but the technology and the ideas are exciting, and I think it’s a very interesting subject for discussion.
The Nikon D90 and the Canon 5D MKII became the first DSLR’s to offer video recording in respectively 720p and 1080p resolution when they came out in 2008. Overall reception was a bit sceptical, what would you actually want to use this for? Very few professional photographers could see themselves shooting video all of a sudden. But that was before a man named Vincent Laforet entered the field. He realised the huge potential of the format, and convinced Canon to lend him a pre-production model to shoot this sample video with:
Admittedly an early attempt, and maybe not a work of art in itself, but the video showcases some of the main strengths of the format, and why it has become so popular. There is an amazing amount of detail and colour saturation in these low light scenarios which were shot with almost no extra light at all, and the aesthetics of the image resemble 35mm film, mainly due to the shallow depth of field achieved by the camera’s full frame sensor (which is actually physically larger than the standard Super35mm gate).
The Good
The 5D can shoot up to ISO 6400 with surprisingly little noise. The highest ISO motion picture film available is 500 — in other words, doubling the amount of light more than three times to achieve proper exposure. With their next generation, Canon and Nikon are bumping the max up to 12800, with a ridiculous boosted ISO of 102400 (I think more than a few people thought this part of the spec sheet was a typo). Vincent Laforet tried the new camera in October, and was very excited (read here). These cameras see more than our eyes do, and somehow manage to make bright pictures of it. This is the video Vincent Laforet made with the pre-production 1D MKIV:
Some of the images were even shot on ISO 12800, because he left it on by accident and didn’t notice(!).
The Norwegian national broadcasting (NRK) also did a test (here and here) with cinematographer John Andreas Andersen where they compared the 5D with the RED, and pretty much confirmed that the 5D beat the crap out of RED in low light scenarios (but not much else).
Realistically you would probably not shoot ISO 12800 on a feature for quite some time, but the point is that when they bump the max to 12800, that means the 6400 is looking better than the last generation. Next time they might bump it to 25600, and all of a sudden 6400 is starting to look really attractive. Both Arri and RED are moving up to a “standard” ISO 800 in their next generation cameras, which I presume will still be virtually noisefree.
When it comes to the old film vs. digital battle, I think low light sensitivity will be one of the main differences between the formats a few years down the road (not saying I think film will go away, more on that later!).
The Bad
We’re talking about a camera system that in general costs less than 10% of it’s high end motion picture counterparts. It’s an unfair comparison.
First of all, the stated resolution of 1080p is not really a technical reality. The eminent Stu Maschwitz (“point it at a resolution chart, and you’re in for some hurt“) has highlighted this for some time (read here). The DSLR was not designed for shooting 25 frames or more per second. In order to achieve this, the camera does not use all the photosites available for each frame, and uses aliasing to recreate the details that are not actually there. This can be very bad.
Second of all the video is recorded in a highly compressed format (H.264) which leaves very little freedom when it comes to treating the images in a post production workflow, and highlight detail is somewhat suffering compared to the professional competition. There is much less data per frame, and no matter how intelligent the processing is, something is missing. Generally you want as much as possible (one of the main arguments for still shooting film).
There is also the issue of ergonomics and compatibility. A DSLR is designed for shooting stills, not for shooting movies, which means that practical implementation in an established workflow can be troublesome on a regular drama production. Whether it’s handheld, on a tripod, or from a crane or steadicam, you might need a lot of special equipment and adapters to be able to operate like you’re used to.
Also stills lenses are less than ideal in practical use, the focus markings are compressed and less accurate, and they breath a lot when focusing. You can get a PL-mount adapter for your DSLR (check out ASC-member Shane Hurlbut with a Primo on his 5D), but then you need special gear to utilise standard accessories like a mattebox and focus wheel. Definitely an option, though!
The final big issue in my opinion is the still unsolved rolling shutter issue caused by the CMOS sensor in these cameras (explained here). Also called the “jello-effect”, this is an attribute of how CMOS sensors capture an image that cause moving shots and handheld shots to wobble. This is actually a major gripe with the format so far, making a lot of footage shot on HDSLR’s feel a bit “liquified” (check out this video, and maybe this more humorous approach to the subject).
(There is also the issue of audio recording, but I’m assuming external recording of sound here.)
Conclusion
What makes the HDSLR bad in terms of ergonomics is also what makes it an excellent niche product. A small, inexpensive camera that deliver high quality images seems like a perfect companion for stunt work and small, cramped locations. When the cost of the camera is so low that you can break it on purpose, I see the possibility for lots of destroyed cameras with exciting footage inside. The unconventional size also allows for unconventional images.
Also, the promise of a low-cost alternative to 35mm film that deliver similar visual qualities is an exciting prospect for independent drama productions, and low-budget music videos and commercials. This is not just an assumption, even the venerable Lucasfilm have shown a clear interest in the technology, inviting cinematographer Philip Bloom to Skywalker Ranch to demonstrate (interesting read and video!). They had a closer look at the footage on a large screen, and was reportedly still impressed.
But the actual quality of the images is still nowhere close to 35mm or 16mm film, nor can it compete with the professional digital solutions available today. Except for low-light, which is an exciting evolution of digital image capture, and maybe what most obviously is starting to create an advantage to digital capture over film.
It feels like this is the direction where digital is going, and why you will decide to shoot your next drama on a digital format. But I still think film will stay king of ultimate image quality and information. And even when digital finally catches up, I think film will stay an attractive alternative because of the aesthetics that you might want for certain projects.
What do you think? How will HDSLR affect your reality? What are your pros and cons?
Additional Resources
If you’re interested in reading more about HDSLR and what it can do for your future film projects, I can recommend reading up on the following people and resources:
ProLost
Where the eminent Stu Maschwitz provides valuable insights and thoughts
Philip Bloom
Who travels the world and tests HDSLR on different projects for different purposes
Shane Hurlbut
Another cinematographer who explores the world of HDSLR
Creative COW
Now with a dedicated HDSLR-forum
Vincent Laforet
The grandfather of HDSLR video, but really a stills photographer
5DMk2 blog
News and links about HDSLR
Some equipment to help adapt the cameras for video capture:
Redrock DSLR Cinema Bundle reviewed by Creative Cow
Zacuto Z-Finder v2 reviewed by Philip Bloom
And finally a couple more examples of HDSLR video:
Kommer Chomets Illusjonisten på norske kinoer?
Flashback: Kom og se
John Woo tilbake på banen
Bind meg, elsk meg! (1990)
Slik blir soundtracket til Coppolas Somewhere
Toy Story 3 – fullførelsen av det fullendte
Varg Veum: Skriften som burde vært bedre
Pernilla August til Venezia med Svinalängorna
Ingen virkelighet i Armadillo uten død
Begrepet: Nuit américaine
Montages er ikke redaksjonelt ansvarlig for innholdet på bloggene, og meninger og ytringer som kommer frem i de respektive bloggene er ikke representativt for Montages sitt syn.
Ønsker du å forhøre deg om mulighetene for å opprette en blogg under Montages? Send en epost til post@montages.no
RSS Feed