Author Topic: When a Camera Manufacturer Hates Me…  (Read 1198 times)


  • Administrator
  • Full Member
  • *****
  • Posts: 157
    • View Profile
When a Camera Manufacturer Hates Me…
« on: April 26, 2015, 10:52:14 PM »

Being called out by Jim Jannard in his final post to Reduser got me thinking: do I really have a problem with RED?

By Art Adams | August 29, 2013

It's a dubious distinction that I was called out by Jim Jannard in his final post to Reduser. It's true that I've been critical of RED over the years; while they've made my life easier in some ways they've made it harder in others, and I've been rather outspoken in asking them to eliminate the "harder" part.

I remember being shown a RED ONE handle engraved with a serial number, a symbol of the owner's commitment to buying a RED ONE.  It struck me as odd that so many would bet money on a camera they'd never seen before, but the excitement was palpable: here, finally, was an affordable digital cinema camera that nearly anyone could afford.

The RED ONE showed up at an interesting time in history. As best I can recall there were only a couple of single sensor cameras in use, if even that many. While the Panavision Genesis broke new ground in Hollywood the rest of us were cramming Pro35 adapters onto three-chip cameras to emulate the "film look." It was quite common in my market (San Francisco Bay Area) to have video engineers on set painting cameras and almost nothing shot on video saw a color grade. When this new "RED" camera was announced it seemed too good to be true as it cost considerably less than many of the three-chip cameras available at the time while giving us the advantage of an actual "digital negative." The video engineer disappeared, and color grades returned to the video mainstream.

It took a while for the cameras to trickle out. RED missed a deadline or two but the cameras finally started to ship, although the company made it clear that they were still in the very early stages of development. In spite of this warning many of these cameras immediately went into professional service and I know of at least two once-in-a-lifetime shots that were lost due to early software crashing. RED did make it clear that their cameras were only being released so that owners could participate in the development process, but they'd marketed their product so successfully that new owners couldn't wait to put them to work.

I remember seeing a RED ONE for the first time and thinking, "What is this thing?" It didn't function like any other video camera I'd ever seen. The dynamic range changed with the ISO settings, which is exactly what happens with film. Nearly every look function was metadata, meaning it was almost impossible to screw something up that couldn't be fixed later. The menus were complex, but once the camera was set up it was seemingly dead simple to use. It really did seem that an affordable digital replacement for film had arrived.

I called one of my favorite rental houses and asked to be notified as soon as their first RED ONE came in. They did--and they also told me it went right out again. Apparently if a PL lens mount isn't lubricated in a specific way it will fuse itself to the lens if the two are locked together for any decent length of time. A technician opened the box, pulled the camera out, checked the lens mount, put the camera back in the box and sent it right back. As promising as the camera was, it was clear that the company itself was still inexperienced in the craft of making cameras.

The positive side of this is that I learned more about how cameras worked by watching the RED ONE being built before my eyes, and asking questions of video engineers that I'd never known to ask before. For example, early software builds made clipped highlights bright yellow in daylight and cyan in tungsten light. I thought this was a fatal sensor flaw but a subsequent software build resolved the issue. That was quite a surprise as I had no idea that this was a software fix and not a hardware fix. "Oh sure," said the smartest camera engineer in town, "all cameras have to deal with this. One color channel always clips first and the trick is to desaturate the highlights to avoid color shifts."

I did not know that.

The camera didn't look so great under tungsten light so I borrowed some DSC Labs test charts to find out why. Bright colors looked fairly nice in daylight but under tungsten light they became dark and muddy. One of the first things I learned about color in film school was that any bright hue can be desaturated by adding blue, and that's what was happening here: every bright color the camera could reproduce outside of pure red was being contaminated by blue. The closer I brought the white balance to daylight the less of an issue this was, so I made it a habit to use blue filters when shooting under tungsten light to try to reduce this muddiness. Fortunately it didn't take much blue to fix as 1/4 to 1/2 daylight correction was often enough to eliminate the blue mud.

(RED's software tools have vastly improved over the years. The great strength of raw file formats is that you can revisit the footage years later and make it look a lot better than you could at the time you shot it, although that's of limited use when delivering a project on a schedule.)

After that I looked at dynamic range and discovered that while the camera was usable at ISO 320, RED's stated based ISO, I preferred to rate the camera at ISO 160. I lost a stop of highlight latitude but the camera was much quieter, and as I don't like a lot of noise in the image unless I'm creating a definite effect, and as I shoot a lot of visual effects work and compositors hate noise, I opted to take that dynamic range hit in favor of a nice clean image. This meant that if I shot under tungsten light and used an 80D filter to partially correct the color of the light to daylight my effective ISO was 100. That's not very light sensitive, but three-chip cameras of that era weren't a lot faster and the reduced depth of field generally made the compromise worthwhile.

I really liked working with the RED ONE for quite a long time. It was a little frustrating due to a number of quirks, such as software bugs that caused crashes or I/O faults and fans that could never be completely turned off, but it made the Pro35 adapter instantly obsolete and I was quite happy about that. Still, it was a bit of a wild card to work with. For example, I never had a RED ONE overheat but I'd heard stories from people who had, and an engineer at a rental house pointed out that the camera's exhaust vents were along the bottom of the camera, while heat rises. It was clear that there were a number of design choices that were driven more by marketing than practicality, but that was acceptable because the RED ONE really was a viable film replacement that cost a lot less than a Sony F35 or F23.

Writing about these quirks didn't make me any friends at RED, but at the same time I felt it was important that people understand what they were getting into. Although the camera was foolproof in some ways it was also completely different to the point where many people had a hard time understanding the few anomalies that might screw them at a later date. For example, prior to the RED ONE color was taken for granted: no one expected a camera to look substantially different under tungsten light verses daylight. The color temperature of the light really made a difference to the RED ONE, and this was not something you could undo later short of an upgrade to RedCine. There were other issues, such as rolling shutter flicker (which I discovered accidentally and found to be particularly insidious, and which was not unique to the RED ONE other than that it was the only camera around for a while that had a rolling shutter) and a shiny red OLPF that made lens flares a lot more noticeable. Still, the RED ONE changed everything: as imperfect as it was, it was perfect enough at a great price point.

Then came the competition.

I firmly believe that the RED ONE gave us the current generation of cameras at dramatically lower price points. Without RED we wouldn't have the Arri Alexa, the Canon C100/C300/C500, the Sony F5/F55/F65, the SI 2K, etc. The difference between all of these companies and RED is that they market directly to their user base, while RED markets to their fan base. The two aren't always the same. Sony, Arri and Canon sell great images in a reliable package, while RED sells the dream of competing as a filmmaker at the same level as a movie studio. That's a much more powerful message, but it has caused me a lot of headaches.

The chief of these cranial annoyances is that the camera has become more important than the cameraperson. Prior to HD a cinematographer had a pretty reliable chance at a long career if they could consistently deliver beautiful images no matter what, especially as they were the only ones who knew what the image was going to look like when they shot it. Producers and directors had to hire a DP based on their track record and then trust them implicitly to deliver the look that was agreed upon in prep. This system seemed to work as most of the greatest films in history were shot this way.

Video removed some the mystique and control of the DP because now everyone could look at an image and critique it, regardless of whether they had the artistic experience and taste to do so. Rather than continue to trust the visually sophisticated eye of  the craftsperson hired to create images that perfectly captured the mood and emotion of the story, it became simpler for those who were less sophisticated to say, "Make it brighter;" "Make it warmer;" "Aim the camera that way." Media production is inherently a collaborative medium but somehow video paved the way for others to jump into the DP's realm and make creative choices in spite of the fact that they really didn't know what they were looking at.

Seeing an image on a monitor is not the same as understanding it. Anyone can watch TV but very few can make it. The ability to look at an image doesn't automatically make one an art critic, and those who are not art critics don't see or understand the subtleties that an art critic will. To remove someone with that kind of artistic eye--the DP--from the creative process does no one any favors, but unless you know enough to trust that a dedicated artist will make prettier images than you will it's easy to second guess their every decision and hobble them to the point where they're almost literally painting by numbers.

RED took that a step further by implying that owning a RED camera puts on one the same technical level as a movie studio. This is very appealing because it removes the guesswork from film production. You don't need talent, years of experience, an eye for cinematography and storytelling; all you need is this one amazing camera. This marketing strategy worked because the industry hates trusting artists. There's a lot of money at stake and creativity is, in the eyes of many, unpredictable and unreliable. Many would prefer to trust a predictable piece of equipment over the vagaries of an artist's vision.

Cameras are easy to find and fairly reliable; artists appropriate for a given project are a lot harder to find. Wouldn't it be nice not to worry about the artist anymore? This is definitely not RED's position, but by marketing the camera so well they effectively empowered producers and directors to select the camera first and the cameraperson later. Non-cinematographers aren't qualified to make this decision on their own as they are working off a company's marketing materials and don't have the kind of experience that a career cinematographer brings to the table, but as the cinematographer's craft is truly understood by very few it's somewhat comforting to change artistic decisions into material ones and take them out of the hands of the "unpredictable artist."

If it really was all about the camera then we should have seen at least one really marketable feature from every RED ONE owner. Sadly this has not happened. An amazing camera isn't enough. You still need the right people behind a project.

Still, this thinking has dramatically changed the way I work. Often the camera is now booked before I am. Somehow, in spite of the fact that I've worked in the industry for 26 years in both film and digital and I've spent all that time doing nothing but camerawork, a director or producer knows better than I which camera is best for a project. Most of the time the camera chosen is a RED regardless of whether it's technically or fiscally the best choice. I've lost track of the number of times I've shot projects with REDs that could have been executed with another camera for much less money with no change in quality. I'd love to save production money, but once they make their choice the odds are very slim that I can change it.

Years ago I got a call from a company that wanted to shoot 120fps slow motion in front of a green screen. At the time they'd already selected the one affordable camera that could pull this off--the RED ONE--but to do this they'd have to drop down to 2K resolution.

They were finishing in HD, and a good rule of thumb is that Bayer pattern sensors lose 20% of their resolution right off the top due to the deBayering algorithm blending colors from adjacent photo sites into distinct pixels. From experience I knew that 2K looked soft in HD, even though 2K is technically higher resolution than HD if you only look at the numbers.

I was concerned. I knew that keying a soft image was going to cause them problems, so I told them that they'd have better luck shooting with another camera at 60fps (the maximum framerate available in an affordable camera at the time) and then doubling the frame rate in post after compositing was completed. I sent them a multi-page email explaining how this could be done along with sample frames from 2K images I'd shot that showed how soft they were at HD resolution.

They thanked me profusely… and hired someone else who said, "Sure, no problem, I'll shoot that on a RED." I don't know how that project came out, but I hope post wasn't too painful. It fascinated me that I lost a job because my expert technical opinion didn't match their expectations... and my professionalism lost.

I told another DP about this and his response was, "I wouldn't have tried to talk them out of using a RED. Agreeing to do it their way gets me closer to a paycheck." I didn't feel right doing that… but it occurred to me that this may be the new world I have to live in if I want to make money as a cinematographer: let the clients make the decisions they think they can make better than I can and then try not to get blamed if they don't work out. Keep mouth shut, cash check.

It's a tough call, but if trying to do the right thing means not getting the job…

Every camera has strengths and weaknesses that make it great for some jobs and not so great for others. I know what all of those variables are, but somehow my expert testimony is no longer required. It's baffling. I liken it to hiring a world famous landscape artist to paint a picture of your estate. You hire them because they can paint a lot better than you can, but then you dictate which paints they are to use, what kind of canvas they'll paint on, which brushes are allowed, etc. Rather than letting the artist do what they do best, which only benefits you in the end, you hire them and then second guess a number of the creative decisions that allow them to do the quality of work you hired them to do. Instead of trusting the artist, whose skill you appreciate but don't really trust because it's so intangible, you instead rely on the advice of those who make the tools the artists uses because those are things that you can see and touch… and control.

Why would you trust a company's marketing over that of a true artist and craftsperson who does nothing but paint amazing pictures every day? I can't quite figure that out. It has something to do with trusting those who make the tools more than trusting the artist, because you can touch the tools but you can't touch artistry.

A subset of this phenomena happened to me in my early days shooting video. Some producers and directors would second guess my decisions by running them by the video engineers before allowing me to implement them. Somehow the person turning knobs was more grounded than I was because they were dealing with physical imaging tools while I was creating images simply by using my brain and placing lights in the right spots. This caused a lot of problems as video engineers were typically not the most artistic of people, as picture that was artistically perfect by my standards might be technically underexposed according to their instruments. The instruments usually won that argument because they were tangible, whereas my artistic taste was not.

Another thing RED did was try to expand its user base in ways that put cinematographers in a bit of a bind. An example of this is when RED tried to push its way into the still photography world by selling the idea that moving images and perfectly sharp still images could be captured at the same time. It's a nice thought but the technology isn't really there yet. Still photographers can carry the sun in a briefcase because their lights only need to be bright enough for 1/125th to 1/250th of a second to freeze fast-moving action. Moving pictures require continuous lighting, which requires quite a lot more power and equipment to achieve the same shutter speed, but RED never explained that. I had a number of friends call me frantically during fashion shoot preps: "How do I do this? I'm supposed to capture moving images with a RED Epic that can be repurposed as stills but my lighting budget isn't large enough to give me the shutter speed I need to eliminate motion blur. I don't see how I can do this well but my client seems to think this can be done and I can't convince them otherwise without losing the job. HELP!"

Not a good scene.

That's the difference between Arri, Sony and Canon: they market to us, the people who have to use the gear. RED markets beyond us, or around us, which puts us in the position of having to explain to non-technical creatives that their emotional response to a marketing message does not guarantee that they've chosen the right tool for the job. The problem is that creative people hate to hear that kind of thing. They were sold a dream, and the person who tries to wake them up usually doesn't get the job.

That's been my primary complaint about RED: they've taken decisions out of my hands with their marketing. That doesn't feel good. I can forgive the fact that their cameras have noisy fans and occasional glitches and the wrong connectors and not enough of them, because the price point for what they deliver is fantastic. There's a real need for tools like this. Just don't go to my boss and convince them to use your tools without talking to me, because I'm the one stuck having to make the tools work--and if the tools are wrong for the job then that doesn't help my career.

Another frustration is that RED's marketing falls down in one area where I really want them to succeed. Recently I've seen several Red Dragon demo videos that make the new sensor look very, very promising, but I don't really know what I'm looking at as the data behind the images is missing. One video in particular was very well shot but the technical information that accompanied it was incomplete, and therefore useless. Another video demo was shot handheld, in available light and with still lenses, and was meant to showcase how the new sensor handles flesh tones, but the images were horribly graded and did no one any favors.

Director of Photography and owner of the Cinematography Mailing List Geoff Boyle, FBKS offered to shoot an objective test with the new sensor to see what it could really do. According to a message he posted to the Cinematography Mailing List, RED's response was to say no, they weren't going to send him a camera because they couldn't be sure he'd say only good things about it.

Rather than send a camera with a new sensor to a cinematographer who knows how to do extensive objective testing and risk him discovering that Red Dragon, like every other sensor out there, isn't perfect in every way, they'd rather send it to people who don't know how to do any of that but will say nothing but nice things about the new sensor. That's great marketing, but not very helpful on a practical level to those of us who will actually have to work with the new sensor, possibly within days of its release, without knowing what it will really do.

All of which boils down to this: we won't really know how good the Red Dragon sensor is until we can touch it ourselves. RED is showing lots of pretty pictures from the sensor, but any camera can make pretty pictures under the right circumstances. I'm not getting enough information in these videos about the circumstances under which the footage was shot, which prevents me from being as impressed as I might be if I knew what was really going on. I'll just have to wait and see, and hope that the only reason that they won't send a Red Dragon-updated camera to Geoff Boyle is because he makes them nervous from a marketing perspective, and not because there's something about this sensor they don't want him to discover.

I have a lot of respect for RED. They changed the world in a lot of good ways, the most important of which is that they've kicked their competition in the ass to lower costs and make higher quality cameras. That's huge. Unfortunately RED has also made it acceptable to ship cameras that aren't complete yet and empower our clients to take important equipment decisions out of the hands of the very people they hired to make those decisions.

This is definitely a mixed bag.

I'm hoping that the new sensor turns out to be as amazing as RED says it will be, partially because I want more choices as a cinematographer but also because many of my clients will believe RED's PR that Red Dragon is the most amazing imaging device ever. Things will be a lot easier for me if that's true.

I'm also hoping that the new face of RED will be a little more professional. I've been critical of other camera companies without being called out on their websites, and in several cases they've actually found my criticisms helpful. When the C300 first came out I found a glitch in the white balance setup and wrote an article detailing how to work around the issue. Rather than complain publicly about me Canon fixed this bug within a month. Sony did the same thing: when their first F5 and F55 cameras came out they had a hiccup in the highlight sharpness settings that put black lines around extreme highlights. Myself and several others brought this to their attention and they released a software update less than ten days later. I have yet to be called out on Sony's website for this.

I don't complain about cameras because I have some sort of hidden agenda. I complain because I want more out of them. The better and more reliable they are the faster I can work, and that's particularly important these days as budgets shrink and schedules become more and more optimistic. Every camera has something wrong with it--there is no perfect camera yet--and ultimately I think it's more productive to fix those problems than to be mad at the people who point them out and show others how to work around them.

I don't complain about a company's marketing unless it consistently puts me in awkward situations that distract me from  the work at hand. I've been trained to work very, very efficiently and squeeze the most out of my client's dollar. If a client chooses a tool without my input and demands that I use it, and it's the wrong tool for that particular job, they're going to waste a lot of money forcing me to work around it. I don't think that's fair to anyone. I reserve the right to complain, loudly.

I'm not a successful camera manufacturer. Clearly RED has a very different agenda to mine. They want to sell as many cameras as possible; I want to make really pretty pictures on time and under budget. When their marketing messages interfere with my ability to meet my agenda I reserve the right to offer them public feedback as to what I like about their products and marketing and what I don't. Anything they do to address my concerns will result in my becoming a fan of their products. I won't be a fanboy, whose admiration for a product is driven primarily by emotion, but rather a solid professional fan who works in a tough industry and always delivers on time.

Note: Geoff Boyle's response to Jannard's post can be read in the comments here. I would have responded directly on Reduser but comments have been closed.

Disclosure: I have worked as a consultant, both paid and unpaid, to a list of camera companies that includes Sony, Canon and Arri. My early experiments with the RED ONE and DSC Labs Chroma Du Monde charts led to my working for DSC Labs as a paid consultant.

Art Adams

Director of Photography

Twitter: @artadams