The Camera Convergence
No camera is perfect but in 2015 a camera was released that might be as close as we have ever come – so why didn’t we realise it?
I have thought a lot about what would make the ideal camera for me. We all have this dilemma when choosing how to spend our hard-earned money on new gear. We pour over spec sheets and click from youtube video to youtube video. If we are lucky enough to get our hands on an actual camera before buying, we shoot test footage, pixel peep it for hours and agonise over the colour science and noise.
Filmmaking is an expensive game to play and the kit you have will affect how you shoot. Your style, your techniques, your creativity will all be affected by the camera in your hands. Are you a run-and-gun youtube shooter travelling to cool locations with a GoPro in your pocket? Or a corporate doc-maker locked on sticks with good audio monitoring? Maybe you are aspiring to be the next great film director capturing a kinetic Steadicam look.
In short – the kit matters.
If I were to sum up my camera needs they would be to have a flexible camera, capable of shooting for broadcast, sometimes with a crew but often solo, and with the ability to shoot a wide variety of shots. That’s vague but I’ll tie it down below.
When I am looking at cameras, I evaluate 3 things – image, usability & movement.
1 | Image
In 2012 I attended a one-day workshop with Philip Bloom in New York. It was an interesting day and the topics addressed one central question: What makes the perfect camera?
A list was drawn up on a whiteboard of every desirable feature in a camera and it went something like this:
“HD, no wait – 4K! Actually why stop there, surely it should be 8K for proper 4K?
Slow Motion – how slow, 120 fps? High bit rate obviously and 4:4:4 colour sampling.
Speaking of the software what codec should we use? Well, how about raw? That makes sense.
So where is the image data actually coming from, a S35 sensor? Why not full frame, actually why not Vista Vision?”
The conversation went on and the ridiculousness of this uber-camera was apparent. Looking back on that list from 2012 it is amazing just how many of those features are available in the cameras of 2019 – but not all, and certainly not all at affordable prices.
Red, Arri and the likes lead the way in image quality with high resolution, raw and compressed images. Phantom Flex cameras shoot into the 1000s of frames but loads of professional and pro-sumer cameras happily shoot up to 120fps, 150fps, 200fps…
Sensor technology improves year on year with lower noise, better high ISO performance. High powered processors run software that can realise HDR, HyperLog Gamma and all while giving users intuitive face and subject-tracking autofocus.
Recording media has improved (see Moore’s Law in action) to capture the TBs of data that would have been unthinkable 10 years ago.
So why have we not reached the goal of one camera with all the dream features of even 2012?
Technology is certainly part of it. Developing new sensors takes time and investment. Delivering large volumes of data from the sensor can cause overheating, as can recording for long periods of time. The huge volumes of recorded data can be prohibitive.
There are a great many technological hurdles still to vault but what the camera companies have achieved thus far is genuinely incredible. The Sony A7s series with its amazing low-light performance, the Panasonic GH5 series of feature-rich M4/3s cameras and the BlackMagic industry-disrupting models with incredible headline specs. At the higher end, Red push the boundaries of cinematic images with massive sensors and awesome software.
Without wishing to stray into conspiracy theory territory, is there a business incentive for the manufacturers to hold back new technology? Certainly. If they released a perfect camera they would never sell another; just ask the TV manufacturers who all made lovely flat-screen TVs, drove the price down and are now desperate for another innovation to sell another round of products.
Competition in the market is driving innovation and starting price wars where the consumers are the winners. There are now dozens of excellent cameras that will shoot high quality, cinematic images for affordable prices. Even the phone in your pocket has been used by Hollywood directors to make feature films. What a time!
2 | Usability
The promise of the DSLR revolution was ‘cinematic images in small form-factor’ – great! In practise tho it just isn’t that simple. Anyone who has built a ‘franken-rig’ to add zoom recorders, external monitors and handles, various ND filters, or recorded 2-system sound to sync in post because their dslr had no headphone jack, will know what I mean. The camera has to be usable.
I would also include in this category the perception problem. Showing up to a well-paid corporate shoot and suspecting the client is disappointed when you pull a ‘little camera’ from your bag is a real issue. It shouldn’t be, but it is. The results should speak for themselves and winning business is not about how big a camera you can offer, but let’s not kid ourselves, it matters.
Companies have sprung up worldwide offering every add-on imaginable to convert your ‘cheap’ camera into a cinematic powerhouse – but all at a cost, making that camera not quite so cheap anymore. External recorders, top handles, shoulder rigs, audio recorders… you name it, you can buy it. Wouldn’t it be better to just buy a fully functional video camera instead?
Cameras like the FS7, C300 and BlackMagic URSA sit in this sweet spot. Large sensor look, high quality image, audio and ND filters built-in, handles etc. and as such they have become incredibly popular. The FS7 and C300 are considered industry standards. They don’t have everything – no touch screens, auto focus is not great, no follow-focus system or modularity – but they are widely popular for good reason. They do everything professionally.
3 | Movement
So often it seems that the movement of the camera is an afterthought rather than an integral part of the system. I find this to be one of the most challenging aspects of modern cinematography.
Hand-held, sliders, tripods, dollies, gimbals, car mounts, drones, action mounts, Steadicams, jibs, cranes… the list of equipment for moving the camera goes on and on. This is where the idea of cinematic movement starts to bump into the notion of cinematic cost.
The bigger and more ‘professional’ the equipment for moving the camera, the more expensive it is. At IBC in 2018 I enjoyed playing with a Technocrane and a camera motion robot, both incredible pieces of tech that I will likely never have a chance to use professionally… but never say never. Steadicams are prohibitively costly and difficult to use. Some of the crazy gimbal set-ups available featuring body harnesses and even Segways are turning camera operators into military vehicles.
Quite simply – camera movement is essential to creating cinematic, engaging images. That doesn’t mean that a static shot on sticks or handheld isn’t good, but let’s look at them in the bigger picture of camera movement.
So how are these 3 concepts related?
I think of it as a triangle with points at Image, Usability, and Movement, and if you pull on any one of the points it pulls the others. Let me explain.
A very small camera, eg. a GoPro, is fantastic for movement. We can move it just about any way imaginable, carry it for hours up a mountain, hang it off a drone, dip it underwater, stick it to the side of a car, mount it on a helmet etc., but the image and usability suffer. The image comes from a tiny sensor with lenses that are fixed and it has a tiny fixed screen and no audio monitoring, no NDs, no XLRs etc.
A camera that prioritises image quality like the Red is heavier, more difficult to move and often sacrifices usability. Red cameras come without any decent means of recording audio, no built-in NDs and their modularity, which is a benefit for cinema production, is expensive and complex. Moving a Red is also more difficult due to the weight, expense and operation logistics. If you have ever piloted a drone carrying a Red camera you will know the fear. To crash a DJI Phantom is one thing, to crash a Matrice with a Red Gemini on board is quite another.
Even in the middle ground, the sweet-spot, the FS7, C300 style cameras are challenging to move. On solo shoots I am often required to bring a camera, sound kit, some lights and my tripod. To bring movement in the form of a slider or jib to my shots simply isn’t practical. Up steps the DSLR-style cameras but again the image suffers as the size and price reduce and the usability is a problem.
So is this inherent? Is there no way to solve this conundrum?
Let’s return to 2015. DJI had released the Inspire series of drones; high quality, impressive machines that now offered a new camera option – the X5. A micro 4/3s lens mount for affordable, interchangeable lenses. A 4/3’s sensor with good dynamic range. High quality internal recording CODECs. It was designed to fly on the Inspire drone but suddenly there was another option – the OSMO handle.
A simple handle with batteries inside that allowed the X5 gimbal to operate without the drone. The previous OSMO models had small cameras, giving a picture more like a GoPro, and they had their uses. I frequently took one in my kit bag for a walking shot or to mount to a car window. With the addition of the X5, the OSMO had grown up.
A phone or iPad was used as a monitor or, if you preffered, a DJI branded monitor – both being larger than the usual screen on most cameras I currently use. For a few hundred $’s more you could add a handy follow-focus system that was small, easy to use and responsive. Adding a follow-focus to an FS7 requires rails and 3 hands but the OSMO had it figured out.
The OSMO could be mounted to a pole for huge jib shots with wireless monitoring at the base point. Remotely controlled for dynamic moving shots when attached to the outside of a car or inside window. Sliders and dolly shots were done by hand, using a system that didn’t weigh a ton like its sister, the 3-axis gimbals we all know so well. Having used a GH5 on a Ronin-M and having tried A6400s on Ronin-S all the way to C300s on Movi15’s I can say from experience that I can’t use one for long, and even for short shots it is taxing on the my shoulders. But the OSMO is light and simple.
Best of all is that I can now use one single camera for not only all my ground-based shots but my aerials too. It’s the same camera 200ft up in the sky as pulling a smooth slider shot indoors. It makes colour matching a thing of the past.
So was it perfect? Well, no. It had a rather noisy fan that could interrupt audio recording and speaking of which… while it did have an audio input, I wouldn’t suggest you record much audio through it. It didn’t look like a ‘normal camera’ so clients may have found it very confusing. I’m not aware of any underwater housing options for it and it didn’t offer the same time-lapse features as DJI’s current OSMO pocket for example.
The OSMO X5 and its successor the X5r offer high quality image (even including RAW recording), many usable features and unparalleled movement options for a very reasonable price.
When the X7 on the Inspire 2 drone was released in late 2017 I thought the moment had arrived. Time to sell all my bulky kit, all my sliders, jibs, big tripods, dolly track… but it never happened.
The X7 camera, the successor to the X5 range looked fantastic. A proprietary lens mount meant lightweight, custom primes. The sensor was S35 and leaps ahead of the X5 in its technology. Good dynamic range, even low light performance. Could this be the camera of my dreams? The advertising certainly pointed that way with numerous shots showcasing the inspire sans props, and the X7 shooting everything.
I waited for the release of the OSMO handle and associated accessories, but it never came. Rumours swirled that a deal had been struck whereby DJI would refrain from releasing traditional-style cameras in order to stay on the right side of supplier Panasonic - I don’t know if this is true.
DJI’s Ronin-S came along soon after, a single-hand 3-axis gimbal. It’s nice, but not the system I hoped for. It’s heavy, lacked a non-mechanical follow-focus system, took set up time and was often being used with external monitors or recorders that added more weight. I couldn’t use it for more than a few minutes at a time. However, it is popular.
The DJI DL Lens mount system would benefit from a zoom lens option and a built-in electronic ND filter, much like Sony’s, would seem like the perfect addition to any drone camera, controllable from the ground. Time-lapse functions like those available in DJI’s newest gizmos would be nice… but wait a minute. Am I back in that 2012 workshop wishing for a perfect camera that will never exist?
DJI struggle to be taken seriously as a camera company; they are Chinese drone makers, surely they aren’t welcome in the camera club of Canon, Nikon, Sony and the likes? I think that’s why DJI bought Hasselblad, the most ‘camera’ company of them all.
I’ll end this with a plea: DJI, please complete the circle. If you release an OSMO handle for the X7 (or even better, the X8 as has been rumoured) then I will ditch my conventional camera kit and switch. I think that this technology has the power to truly change the filmmaking landscape in the same way that the DSLR revolution did.
DJI did it once and I’m sorry we didn’t notice. Maybe next time we will take them more seriously.