Why are TV Cameras still huge and expensive? (2019) [video]
youtube.com
youtube.com
Most of my work has been as a contractor or freelancer and I’ve always believed in buying excellent gear. Yet I’ve never had any producer or editor ask “what did you shoot this with?”*
Durability and reliability is the prime reason pro gear is preferred but for me anyway that calculation has changed in recent years.
I just returned from Borneo where I was mostly using Sony RX-10 prosumer cameras. I carried 4 RX-10 bodies and never missed a shot.
Everything was shot in 4K. A top end pro camera from Sony would have an edge on quality in terms of pixels and dynamic range. But not enough to spend an additional 25k.
Plus with the cheaper cameras I much less hesitant to put them in harms way. When you are less concerned about destroying the camera it provides opportunities that otherwise might not be considered.
I actually just hired a production team for a project I’m involved with as a producer. They are going to be using top notch Arri Alexa cameras with prime Cine lenses. This is GREAT gear!
But I’ll never forget the $25 Panasonic M42 28mm lens I used on a RED camera that was fantastic. It had a bit of chromatic aberration but this was easily corrected in post. Nobody questioned the results.
Our iPhone or Android cameras rival what Panavision was providing in the 80’s.
* the exception is always super slow motion or incredible low light performance.
I went back and forth on this balance for years. I wanted that large sensor look and feel, with great full-frame lenses, but trying to use it live was a nightmare workflow. This was before Canon started making the C series that mixed the world of a camera with an EF mount that also had SDI I/O. I tested those, but still couldn't trust them in a live broadcast, mainly because of the lack of a parfocal lens I can focus remotely. I think with cameras like Blackmagic's Ursa and some more "prosumer" lenses that are parfocal we may be getting closer to having the best of both worlds, but I haven't researched that in a couple of years.
Broadcast optimized gear is like having Kubernetes. Everything built for broadcast talks SDI/NDI so it's generally possible to operate or at least monitor the device remotely. Then you have routers and mixers designed to modify those video feeds in real-time with as little latency as possible. One of my first gigs in broadcast involved learning how to remotely white balance cameras from the control room or patch the output of one video feed from a satellite truck into a monitor on set, while also recording that same feed in a completely separate room. I've also spent time doing "documentary style" and film production and was amazed at how different everything was, and how some workflows that are super straightforward in a newsroom are basically impossible with things like GoPros and DSLR setups.
The same compensation behaviour can be done in reverse to prevent focus breathing (and yes, that 500$ lens does that, too).
Sadly outside of Sony's 18-105G, 18-110G and 28-135G, most lenses for mirrorless or cinema cameras are designed to be manual and don't even attempt it.
Weird that there are no 3rd party lenses that exploit "software"/"motor" to escape the parfocal optics restrictions while properly exploiting the short backfocal distance of something like Sony E-Mount (vs. e.g. EF-S). With the ultrasonic, very fast AF motors, they can't tell me that an encoder ring for focus pull haptics in lieu of camera-side controls designed for the task would suffer too much input lag. And at the better quality end, the freedom from decoupling zoom and focus should make up for the cost of putting a motor inside, instead of just using an external remote focus pulling clamp. (That said, just adding a focus ring rotation sensor to offset the AF might even be possible to hack onto an existing lens, depending on the AF motor control scheme. A tiny dongle on the side near the base where one doesn't grip seems unintrusive to me.)
If fast (aperture) zoom lenses for sport photography had ultrasonic ring motors for auto focus 10 years ago, at low-to-medium quad-digits for the entire lens, and minimal (AF-included) shutter lag being a major selling point (next to the lack of audible frequency motor noise), there can't be a true technical limitation preventing lenses for "cinema" use from relying on a pair of motors to allow a mechanical/optical coupling between zoom and focus.
I did not mean to use an autofocus on the sensor side; rather just on the actuator side.
What is not talked about so much is that as a professional in a field disruption feels like your tools getting worse each year and your work day filling up with annoying work arounds. Overall there is progress but each step feels like a step backward in to the prosumer land.
Can you imagine if Kevin Smith had a reasonably modern phone, a semidecent PC, and a copy of Resolve when he started out? Of course, that's happening, except it's not Kevin Smith, it's some other snotty kid whose name will become associated with cult cinema in a few years. They're still in their backyard making daft wee movies with their friends and sticking them on Tiktok, and one day you're going to be buying a ticket to their film.
> I just returned from Borneo where I was mostly using Sony RX-10 prosumer cameras. I carried 4 RX-10 bodies and never missed a shot.
Even at that there were (in the early 2000s!) real proper feature films shot on standard-def DV, including one of my favourites, 24 Hour Party People. Every time GAS starts to bite I look at my Sony DSR-PD150 and think "you know what, if it was good enough for the guy that did Open Water or Super Size Me, I ought to be able to get by with it" :-D
“What are ’tickets’, Grandpa?”
Anyway when he's a teenager everything will be shot in Willam-Gibson-esque neural stimulation implants by actors with Zeiss Ikon eyes shooting 30720x17280 240p, and he and his mates will be running around the woods with an old banger Arri Alexa 35 "because you just don't get the same organic quality in 32K that you do with these old 4K cameras, it's just got so much more life", same as we did with 16mm and VHS.
Going to see a film on a huge screen with a hundred other folk will be just this incredible oldschool vintage thing, people dressing in authentic 20s fashions for the occasion.
And everything old will be new again, just like it was last year and the year before.
So basically Joel Haver: https://youtu.be/czlXuKfPI6U?t=277
My larger point was really about the distinguishing characteristics between “pro” and ”prosumer” are diminishing rapidly.
When a minor correction in edit is all that stands between $50,000 in expense the incentives are easy to figure.
But there’s a huge difference between a YouTuber with 7 GoPros strapped to their car/bike/wingsuit, and the on board cameras used for real time broadcasting for F1/MotoGP/NASCAR. I’d bet there’s about three orders of magnitude more money involved in getting the cockpit shot from Lewis Hamiltons car into a live race broadcast compared to even “big budget” YouTube “fix it in post” production houses like Hoonigan.
I imagine you also shoot under reasonable low light ‐ for high ISO shots, a 1 inch sensor doesn't capture as much light as full frame or even APS-c.
However, a7 IV has pretty good performance too (except at extreme ISOs), and is much more useful as a photography camera.
A7S III is also not getting video centric firmware updates, unlike the FX3 which has same internals.
But that only makes sense for narrative media, where you can repeat the same scene again and again. For live, or sports, the camera and lens need to be extremely overengineered so every take is perfect, because you won't get a second attempt.
what does this mean?
If you are broadcasting a football game, and the camera fails during the goal, the football players aren't going to re-enact the play for you. If you need to switch the lens for a better shot, you can't.
With sports you can’t do that you have a limited number of positions and a dynamic scene so you never know what is going to happen and cameras have to be able to track the action and operate in a much wider range of framings.
Sport lenses aren’t just over engineered for reliability it’s essentially like having an entire selection of lenses in one and being able to seamlessly transition between them.
Also the action tends to be very fast so the cameras and lenses have to be extremely fast and always give you perfect motion clarity and focus.
There’s no second chances but most importantly there’s also no post so you can’t crop out a frame for a better zoom or sharpen something that was a tad out of focus.
Similar to saying why is it hard to live in Antarctica, because it's extremely cold, but the details are what make this a meaty video.
Which means that you can zoom with it without having to refocus, which is not the case with less expensive lenses.
It's not perfect — there's always a lag of about 6 frames during fast zooms — but it's "good enough" for most situations (which means it's not ready for broadcast).
One thing to keep in mind though is in big production sets they are not using automatic focus at all, and cinematic lenses don’t tend to have focusing motors. (Also a focusing motor for a 100lb lens would likely be very noisy).
Older box lenses use a convoluted set of gears to achieve the same thing.
Some handheld ENG lenses for shoulder carried cameras get close to parfocal by making the lens elements travel in a complex way, but its not as perfect as the box lenses get.
The zoom position and focus position is worked out by moving the front and back elements of the lens, with a bit of adjustment so they track together properly. I actually have a nice old Sony V1E where the back positioner has failed (or become disconnected, possibly) so it never really looks right, and it would be too expensive to repair. It works great as a DV playback deck, which seems like a bit of a waste I guess, but repairing it would be about the same price as buying a working one.
It's far hard to aim a photon at the exact correct pixel on a sensor when both the lens and sensor are smaller. With larger glass and sensors, even if something is off slightly, there's a better chance it hits the right pixel.
This is likely more noticeable with chromatic aberration different light colors bend slightly differently. With larger equipment and long lens lengths, there's less separation between colors.
Just compare ENG broadcast sensor size to APS-C/Super35 or Full Frame: https://i.k8r.eu/SC-yYw.png
The BMD Ursa Broadcast used in the above video is an exception, but its also a cinema camera converted to take B4 mount ENG lenses and not built from the ground up to do live broadcast video. Its also from a company that has weird ideas about how broadcast equipment should work.
There's only a handful cameras with three individual sensors with global shutter on the market nowadays, and the price point is expectedly astronomical (~70k USD)
There's a couple exceptions, like the Super-35 HDC-F5500 and HDC-4800.
But you're right, they still have a few good options for now :)
But folks are using much smaller equipment for some production work too.
Has the quantum efficiency of silicon improved?
How much?
It’s possible to create a camera sensor that isn’t limited by the quantum efficiency of silicon. In fact, it’s likely that no sensor is “perfect” and that tradeoffs are made between cost and suitability for purpose.
My 2010 BMW gets better gas mileage at higher speeds than my 1973 Oldsmobile, even though the energy density of a gallon of gasoline hasn’t changed. And neither of them is anywhere near the theoretical maximum.
Well, the QE pretty much is the limit. Sorry.
Sorry. Again, no. Not a single well designed camera will do this. Quantum efficiency is, to use a term from a time gone by, the "emulsion". Now, if you are talking about a $15 security camera or some garbage design, sure, whatever, anything is possible. I have been involved in the design of cameras for terrestrial, orbital and lunar applications. You don't throw anything away unless you absolutely have to.
QE is the "fingerprint" of the sensor, it si based on the material used to make it and it all follows from there. Put a different way, it's how a sensor converts photons landing on a pixel into electrons. You can't degrade or enhance this externally. If less photons land on the pixel (bad optics, ND filters, etc.) the QE is the same. If more photons hit the pixel (back illumination, better optics, etc.) the QE is the same. If you have shitty digital processing, the QE is the same. Etc.
Going back to statement that "Digital sensors these days are way better at capturing light", still not sure what it means. For example, a better way to capture light is back illumination, which has been around for many decades. Interested in learning if there's something in modern sensors (in production) I might not know about.
I have a database of nearly every imaging sensor in production up to about a year go. It covers the SWIR to visible range. Every sensor you can buy. If there are any advancements that cause a sensor to be "way better at capturing light", I would sure love to know about it.
Note: I studied CCD and CMOS sensor design some 15 years ago under the scientists who designed a great deal of the sensors that flew in spacecraft. Jim Janesick was one of my instructors.
https://www.amazon.com/Books-James-R-Janesick/s?rh=n%3A28315...
No. It's about materials properties and physics.
1. The lens elements are really large to gather a lot of light so that it can work well in a wide range of lighting conditions. The front most element is as large as the front hood.
2. Every set of lens elements, some being just one lens, is moved individually, there is a controller that calculates the proper position for each lens and moves them based on input from the operators controls and the camera. This allows the lens to zoom really fast without loosing focus.
Edit: In my experience, the mass of the lens is not as important as the tripod head when it comes to keeping operator shake to a minimum. A good head will smooth out the operators movements and a bad operator is perfectly capable of making 100lbs of camera and lens shake.
I think I was just making sure it was clear (to others, since you probably are aware) that it's more than just lighting/apertures. Being good at a variety of lighting conditions, and being able to do that from 18mm-1700mm with a variable speed zoom without losing focus is where things get really big. There's a lot of features in the camera body to support this specific type of production beyond the lens, too. For live TV, an operator is making split-second decisions around composition (including adjusting for graphics on the screen), lighting, focus, and movement, while also monitoring on-air status and listening to both director and (sometimes) program audio. And then you have to add on the massive rigs that make sure everything's in perfect balance so you can whip the camera around quickly and have it stop on a dime. It's a whole system. A good high-end rig is like driving a Porsche. Sure the engine is a big part of the power and engineering, but if you don't also have a properly tuned suspension and grippy tires, you aren't going to have much fun driving it.
Are there any other industries on the peripheral that would also benefit from market innovation in optics?
A few years ago, a friend of mine, who is a video engineer, introduced me to the DJI Osmo Pocket[0]. That was an awesome little camera. I suspect that it may have improved with age.