Meet Lytro Cinema: 755-Megapixels, 300 FPS, Refocus and Change Depth of Field in Post

Following the introduction of its Immerge virtual reality system, Lytro, the "light field" camera company whose consumer models we now see discounted nearly everywhere, recently left the consumer space to concentrate on and introduce its new product, Lytro Cinema. Offering a complete solution with an included server to handle the 755 megapixels of data at up to 300 frames per second (not typos), the Lytro Cinema is a new kind of too-good-to-be-true beast. But the most incredible thing about the camera? It's no lie.

As the questionnaire for a trial of the system asks for production budget estimates of the largest film you're working on in increments of "Under $10M" all the way to "Over $50M," you can be sure the Lytro Cinema won't be cheap. It must take a supercomputer to run all that data regardless. And as photos show the form of the camera is more that of a larger studio camera as opposed to a RED-type system, it won't help directors win awards in fast-action sequences of flying cameras. But the Lytro Cinema offers something other systems cannot (apart from its ridiculous resolution and maximum frame rate).

Since every pixel has full directional information, depth can be measured at any part of the image. This allows editors to refocus, change the depth of field, and, perhaps most importantly, pull out and edit specific objects or entire sections of a scene in post without the need for a green screen or otherwise more complicated effects.

After some experimentation in the consumer market, perhaps Lytro has found its true calling.

Log in or register to post comments


Isaac Alvarez's picture

Awesome read! This is a beast, want one.

Ryan Cooper's picture

Rumours I've been hearing is the thing starts at $125,000 to rent for a small production. Start saving your pennies!

Jonathan Lobb's picture

This looks amazing, and I hope that Lytro is successful in selling it to the industry. Even moreso than this particular camera this hints at their future developments as they make them smaller and more capable. Imagine how crazy a RED-sized version of this would/will be 10 years from now?

Joonas Nieminen's picture

Making the sensor smaller would hinder the ability to re-position the camera since the sensor would cover a smaller area around the subject. But I consider the re-position ability less important compared to the other attributes this camera has and thus making a smaller-sized version would of course be the next step in product development. This way the camera could be placed on a regular HeavyDuty-tripod and not on a pickup truck.

Chris Kennedy's picture

It will be very interesting to see what this leads to in the years to come. It could definitely cut a lot of production jobs (ACs and grips specifically), cut the need for multiple cameras, and streamline vfx.

But you'd lose the collaboration which goes into these productions and lose the look of the film/video. Lenses, filters, and the type of cameras chosen do so much to create the look of the image-and computers can't quite replicate the organic feel of film, or the distortions a vintage anamorphic or Cooke lens gives... I'm not sure I like the direction this would lead to from a cinematography standpoint.

Chris Kennedy's picture

It's those imperfections which cinematographers use to help give their footage a distinct look/feel which filters can replicate but still haven't quite truly captured yet.

Resolution has nothing to do with it, 35mm even though it's scanned at 6K/4K still has more resolution then the 8k Red Weapon, Alexa 65, etc. which all still use bayer pattern sensors so that isn't what the actual resolution is. These digital cameras still don't retain highlights nor the smooth rolloff from the highlights that film has and just clips.

Sure digital is basically on par with film-and will very soon surpass it, this camera system will definitely change how we perceive the term to "Fix in Post"

Chris Kennedy's picture

I said nothing about those 'imperfections' as being desirable-but lend to give footage a look and feel. Don't get me wrong I'm all about digital cameras now and have no plans on returning to the good 'ole Arriflex 435. Since post houses these days aren't as equipped for 35mm and you can experience dropout frames-it's just not as reliable as it used to be since practically the industry is entirely digital.

I mean 'organic' as clearly not appearing like an electronic image, not as "appearing as natural as the real thing"-nothing to do with resolution. Guess it's just a term that us old film buffs understand.

The bayer sensors' number of pixels is different from the resolution after debayering the footage in post. i.e. 4k from a Red debayered is 3.2k, the same is true for the Alexa and everything else with a bayer sensor.

An original camera negative of 35mm has more resolution then 4k; obviously it's an analog format and you'd get sub 2k during projection so that's why it's always perceived as soft. I've gone through the telecine process dozens of times with 35mm-the resolution is there.

Dynamic range depending on the film stock would be around 13-15 stops, sure that's what digital cameras do these days and are about to surpass that. However digital cameras get most of those stops in the shadows and less in the highlights, it's opposite that for film. And it's that 'smooth roll off' in highlights which allows cinematographers like Robert Richardson to push his highlights over 4-6 stops from his key when he shoots film... Yeah can't do that digitally right now.

I've only done HDR with Reds, they create some motion blur artifacts in the highlights and sure work for some shots-but aren't ideal for everything.

You're trolling at this point

John Riedy's picture

Sure, but will it mount to a drone? Or a Steadicam, for that matter? :-)

Richard Johnson's picture

Hum...Just like their other cameras the images look weird to me. Not sure if any footage from that promo was actual footage, but it would be interesting to see.

Sagar Solanki's picture

I read an article where it explains how the demo video is actually filmed with a green screen, to demonstrate what this video camera should be able to do. No one has seen the end product yet. And 775MP ~ 40k resolution sounds cool but lets see if they can deliver. Also, how do they plan on handling 400GB/s transfer speeds?

I guess time will tell if this will be revolutionary or just another flop like the illum that was sold for 80% off.

Graham Marley's picture

I can't help but be skeptical. "We just crushed every digital motion imaging company by whole orders of magnitude" prrrrrooobably comes with a catch.

Sean Molin's picture

Looks like Lytro realized the "Tesla Model" of development was the way to go. Start with the expensive, luxury, low-volume technology and use that to fund future consumer tech.

I know they started with the consumer camera, but those were an absolute toy; a proof of concept. By really starting over from the top (Roadster), next they can fund professional photography cameras (Model S/X). Then with the profits from that, they can fund serious consumer cameras (Model 3).