Sony to show 4k motion picture camera prototype at NAB

image credit: Sony

According FDtimes Sony will show a new 4k motion picture camera prototype at NAB (April 9-14).

"It doesn’t have a name yet, but instead of calling it the Next CineAlta Camera, let’s give it the working title of Cine Altissima. She is not a mere 4K motion picture camera. Behind the PL mount, there’s an 8K sensor: 8768 x 2324 pixel single CMOS sensor (that’s 20.4 Megapixels) — Super35 3-perf size,  16-bit RAW output, 16:8:8."

Some additional specs: 1-72fps and 1-120fps in "High Frame Rate". The camera will have a dockable SR memory card recorder and at 24fps, 1TB memory card will record one hour of video. The camera doesn't have a name yet.

In December last year Sony announced their 4k video roadmap.

This entry was posted in Sony and tagged . Bookmark the permalink. Trackbacks are closed, but you can post a comment.
  • Craig Houdeshell

    This announcement will get RED’s attention. Maybe.

    • Is no one else weirded out even just a little by that image?

  • Scott L.

    An 8k sensor is impressive. Thats a lot of overscan for a 4k video target.

    I think Red and Arri only use 10 & 12 bit raw, with Red using lossy Jpeg2k compression on their Redcode raw file, so not really being “raw” at all ( raw is uncompressed). Even dSLRs are only using 14-bit raw. 16 bit raw would mean huge files, and possibly a 16 bit a/d converter like the Hasselblad/PhaseOne medium format cameras (that also use 16 bit raw files)

    • Mark

      Why does raw have to be uncompressed? There’s no pixels in raw data (yet), it doesn’t need them. You get the pixels, and file size increase, when you convert the files with your raw software.

      Are you saying my 15MB raw image files that end up being around 70MB TIFFs are compressed?

      • Contax Man

        There is pixels is RAW data, RAW data is just the image captured by the sensor before interpolation.

        You can compress a bayered sensor data, but compression for RAW kinda defeats the purpose, especially if its compressed into a JPEG-based wavlet format like Red does with their REDCODE. Basically you’re getting a compressed bayered file, debayer it, and compress it again after you’re done. You’re going through multiple lossy steps, and the whole point of RAW is to be able to modify your image without destroying sensor data.

        BTW Arri’s raw format is a true uncompressed 12-bit RAW file, Red’s isn’t.

      • Sky

        There are 2 kinds of compression Mark – the lossless, like ZIP or one used in TIF files: LZW – and lossy compression like JPG or GIF.

        If red has lossy compression in RAWs than it seriously sux.

        • Stephen

          Clearly you’ve never worked with any RED footage. It is lossy compression, but it’s visually imperceptible.

          You’d never actually want 4K uncompressed raw. at 24fps you’d be looking at 324 MB/s. That’s megaBYTES not megabits. In other words, a 2 terabyte hard drive would store just over 6 SECONDS of footage!
          You would need a massive RAID array the size of a mini-fridge to get enough hard drives in there to stripe a bunch together to get enough write speed to handle the datarate, and then go absolutely nuts with the number of those stripes so you could store even a few minutes of footage.

          • chris

            nice math you learned there 😛
            324 MB/s multiplied by 6 seconds is 1944 MB
            so one Minute is 19.440 MB
            and one Hour of footage is 1139 Gigabyte which would fill up a little more than half a 2TB hard drive which are common now.

            still a lot of data 😉

          • Stephen

            Yeah, massive brain fart. Just over a terabyte per minute is still utterly insane.

          • George

            @ Stephen

            You mean 1 terabyte per hour- not minute. You really aren’t good at math….

            And if you’re spending $30-50k for a camera, you can spend the extra chump change for ~1TB/hr of video. Red using JPEG2k compression and calling it “raw” is just a a joke. I getting sick of all of Red’s outright marketing lies:

            The Scarlet was announced before dSLRs did video (before the D90), ancient history, and they still haven’t released it. Their “raw” is a joke, they call their original Red One 4K only to be resolving 3.2K of real resolution (hence justifying people upgrading to “5K” Mysterium). Their HDRx is just a cheap trick that cellphone can do now (two stream, different exposures).

            Then they go ahead and rip on every other camera maker, calling themselves ‘revolutionary” their Epic a “Nucelar reactor”. I used to love Red, but there are too many broken promises and lies from Jannard at this point…

          • @ George

            What is the point of shooting uncompressed raw when you can see no difference with compressed raw? It is still raw even if it is compressed since it is recording sensor data, not debayered RGB pixels. 1TB of video per hour is not reasonable by any means especially when you consider that you need highspeed drives to deal with that kind of data rate.

            In fact, ProRes422, which has become a defacto standard in editng and delivery is also technically a lossy compression codec, but it, like RedCode is visually lossless. Isn’t that all that matters?

            While it is true that the 4k red only resolves 3.2k of information, this is true of every single chip camera. A 21 megapixel 5DmII does not actually resolve 21 megapixels of information, because it is a bayer pattern. I never understood the RED naysayers, a lot of great work has been produced on a RED, isn’t that enough?

          • George


            Compression being “imperceptible” is a joke. The reason why people use 4:4:4 and uncompressed is because each footage is going to ultimately go through a ton of grading and manipulation. You don’t want to go through too many lossy steps as down stream you’ll going to have to compress even more. That is where it becomes perceptible.

            Also in regards to RED only resolving only 3K for their 4K camera, unlike the Canon 5D Mark II, a cinema camera is going to be delivered in a specific resolution format.

            A 4K camera needs to resolve 4K because that is what the end user is going to see. Not 3K, but 4K. And this should be expected for gear that labels themselves ‘professional’ and charges what it does.

            Also, this is the reason why Sony is using an 8K sensor for their 4K camera.

  • amien

    red sits in the red.

  • Bigus Dickus

    I smell… like there is burning lots of plastic RED cameras

    • Stephen

      …Except for the fact that they are made out of aluminum alloy.

  • this sounds impressive…. and very expensive!

  • Back to top