Camera technology can go in interesting directions

Discussion in 'Casual Photo Conversations' started by Mark Keefer, Oct 3, 2018.

  1. I came across what looks like a fancy point and shoot, maybe a bit of an upgrade for $1950 camera that has 16 x 13 megapixel sensors and lens modules, looks like it takes several photos at once and stitches together a 52 megapixel image, claiming a 28 mm-150 mm full frame zoom equivalent and 13 stops of dynamic range. Looking at all the sensors is what caught my eye.

    It sports a 5 inch HD Touchscreen, supposed to have a battery that can last 8 hours. It is supposed to get firmware updates wirelessly.

    It says image captures in LRI (RAW) and JPG formats? I just thought it was an interesting development. All I know is what is in this link.

    Last edited: Oct 3, 2018
  2. Using multiple lenses and sensors to capture different parts of a scene has its advantage, because the exposures of the bright and dark areas can be independently controlled. The result would be a definitive improvement in dynamic range. This is an innovation in the right direction.
  3. I thought that this was going to be a terrific product. I was disappointed. It was not in vain, though, as it will spur further innovation that will actually achieve something.
  4. I don't think we will see Canon, Nikon, Sony making pro bodies with multiple lenses soon. But this might be a cool camera for the phone camera point and shoot crowd that have iPhone disposable income. You are right though, multiple sensors could expose for different parts of the scene.

    Perhaps the direction we may see one day on our pro cameras will be more powerful processing sensors where every pixels could independently compensate for bright and dark adjust ISO and white balance to some degree, giving us better dynamic range and color from sun and shade. Possibly this could be a feature adjustable by the photographer as to the degree of compensation. It will take cameras and sensors with a lot more processing. You wouldn't need multiple lenses, just independent autonomous pixels. A smarter sensor so to speak. And the way the information could be stored in a RAW file that may allow for the degree of this compensation to be later adjusted in post.

    Also imagine every pixel being recorded simultaniously, no more jello effect or bending of the swinging golf club. No more banding from LED lights flickering at 50 or 60 Hz because your shutter speed on your mirrorless is faster. No more high-speed sync issues on manual flash. Each pixel records the data at the same instant and keeps it in a buffer until it can be written to a file.

    Perhaps one day.
  5. They were called CCD's.
  6. Leaf shutters didn't have this problem either . . . It was restricted to focal plane shutters . . .
  7. I was referring to CMOS (complementary metal-oxide semiconductor) with a perpixel buffering. As far as the CCDs (charge-coupled devices) Perhaps with battery technology improving we might see CCDs again one day. Wasn't it the power consumption that made camera manufacturers move to CMOS?
  8. Sony hints that they have a CMOS sensor in development which has a global electronic shutter.
  9. Sandy Vongries

    Sandy Vongries Administrator Staff Member

    Back to the camera / concept? Please!
  10. Back to the original concept - a multi-lens, multi-sensor camera that stitches up a 52 MP image. Practically speaking, every image would be an extreme wide angle shot. Is that really useful? In my own experience, fewer than 15% of my images would be considered "very wide," e.g., 24 mm or wider. We have seen these "innovations" in use, for virtual reality (having given up on actual reality). Even audio is not immune. There are microphone arrays in a stylized human head for VR sound too. The results are less than spectacular, far less than one can obtain with conventional spaced arrays.

    Perhaps someone can impress me with results from a multi-facited camera, a VR praying mantis so to speak ;)
  11. Looking at a really good view of the lenses from the link provided by NetR's post, THE VERGE article, it looks like there is an array of different mm lenses, giving them 28mm through 150mm, perhaps some duplicated. Not all are wide angle and not all 16 lenses were used to make the panoramic landscape. From the first article it looks like the camera is also allowing full depth of field adjusting in post if I am reading that right. So it may be using combinations of F-stops and lens and possibly portrait lenses and wide angle lenses depending on what you are trying to accomplish. It sounds complicated, there would be so many possible combination if the software is advanced enough. Portrait lens getting part of the shot spot on, wide angles getting the background and depth of field moving from lens to lens depending on what the photographers post edit wants to create. Depending on how intuitive the post editing software is would determine how complicated this will be to work with. SInce it gets wireless updates, they may still be tweaking the software.

    If they can survive and keep pushing ahead with this, it is going to get interesting. With the use of multiple lenses, I don't see why 3D VR isn't one of the directions it will advance into and manipulation of stills perspective to some degree.

    Hopefully we will see advances with this camera system and prices lowering and they get enough momentum to survive. It looks like they have already overcome some of the big challenges in getting this thing developed and pushed to market, they are in partnership with Softbank Vision Fund and Leica Camera and plan to do more than just consumer cameras.

    Here is a link to the Light Company story. LINK

    I think it would be fun playing around with this new camera tech. I know this is the kind of thing my 29 year old son would enjoy. He loves taking photos with his iPhone, no interest in dad's big dslrs and lenses. If this can fit in a big pocket and easily carried around, and produce better than iPhone images he would love it.
    Last edited: Oct 6, 2018
  12. As well as I know it, the sensors used on simpler cameras and phones can simultaneously and electronically capture pixels at the same instant.
    It requires additional electronics at each pixel position, such that the actual pixel sensor is smaller, so isn't done for high-end cameras.

    It seems to me that you could make different pixels in the array with different sensitivity for better dynamic range, if that was wanted.
  13. From what few examples I have seen, the RED Hydrogen got it right. It does a really good job of depth mapping with only *two* camera modules. Two are certainly better than one, and apparently better than 16.

    Video (13:23):
  14. Computational photography like the L16 does is the future- like it or lump it. While that camera won't match a FF or APS-C dSLR at it's max resolution is does get scary close under ideal circumstances.

    Dynamic range increases are worth the R&D alone.

    Still, all the computational gimmickry in the world doesn't get you past the limits of RBG capture. Note the dark skin model wearing the bright red scarf on the main page of the camera site. The reds are blown out just like my stupid 60D, 1990's consumer print film and other dSLRs with the typical red channel problems. We need to get beyond crude RBG capture to fix this and move to 4/5/6 color sensors.
  15. I think the simulated depth of field bokeh is poor. Hair fly aways get blured, objects seen through bike tire spokes remain sharp are just a couple examples. There may be the occasional image where it can be pulled off but it is a gimick. It is right up there with the claim the iPhone will be the DSLR killer. It is a selling point that non camera people may fall for. What a disappointment bordering on scam. The shoddy broken would anger me if I paid over $1000 for a smartphone in hope this camera app would come close to DSLR quality. Just shoot the image flat and at least get a true image for the type of camera it is. Just my opinion.
    Last edited: Oct 20, 2018
  16. Even though I was a software engineer, I was a reluctant convert to digital cameras (though not being a very good darkroom man and not having room for a darkroom, scanning and printing was a relief). At some point, shooting film began to require a recognition that your image quality wasn't going to be as good, but you just enjoyed doing it more. For me I mostly shoot digital but still some film. So computational photography makes me think of that transition.

    What if your phone, say the iPhone 15 or something, maybe with multiple lenses, better sensors, and the benefit of computational correction of the image, gets to the point that the image quality is actually better than most (at least lower end) digitals? Now those digitals are likely to start using some of those same tricks so the phone might continue to be limited by it's sensor size somewhat. But I just find the iPhone 6S Plus that I have takes good pictures, but using the phone as a camera is awkward, mainly because of the control layout and the lack of any real control in the basic apps. I guess you could have a case that gives you the dials you expect and a decent shutter release but I haven't seen one that works yet. I guess it would bother me though, if somehow you could get BETTER pictures on an iPhone than I was getting on my digital body.

    One of the Fuji cameras takes a series of pictures over a second or less which effectively does focus stacking. You can have everything sharp or select your depth of field. Now this is a gimmick NOW but it's a pretty cool gimmick. No matter what we do, the world will change around us, including our hobby (in my case). I guess the only thing we can do is to try to keep an open mind.
  17. jakenan

    jakenan Guest

    there used to be a digital capture thing called a lytro which just collected light and
    you put it into your computer and did the rest. the entry level models didn't cost much
    ( less than 100USD ) the higher end models cost between 500 and 1000clams. it came out
    a few years ago ( 5? ) around the same time as i remember seeing things about the L16...
    at least the L16 looks like a p/s camera, the lytro looked like a weird rectangular tube.
    its good to see the L16 is still around, the lytro isn't made anymore and the company
    is now heavily involved with VR ...
  18. I got a second hand, but I believe unused, original model Lytro for about $20.

    It is sort of fun, but I haven't actually done much with the pictures.

    Among other things, it can generate stereo pairs, even with only one lens.

    What it actually does, is image a 4x4 block at each pixel position, covering
    the light coming from 16 different directions. The first step in image processing
    is a four dimensional Fourier transform.
  19. jakenan

    jakenan Guest

    id love to hear how it works when you start making images with it,
    i always wanted one but never had the $$ and energy to get one :)
    have fun!

Share This Page