Joseph S Wisniewski: "The multiple perspectives captured mean you can generate 3D images or video from every shot at any desired parallax disparity"
Except that single point cameras are essentially useless for 3D video, because the occlusion is only correct from the single viewpoint. Moving the viewpoint means that you now need background that was blocked from the single viewpoint.
The end result is that the images look like bad automatic 3D conversions, until a skilled artist retouches all occluded areas.
see the caption in the photo of the camera: "The sensor is housed in the black box behind the orange strut, which appears to be at least a foot wide." we're not talking about podunk super35 sensors. it's a 304mm wide sensor. super35 is 24mm wide.. there's a reason the thing is the size of a car.
light field data allows correct perspective shift within a limited area because it captures both angle and intensity of light passing through an area. essentially it has the background and foreground captured from a continuum of perspectives within an area. the tradeoff is the somewhat ridiculous amount of redundant data and power needed to process it all
This is obviously not going to replace most standard film production crews shooting ordinary scenes of people, no reason to use a car sized camera to get a kiss scene in the bedroom right. I think people are missing the point that this will be able to possibly revolutionize shooting the ultra costly massive special effect scenes like explosion/building collapse/million dollar per second kind of shot where you've got one take to get it right and you don't want to miss something as mundane as focus for the love of God.
PeterOlsen: Looks very interesting. Exciting to see what it will bring in the future.
To me the insane data rates of the quoted 300 Gb/s look like a huge problem with present technology? I take it that the 300 Gb/s is before compression? Even with compression it seems an awful lot of data for shooting a whole movie. Even with the fastest commercial standards it would take a loooooong time to transfer from one medium to another and I dare not think about the time it would take to upload it to the servers. It would probably be faster to physically deliver the data :) I am not in the film/movie business so I don't know if there are any proprietary, non-commercial standards which are substantionally faster than USB3/Thunderbolt??
I hope somebody can enlighten me :)
Have a nice day.
They rely on the moving truck full of hard drive method of data transfer
Woo lytro you did it! Congrats!
You don't just shoot for the moon or the stars more like quasars and cosmic microwave background
LFPCPH: Phone cameras remind me of the happy days of the Philips cassette tape. Manufactures tried for decades to make a good cassette recorder. They never made it because it was impossible from the start.
i think cell cameras like the lg g4 reach closer to high quality than cassette tapes could reach hifi. for a scene with limited dynamic range the resolution and color are remarkably close to what you'd get out of a milc or dslr with high quality lens
god i love photography forum bitterness i'm clearly just as pathetic
FantasticMrFox: Indian trains? Nooo way, never been photographed before. So original.
i know rite they should only award photos of never before seen things like the stuff you take of your bulls
ElviraGrey: lol this isn't surprising in the least. The difficulty of the 'skill' of photography is highly over stated.
hey it takes $20k of equipment and a 10000+ post history with a demonstrated ability to demolish your opponents to prove you're a pro
the quality of camera phones blows my mind for what they have become. that you cannot appreciate what they are is just sadness
lol FF amateurs
the online photography forum world will erupt in flames as suddenly dynamic range increases completely destroy the balance of company fanboy argumentation and Fuji/Panny fanboys laugh themselves with glee to their deaths
Jim Evidon: If I read the article correctly, by employing a global shutter effect to the entire sensor, they could totally eliminate a mechanical shutter; either focal plane or leaf, and simply use the firmware to electronically turn the sensor on or off at user adjustable “speeds” resulting in a lighter, mechanically simpler and more accurate camera.
To the user, the controls could look familiar with dials and wheels, but the internals would be completely electronic. I expect the new product could hit the market within 2 years or much less. I suspect that the project is further along than is reported.
Times they are a changing.
don't worry there'll be a 'RealShutter-Sound (R)" $250 addon that perfectly simulates the click with adjustable volume and additional timbre controls
mxx: What happened to the a6100 and a6200?
i for one am a fan of mirrorless turds
Chris Joy: Once again Sony takes two steps forward and one step back. The new EVF, improved AF, awesome video specs, Slog3 and so on are awesome - but then there's no headphone jack to monitor the great audio courtesy of the new (or old if you're a Nex user) mic input and all those wonderful AF points dancing across the frame but no way to quickly change subjects or focus point. This is especially annoying when shooting video and my favorite feature of a number of other cameras, just touch the screen to move focus. The new PADF should be really accurate and make racking focus a breeze.
haven't you heard manual controls are for hipsters
xeriwthe: a6000 low ISO JPEG is impressive, certainly exceeds the x-pro2 for detail. the noise reduction of the x-pro2 at high ISO maintains a higher uniformity of detail than the a6000 at high ISO, but just barely. i am impressed by the sony jpeg engine too bad i'm still a fuji fanboy at heart. the xpro2 low iso jpeg hits a similar overall detail level to the x-a1 RAW when scaled to equivalent sizes, which is key to me and impressive for a JPEG engine.
interesting how the electronic shutter makes a pretty significant difference in the pixel level sharpness. what causes that?
resummarizing my garbage; a6000 low ISO JPEG is impressive, certainly exceeds the x-pro2 for detail. but the xpro-2 renders smoother details with less propensity to render crunchy artifacts like the bayer sensor. the noise reduction of the x-pro2 at high ISO maintains a higher uniformity of detail than the a6000 at high iso.
overall i am impressed by the sony jpeg engine too bad i'm still a fuji fanboy at heart. the xpro2 low iso jpeg hits a similar overall detail level to the x-a1 RAW when scaled to equivalent sizes, which is key to me and impressive for a JPEG engine. overall detail similar to a 16MP bayer with the benefit of the contrasty, smooth moire-resistant xtrans rendition, all in a JPEG sooc. damn not bad
a6000 low ISO JPEG is impressive, certainly exceeds the x-pro2 for detail. the noise reduction of the x-pro2 at high ISO maintains a higher uniformity of detail than the a6000 at high ISO, but just barely. i am impressed by the sony jpeg engine too bad i'm still a fuji fanboy at heart. the xpro2 low iso jpeg hits a similar overall detail level to the x-a1 RAW when scaled to equivalent sizes, which is key to me and impressive for a JPEG engine.
may be over stating it here but i think fuji actually may have 'fixed' the watercolor in the sooc jpeg. omg
watson076: Over the last few years manufacturers have really lost touch with what photographers love, but I believe Fuji really hit the reset button. 40-50 mp this and that, 4k video, blah blah blah - who the f**k cares ? What I want is a true photographic experience (and I speak for many friends of mine who feel the same way). Fuji concentrates on what true photographers want, not what focus groups tell them to manufacture. Thank god someone is listening.
nothing like focusing on the 'experience' to excuse the slower pace of R&D.
never going to give up my fuji fanboy dedication after they granted us the x10-series designs (goddamn i hope they continue the series) but it does hurt a little being stuck a few years behind in technology. 4k video would be nice for some things
Joed700: I shot Fuji XT-1 along with several prime lenses, 56mm f/1.2 being one of them, for about a year. It's a well-built lens and a good performer. However, Fuji X series bodies and lenses don't hold their values. If I search my memory correctly, it dropped at least a few hundred dollars since it was first introduced...anyhow, I sold the entire Fuji system because I could get the same/similar results with either Canikon 85mm f/1.8 for a third of the cost and shooting FF. The fact that Fuji systems don't hold their values is a deal breaker for me. IMO, paying too much for APS-C...
if you buy it for cheap as a second hand user are you paying too much?