Details of the Computational Camera “Lytro” are finally emerging, and it’s combining an Apple sense of minimalistic style with some high-end photo processing features, all into one tiny sleek package. Engadget was lucky with the first hands-on:
Right off the bat, you’ll notice its diminutive size. If we’re frank, we were expecting something more akin to a DSLR — visions no doubt conjured by the immense arrays used to do this sort of wizardry back in the day. Its small, beautifully made, anodized aluminum body reminds us of a cross between an iPod Nano and an iSight camera — inspiration that isn’t only skin deep, as Lytro’s desktop software (which you’ll need to open and export .lfp files) is currently Mac only, with a Windows version in development.
I must admit, I’m also surprised by the form factor but it truly is beautiful. Given how much can be done after the moment with this camera, it kinda makes sense that it would have almost no on-device controls. Also, they’ve created a neat flash/javascript based viewer that you can play with on their website, letting you export the raw LFP files to sites like twitter and facebook so that viewers can adjust the focus on their own.
Hopefully they’re working with groups like Adobe to create Photoshop importers for their format, so that eventually we can do some even more amazing things like photos with multiple focal depths (put the foreground and background in focus at the same time) or pseudo-HDR.
Working with giant screens like IMAX typically requires special cameras configured for panoramic shooting, with a lengthy post stitching step. However, once you move to 3D the typical setups break down because you have to get much more information in the frame:
Generating 3D video panoramas, however, is much more complicated. Whereas the stitching of each panorama for the left and the right eye requires a parallax-free recording of sub-images, the stereo representation itself needs a sufficient amount of parallax for proper 3D reproduction.
This problem is a tough one to solve, but Fraunhofer has come with a (Rather extreme) solution. Combining cameras and mirrors into a omni-directional camera module, they can put multiples of them together in 24-degree increments to get all the way up to a 360-degree circle for a massive 15,000×2,000 pixel image.
Currently in a rough prototype, it looks like a truly mammoth camera to try to work with tho.
The brains at Stanford University that brought us the SynthCam app and the Frankencamera are about to spin off into a separate company and offer their first commercial option: The Lytro. Boasting the same plenoptic lens shown at other events, it allows you to snap a single photo, a single instant in time, and then post process it later to change parameters like exposure, focus, and more.
The main lens is fixed in place; there is no auto-focus, auto-aperture, or other gubbins. This limits the number of moving parts which need to be adjusted every time a photo is taken, and which cause a lag between pressing the shutter-release button and capturing the image. Lytro’s snaps, by contrast, will be truly instantaneous, just like old film-based snapshot cameras. The light-field approach means they will always be in focus (since the plane of focus can be moved at will after the photograph has been taken). And the main lens is preset so that it always captures the greatest amount of light possible. This means that exposure time can be extremely short even in poorly lit conditions.
Unfortunately, this first offering will be useful only for hobbyists thanks to its 525×525 resolution. It’ll be perfect for home use (I can only imagine how many photos I could have saved if I could have adjusted the focus after the fact, and I’m actually somewhat knowledgable about cameras), and is actually targeted at internet sharing sites. They’ve already raised $50m in venture capital, so hopefully they’ll start ramping up production soon.
This week at the City of Science and Industry event, Nikon presented some of their concept camera technology including the unusual item above called (according to Google Translate) the “Multi-Ball”.
Multi-ball is an object “multi-views that can capture the atmosphere of a scene.”One wonders what kind of pictures can come out well!
That doesn’t really explain much, but I have to wonder if this may be a possible commercially-available Plenoptic Lens. Adobe’s been working on software for plenoptic lens for a while now, and it only makes sense that the would partner with a premiere camera maker to implement it.
At NAB, WASOL was demonstrating a new lens system called the ’3D Lensys’ which allows any traditional lens to be converted to a 3D lens. It’s a hefty lens at 21kg, but it’s a great simple way to add 3D to your existing equipment. Although, if it wasn’t for the fact that people saw it at NAB, the copy on their site reads like a phishing attempt.
As it is possible you to use all the functions such as auto/manual mode, zooming and focusing built in the camera. You can produce any kind of motion pictures and TV programs such as TV shows, sports events, live perfomance, etc, which require several cameras.
You don’t have to spend much time and energy on synchronizing two cameras mounted on a rig in order to converge right and left images, which is one of the biggest problems caused by two-camera-based 3D capturing system. This innovation technique have made us to have many patents in many countries and to be awardrd many technical.
Sony has announced that they’ll be jumping into the 3D Professional Camcorder market this NAB with their upcoming “3D XCDAM” camera.
At 2011 NAB Show, Sony will unveil the production model of the camcorder, which Honeycutt called “the camera of the future.” The 3-D XCDAM shoulder-mount HD camcorder captures images with two 1/2in 3CMOS (six sensors acquiring at 1920 x 1080 resolution) Exmor imagers and records up to six hours on two separate SxS solid-state cards, for left and right eyes; it can also be used for 2-D HD recording. Up to six hours can be captured on four 64GB hot-swappable SxS cards.
Just to wrap up this little piece, the previously reporting Canon T2i 4K firmware that was to be released this weekend was, in fact, a hoax. Resulting in nothing more than a RickRoll (I really thought that meme would die by now), things got more and more ridiculous closer to the release (I was busy all weekend, didn’t pay any attention til today) making it somewhat obvious by the time it was finally “released”.
Well as was suspected just a hoax but with the weakest of punchlines, a rickroll. Not a lot of originality there! A shame that he couldn’t have done something as cool, fun and original as the above video, that would have been a worthy punchline. Just someone after their 15 minutes of fame, well he has had them so time to move on. These things are done on the 1st April for a reason! I did my Canon Raw blog on April 1st last year, got a lot of hate mail but mine was nonsensical from the start even for the non technically minded, with pictures of a Video 8 player as a high end capture device, introduction of interlaced for coveted video look, the loss of the stills function and a lot more utter silliness littering the post. Only speed readers would have been taken in by it and it was not dragged out over 3 days! But as I said then I will repeat now, if we cannot laugh at ourselves who can we laugh at?
Perhaps the best part of the hoax is this parody video which shows why we may not really want 4K on the T2i.
I originally decided not to post this, but if a 4k Canon firmware is on the horizon then maybe it’s a bit more important. There’s a reason why companies like ARRI keep making incredibly expensive cameras like the Alexa, and this video by Nick Paton shows one of the important differences.
This is a short test comparing a Canon 7d with an ARRI Alexa. The purpose of the test being to test the ability of each camera’s images to be manipulated after exposure. I have graded each output to try to get it back to neutral using Gamma and density’s chart. As you can see the Alexa responds very well where as the 7d footage fails in highlight exposure detail and it’s codec compromises it’s ability to be pulled up in the darker regions. It’s interesting and telling and goes part way to explaining the more than substantial price difference.
Granted, you are probably asking yourself why you’ld ever want to shoot at a 7 stop, but if you did then here’s kinda what you’ld wind up dealing with.
LaptopMag has posted a video of the plenoptic lens technology that Adobe demonstrated during the GTC2010 keynote, showing a single snapshop of a girl in a field and how they are then able to dynamically change the focus in the after the photo.
First, the lens’ optics atomize the picture in thousands of tiny versions of the scene, all different from each other, at different angles and positions. When you add these lenses to your DSLR camera, the sensor would capture a huge grid of images made of all these little images. The effect is similar to the eyes of a fly.
Then, the software can take this grid and combine the versions into a single one. Using a simple slider, you can change the focus of the image in any way you want, picking any plane you desire to be in perfect, crystal-clear focus.
Gizmodo mistakenly says Adobe “invented” the technology, and simply calls the whole system “Plenoptic”. In reality, Plentoptic is a type of lens also known as a polydioptric which is actually a multitude of lenses working together (refer to the Wikipedia article) and the technology that Adobe was demonstrating was first developed and demonstrated by Stanford back in 2009. Adobe’s contribution was in moving it to GPU-compute, so that the internal processing comes down from hours to fractions of a second.
Hasselblad currently manufactures the 60-megapixel HD4-60 camera. As if that is not large enough, they recently announced that they they would be making a camera capable of capturing even larger images for 2011. CNET News is reporting that this camera will be capable of 200 million pixels.
Who would ever need a 200-megapixel camera? Well, lots of people I would expect. For one thing, astronomers could certainly use images that size. However, the way that they get to 200 megapixels probably rules out astronomy. Instead it looks like it will be used for advertisements used in print medium.
However, even in the rarefied air of the medium-format market, where image sensors are very large and very expensive, the 200-megapixel “extended multishot” technology won’t appeal to everyone: each shot will take about 30 seconds to capture, restricting the camera to stationary subjects such as cars, watches, and jewelry.
That’s because of the design of the Hasselblad multishot-series cameras such as the HD4-50MS. The current model uses tiny piezoelectric motors to shift the sensor a very small amount to combine two shots into one higher-quality image.
Details of the Computational Camera “Lytro” are finally emerging, and it’s combining an Apple sense of minimalistic style with some high-end photo processing features, all into one tiny sleek package. Engadget was lucky with the first hands-on:
Working with giant screens like IMAX typically requires special cameras configured for panoramic shooting, with a lengthy post stitching step. However, once you move to 3D the typical setups break down because you have to get much more information in the frame:
The brains at Stanford University that brought us the SynthCam app and the Frankencamera are about to spin off into a separate company and offer their first commercial option: The Lytro. Boasting the same plenoptic lens shown at other events, it allows you to snap a single photo, a single instant in time, and then post process it later to change parameters like exposure, focus, and more.
This week at the City of Science and Industry event, Nikon presented some of their concept camera technology including the unusual item above called (according to Google Translate) the “Multi-Ball”.
At NAB, WASOL was demonstrating a new lens system called the ’3D Lensys’ which allows any traditional lens to be converted to a 3D lens. It’s a hefty lens at 21kg, but it’s a great simple way to add 3D to your existing equipment. Although, if it wasn’t for the fact that people saw it at NAB, the copy on their site reads like a phishing attempt.
Sony has announced that they’ll be jumping into the 3D Professional Camcorder market this NAB with their upcoming “3D XCDAM” camera.
Just to wrap up this little piece, the previously reporting Canon T2i 4K firmware that was to be released this weekend was, in fact, a hoax. Resulting in nothing more than a RickRoll (I really thought that meme would die by now), things got more and more ridiculous closer to the release (I was busy all weekend, didn’t pay any attention til today) making it somewhat obvious by the time it was finally “released”.
I originally decided not to post this, but if a 4k Canon firmware is on the horizon then maybe it’s a bit more important. There’s a reason why companies like ARRI keep making incredibly expensive cameras like the Alexa, and this video by Nick Paton shows one of the important differences.
VizWorld.com 
Comments