Slashdot Mirror
Apple Patents Directional Flash Tech For Cameras
tekgoblin writes "A patent application has surfaced that shows Apple's attempts at creating a new way for a flash to work on a camera. The way the new flash works is very intriguing: a user can select a dimly lit area of the photo and the camera will try to illuminate just that area with the flash. The way Apple is attempting to accomplish this is similar to the way the autofocus works on the iPhone 4 where you can touch the screen in certain areas to focus on that area. Instead you will be able to light up that area with the flash. This is accomplished by the camera flash passing through a 'redirector' so the flash can be placed other than directly centered when a photo is taken."
This doesn't make any sense. I'm *sure* I heard Jobs say that he was against this type of technology.
How does this redirector work? the problem with flashes on camera is that they are coming from the point of view of the photo. This creates rather unflattering light.
You can redirect a flash by aiming it, but its still coming from the same point in space as the camera. This isnt ideal or good either.
The best way is to get that flash off the camera... but if you cant, as would be the case with an iphone... it is best to bounce it by redirecting the flash onto a wall to the left, right if you can, or ceiling. Generally up and to the rigth and left work well, as it forces light to bounce off the wall, which in effect makes the wall a large light source.
The problem with the flash being on the phone is that it is still a small light source. Small light sources cast hard shadows. This redirector wont change that, unless it can bounce light off a surface such as a wall. Which i dont see it doing as it has limited mobility being stuck in the back of the iphone. Generally with higher end camera flashes, you can rotate them in 360 degrees left to right and have a large up and down range of movement so you can point it right at the ceiling. you cant do that with an iphone.
We'll see.
Sounds like a cute gimmick for camera novices, but not a new solution to anything other than perhaps interface. Light is light.
I'm a professional photographer and I've been using flash zoom, feathering etc for years to achieve this effect. Guess I won't be allowed to do that anymore without asking Apple for permission first?
http://www.meejahor.com/2008/06/06/feathering-two-lights-for-the-price-of-one/
http://www.meejahor.com/2008/09/29/feathering-its-like-off-camera-lighting-but-faster/
(Just kidding. I know it's a patent for a specific method, not the technique.)
...manually aiming it? People have been doing it for decades with regular cameras...
We don't have to hold it in a certain way to make it work properly.
The only time I bother taking a picture with my phone's camera as opposed to a normal camera is if it's something happening spontaneous and I want to take a shot immediately.
If I'm going to take the time to make adjustments and setup lighting I'm not shooting with my cell phone.
That said, if the camera can auto-select dark spots and light them without over-lighting other areas or otherwise screwing up the shot, I could certainly see that as a good thing.
Photoshop filters are no replacement for actual light.
Photoshop filters are no replacement for actual light.
...Yet.
Eventually, a raw 3D picture with color data can be rendered into any picture you may want.
We will be able to take a picture in awful conditions and then recreate a "fake" picture with the 3D model, the true colors and any lighting we want.
I predict this will exist in under ten years.
How do you redirect light with a solid state system ?
Will this use the same tech the new breed of laser projectors will use?
Actually there exists technology today to gather real time 3d information with lasers. Google ASC 3D.
Eventually, a raw 3D picture with color data can be rendered into any picture you may want.
With no light, there is no colour data.
With no light, there is no colour data.
You need much less light to know the color of a surface than to make a decent picture.
A badly directed light that gives you an awful picture, does give you color data. Take that data, apply on 3d model, create light sources, render picture.
You need much less light to know the color of a surface than to make a decent picture.
Not true – taking a picture is by definition determining the colour of the surface –if you can get the colour of the surface exactly, you can get the picture.
Well, I have to say, it is a novel idea as far as I can tell. I could probably do one better by combining the power of a projector lamp and a DLP mirror system to paint a rather precise lighting system for the purposes of portrait photography. Light can be manipulated with very precise detail, coloring and intensity over the whole scene, not just one point. (Now, someone go patent this idea...) Using this technology, you could photoshop an image before you take it.
As someone pointed out, it is not so easy or as good when photos are edited after the fact than before. The reason why, I will assert, is that there is an unlimited range of variables of light while there is a far more limited range of variables of pixel data. The act of capturing an image on a CCD is already lossy compression of information. By setting up the image before-hand, you are increasing your ability to edit a final product in a more pleasing way.
I would be interested to know how Apple intends to integrate this into an iProduct. iPhone/iPad wouldn't be particularly good at this type of photography I don't think. To accomplish this, a complex focusing system would have to be implemented and while I have heard of liquid lenses (here on slashdot) before, I can't help but believe that the throw distance of such projection technology would be rather short.
Still, all in all, this is a neat idea. And it's not quite a software patent, so I'm okay with it.
Yet another way to really fuck up your photos. If you know anything about professional photography, you immediately know this is a failed "solution". In many cases when you light a scene for photography, it's the DIRECTION that the light comes from that is important together with the amount of light. That's why you rarely see camera-mounted flash used in the studio, strobes (flashes) are positioned away from the camera so as to light the scene in a certain way from one or more directions. With the proposed "invention", the direction light comes from will always be the same, close to the lens. It doesn't matter that it's only lighting a part of the scene.
Now, if Jobs has found out how to use the reality distortion field to his advantage and actually BEND light passing through the air... that would be impressive.
Maybe you block off parts of it with, eg. a low-res LCD - no moving parts!
No sig today...
Not true - taking a picture is by definition determining the colour of the surface -if you can get the colour of the surface exactly, you can get the picture.
That's only partially true. In a photo the received surface's color comes from the position and angle of that surface relative to the camera and the lights. If you know the 3D map of the object you can reverse that composition and gather the original information. That original information is what I called the "color data".
For example, with a picture of a sphere and a 3D map you can determine the real color of the sphere (not the color of the light it's currently reflecting towards your objective) and use that information to render a new sphere with a different source of lighting.
Not only source but also intensity, color, etc. Essentially that lets you render the picture you want over your 3D colored canvas.
I thought Apple hated Flash, stating poor performance and what not.
That's only partially true. In a photo the received surface's color comes from the position and angle of that surface relative to the camera and the lights. If you know the 3D map of the object you can reverse that composition and gather the original information. That original information is what I called the "color data".
Yes, and said colour data cannot be gathered in any way other than collecting light bounced off the surfaces... No light, no colour, it's as simple as that.
Isn't one solution just to take two photos a split second apart (one pre-flash and one with flash), then blend the two images together, using the region of the 'flash' photo that the user selected as the 'flash region'. If its not doing that, then where do I make a patent application? :)
Try to take a picture of a male common damselfly or even a dragonfly in different light conditions. Light is not just reflection, it is also refraction (especially on the more "interesting" photographs). That cannot be modelled realistically based on color data from light falling from the wrong direction. You need knowledge of the actual material to model that and even then it is least likely to be realistic.
There are many examples - the colour of the wings is different depending on the angle at which the light falls on it; same for any other insect; same for rocks, crystals, birds (try taking pictures of a common kingfisher or a sunbird from different angles); open water; in fact nearly anything worth taking a picture I can think of. Even human face if done properly. Try taking a portrait in a standard "halo" shot (works best on pretty blonds :-). Try that with flash and/or additional supporting lights and try taking a "plain" shot and try to get the same effect with photoshop afterwards. You will see what I mean - if you do not have the lighting from the right angle in the first place there is no way you can simulate that "gold shimmer" look from a l'Oreal commercial after that.
In photography light is everything. It is what makes the 10000$ difference between a work of art and POS produced by a point-n-shoot. If the light was not there in the first place and _at_ the right angle the necessary colour data will not be there to record.
Baker's Law: Misery no longer loves company. Nowadays it insists on it
http://www.sigsegv.cx/
Apple does it and they get a patent. When I did it I got arrested.
sounds like Microsoft sold another license for their 'wedge' lens technology.
Actually, that seems to me like the perfect way to aim the flash at just someone's eyes :p
A polar bear is a cartesian bear after a coordinate transform.
Aside: The person who did the beating down uses an iPhone.
Lesson: don't be a twat and call people unintelligent because of the technology they chose to use, many made an informed choice and just had different requirements to you.
>>>>>Photoshop filters are no replacement for actual light.
>>
>>Eventually, a raw 3D picture with color data can be rendered into any picture you may want.
Eh... maybe in the era of Star Trek (2100s) but for now it's not possible. Even digital zoom doesn't work properly, just blowing-up the existing image rather than interpolating missing data. I agree with the Grandparent Poster that actual light is preferable, just as real actual zoom (using a lens) is better.
Still it's pretty impressive how far Digital cameras have come since 2000 (when they were basically junk).
Looks like it's time to get rid of my old S-VHS and 35mm analog units.
"I disapprove of what you say, but I will defend to the death your right to say it." - historian Evelyn Beatrice Hall
Years ago we had industrial photographers come in and take a picture of the equipment
we were building. Lots of curved stainless steel. They brought in a lot of lights, set
them up high and low all over the place, and our boring-looking equipment started
to look Really Good. Pictures came out great.
Let me see if I have this right - you're suggesting using lasers instead of light?
The top scored comments do not consider that the iPhone has enough power and resources to augment photos.
You could take your photo and then wave the camera around in the air or even walk over to the side or closer to your subject, and the phone could be selectively lighting the scene while adding these frames onto the image. It can use the accelerometer with computer vision techniques to understand where the camera is moving. It can learn what the 3D shape of the subject is and computer 3D masking.
What I think this might mean in terms of low-hanging fruit is that the camera will shoot an image, detect the edges of dark areas that need lighting, then provide audible cues for the photographer to swing the camera to the side. The camera would then be able to shoot a series of frames automatically, blur them together, and make the tiny size of the flash appear bigger. It might also be possible for the flash to be actuated based on gesture input or shape recognition to simulate light painting (at least to highlight portions of an object somewhat artistically).
Finally the camera app could allow the user to selectively combine portions of this captured light field so that a single shot could be made to express different moods.
Approaching the subject or actuating a zoom, aperture, time delay, etc. could allow more detail, HDR, blur removal, etc. to be used to augment the image. In the end a processed 3D movie in time and reconstructed 3D scene, potentially including the photographer and what is behind him, or even including different lighting from different times of day or separately illuminated lighting, would result. This kind of a 3D photo-movie with tons of metadata could become the new photo we want to archive.
I hope this is enough to forestall any more patenting!
Not saying this will help you take good pictures. They will probably suck except for some incredibly lucky awesome ones which will be further photoshopped and used for marketing. But a more finely aimable flash combined with motion and processing in the space and time domains will certainly allow smart cameras to capture a lot more data. More so if it gets dual lenses and XY audio pickups. The iPhone seems to be the best platform. If only it would get better photo hardware...
Seriously I could see a next-gen iPhone with a honking big HD video lens and hot-swappable dual memory cards, generating tons of 3D and lighting data that can be editable on a MacBook Pro. Finally a good reason to get a multicore machine!
Or as was scrawled on the wall of a painting studio I took a class in once, "Even shit can be beautiful if the light hits it right."
Free Martian Whores!
The patent looks interesting, but also looks like something that is just a computer controlled version of a technique that has been around for a long time. I have some old potato masher shaped, flash-bulb, strobes. A few of them have the reflector dish that is just slightly off center and a little bit movable. Some call that dents and age, but it looks a lot like Apple's patent. Move the reflector just slightly, and you can highlight a dark area and remove some light from the bright parts. Easy enough and, on the strobes that this was an intended feature, pretty old. Stick a few MEMS reflectors behind the bulb, and you can direct the hot spot of the light with a small in-camera lamp that same way.
The target crowd for this don't want perfect, aesthetically pleasing, artistic photos; they want snapshots. Like the original Diana and Brownie crowd. That still works as you are likely to find a Diana photographer carrying an Apple product now days, as well.
Oh, for goodness sake, when is timothy going to a) learn to link to the original source and not some third-rate blog and b) learn to distinguish between a patent application and a granted patent?
Link to original source: Apple Working on the Next Wave of Digital Camera Technologies
Link to U.S. Patent Application 20100238344
What one fool can do, another can. (Ancient Simian Proverb)
Yes, you can have one.
There are no karma whores, only moderation johns
I'm pretty sure that zooming the flash along with the lens was already being done on P&S cameras back in the 90s. Is this a case of tagging "on a telephone" to an existing patent?
Support SETI@home
Why do all Iphone users insist on telling the world that their platform is okay? Could it be a proxy for homosexual tendencies?
Not really very rhetorical that one – the reason is simple... Because the entire geek world insists on telling iPhone users that their platform is not okay.
No matter how much you spend on computing technology, you will never find the cute boyfriend you seek by displaying Apple devices.
That's okay, I'm happily married.
However, if you think that Apple crap is easy to use, then you're more deluded than the typical meatbag.
I rest my case about twats and telling people apple kit isn't okay ;)
Lasers are light, which is a spatially coherent, narrow low-divergence beam. Light Amplification by Stimulated Emission of Radiation (LASER)
It is what makes the 10000$ difference between a work of art and POS produced by a point-n-shoot. If the light was not there in the first place and _at_ the right angle the necessary colour data will not be there to record.
At least in the 35mm film days, a point and shoot could equal an SLR using the same film stock, so it was about what you do with the equipment rather than what equipment you used; an artist using a point and shoot could take a better picture than some dumb rich guy with a thousand bucks of SLR.
It's more complicated with digital, as SLRs have better sensors (particularly larger pixels, rather than more megapixels), but there are some new compact cameras coming out that redress that balance. Sadly there doesn't seem to be a huge market for small, high quality compacts.
Sigs are so 1990s. No way would I be seen dead with one.
Easy champ. This may come as a surprise to you, but not everyone shares your blind hatred of Apple. I know it's fashionable to bash Apple now, and you want to hang with the cool kids on campus, but being a follower is not going to make you popular.
Whooosh
Citroen DS lights work this way. In order to get a bit more visibility out of the puny late 1960s headlamps the reflector behind the lamp would pivot to point in the direction you pointed the steering wheel.
DS 21M phares tournant
black and white are colours
At least in the 35mm film days, a point and shoot could equal an SLR using the same film stock
Not exactly. The primary reason people would pay big bucks for the SLR camera is because of the difference in lens selection and quality. Other than artistic skill, there's no single component that is going to make a bigger difference to the look of the photograph. All other things being roughly equal, a $1,000 lens on an SLR camera is going to be capable of producing a better photograph than a $99 point and shoot, from a technical perspective at least. There are some photographs that you can get with an SLR that you'll never be able to get with a point and shoot. High speed action, or extremely low/high depth of field shots, for example.
an artist using a point and shoot could take a better picture than some dumb rich guy with a thousand bucks of SLR
This is absolutely true. Still is, even with digital.
Sadly there doesn't seem to be a huge market for small, high quality compacts.
Sure there is. In fact, that's probably the largest market segment for digital cameras. Quality is getting better all the time, even as camera sizes are shrinking. Heck, they're starting to stick cameras in phones that rival the high end consumer digital cameras of only a few years ago.
The best way is to get that flash off the camera... but if you cant, as would be the case with an iphone... it is best to bounce it by redirecting the flash onto a wall to the left, right if you can, or ceiling. Generally up and to the rigth and left work well, as it forces light to bounce off the wall, which in effect makes the wall a large light source.
Simple - you bump two iPhones in "flash buddy mode" - from that point use the accelerometers to determine their relative positions, and have the first signal the second one via bluetooth to flash. The first one can tell the user, "no, point it down a bit" to get the direction correct as the current geometry of each phone are dynamically updated in the controlling unit's 3d scene model (I presume the autofocus has already determined the distance to the subjects).
My God, it's Full of Source!
OUTSIDE_IP=$(dig +short my.ip @outsideip.net)
Technically black is not a colour. I guess you could call it a shade, but that’s as far as I’ll let it go.
Alexander Peter Kristopeit bought his basement from his mommy for one dollar.
no black is a colour. in an additive colour system (e.g. paint) black the presence all colours. in a subtractive colour system (e.g. stage lighting) black is the absence of colours. with white the opposite respectively.
using your logic all colours in the RGB scale aside from red, blue, & green are not colours.
Those aren’t colours, they’re pigments that filter out colours. If you filter out all colours, you get black. It still isn’t a colour.
Alexander Peter Kristopeit bought his basement from his mommy for one dollar.
So if you want to be a fashion icon , MBT shoes is the best choice. While world cup t shirt is a very casual shoes a lot of people become very involved with the addictive appearance. These cheap MBT shoes are a ability account and the ideal additive to acceleration up your confidence! The air jordan 4 is the most popular this year. Supra shoes Designer net supplied perfect service . To buy the chanel purses product, the easier has been formed by cheap air jordan shoes online.
http://www.mbt-shoes.com