When was the last time you looked back at an image and noticed something about the lighting that you wished you could tweak or alter slightly? I’m sure most of us have been in this position at one time or another. Up until now, it’s simply not been possible to even imagine being able to do this. Welcome to the strange new world of “computational lighting design”
Just like Lytro’s Light Field Camera technology allows you to change your focal point after the shot has been taken, we’re moving into an era where software is giving us the ability to actually go back and play around with the lighting when we’re looking at the images on the computer.
Our own Fstoppers staff writer and architectural photographer Mike Kelley was directly involved in supporting the Cornell team that developed this software (a number of his images appear in the video linked in this article).
Mike and the Cornell team worked together to design the software that delivers “computational lighting design”. The software takes over a hundred different frames where the lighting angles are changed slightly each time, using selective amounts of “edge lighting”, “fill lighting” and “diffuse color lighting”, seen in more detail in the image below.
Importing all of these frames into Photoshop as layers allows a user to combine or mix layers and then focus the type, quality, directionality and power of light in particular areas of the shot. This allows them to play around until they have got the desired lighting set up for the scene. Some examples of scenes lit differently by the software can be seen below.
I’m all about doing my best to understand light and get it right before the shot is taken to save time in post, but this sort of technology is interesting if only because it seems such an impossibility to imagine we would be able to do this, even as recently as a few years ago. According to the report, even professional photographers reported it saving time (although it doesn't point out in what particular area this time saving was made).
You can read more in the Cornell Chronicle . The video below explains how the software works in more detail.
Is this the future of lighting on set? Will we be doing more work in post than we do today to light our set after the job has been shot? How would you guys think about applying this to the work you currently shoot? Would love to hear your thoughts in the comments below.
Images courtesy of Cornell Chronicle & Cornell University
All of the "strobist" type of lighting formulas that young photographers are now struggling to learn with off camera flash will be emulated in the future by the equivalent of lighting effects plug-ins. Of course, software is cheaper and less time consuming than setting up actual light sources, so it will eventually replace the need for professional lighting.
Won't happen this way, programs won't emulate light, unless you're doing a render. Plugins are not able to indentify a 2D object, as a 3D object, and iluminate it the way you would like.
The point of this technique is that you can get high end results, with low budget lighting, at expense of more time on post production. I myself did it sometimes when the lighting equipment was limited to speedlights. Myke does it too, and have some really good explanations on his website.
Interesting - do you have any links to some of the stuff you used to do or other examples i could see? Would be interested to check it out
This is not necessarily true. 3D mapping software is getting more user friendly to the point where you can set up a Kinect to map a room. (see http://hackaday.com/2012/06/01/3d-mapping-of-huge-areas-with-a-kinect/). It wouldn't take much of an imagination to predict that cameras could offer a sensor that provides the same benefits without a kinect. If your shot is understood on a 3D level through the computer, this would open up all kinds of manipulation capabilities beyond simply controlling the lighting. Here's hoping that future comes sooner than later!
yeah good point, i remember reading that when the Kinect came out.
I don't think it's that weird to hypothesize and imagine a future in a few years where we can sit a subject down, scan a room in 3D, position a few lighting sources that fire off multiple flashes in different directions and intensities, then throw on some modifiers and get them to do it again that we couldn't run through 100 different lighting permutations in a few minutes and then compose the whole thing behind the computer.
Not saying it's the right way to do anything, but it seems to be more and more when things are going.
If I'm reading this correctly, you have to have taken all sorts of photos with different lighting just to composite to what you want later. Too much work in post, get it right in camera is much easier.
Yeah, it's easier if you have a massive amount of different types of lighting (for certain occasions) and have a laptop to actually look at the results immediately.
Eric this is what i was curious about too. My guess (could be totally wrong) is that this technique allows you to prioritize options of different lighting and is merely in an early stage of development for what the software could do once the hardware (speedlights in this case) catch up to what we are trying to do which is effectively take a number of lit frames very quickly or successively with different light intensities, directions and qualities.
Considering it's funded by Adobe, I'm guessing we won't see this unless it's in a trial or part of existing (expensive) software.
Interesting. I tried a similar, but simpler app some 4 years ago. Don´t remember the name though
This is interesting. While I am firm believer in getting it right in camera, I think that it could come in handy on occasion. I don't think it's a replacement for correct lighting, but something to enhance.
BB, I agree. What's fascinating about this to me (regardless of the 'get it right in camera' or 'leave it to post' conversation) is that the concept is very interesting. Being able to play with light after the shot is intriguing, particularly if our idea of what was 'correctly lit' on set changes for whatever reason once we get back to the editing process
Have we come to the point that photographers can not understand and master light when shooting? What happened to getting it right when the image is created? While the concept is cool, this is one of those things that would contribute to taking the creative process out of photography. We already have a market full of young, willing to work for nothing, photographers. Now we are going to create the "no need to learn the art cause I can do it in post" photographers. Seriously, learn to light. It isn't that hard.
and also, learning to light is FUN let's not forget :)
I think the point is, someone will be having to handle the application of light on the scene or subject, whether it's shot on set or worked up in post. Great photographers who understand light will still be in demand, it's just an interesting concept that - in a very hypothetical world - in the not too distant future, you might now have a photographer who doesn't light a set until after the scene is shot, and then "lights it" via a computer after the shoot. Not saying this is the 'right' way to go about it, just was interesting as a concept for what might be coming.
Seriously. Why do you need to blend "hundreds of frames" when you should focus on thoutghful composition and lighting design? Get the lighting right from the start, do a few test shots, and move on to the next image. I don't need SW to show me how to blend images of different exposures or lighting.
My point exactly. Learn to do it right, not depending on someones magic software to do it for you. Seems a little less then professional to me.
I totally agree with your premise and original post. You're right, we're getting close to the point where a "photographer" isn't going to have to know lighting at all. This presents a problem because photography literally is "writing with light." So, how can a photographer be a photographer without knowing how to light? The answer is that digital imaging is a different medium and is not photography at all. Right now, it's still in infancy and people still compare it to photography. But it the future it will be recognized as a completely separate medium from photography.
Also, professional photography has always demanded function over form. In other words, clients have practical needs and concerns that take precedent over perfect form & technique etc. Any type of software of digital process that saves time and money will eventually become part of a "professional" standard of workflow due to clients' inevitable choice of form over function.
In the long run, what is happening is that actual photography (writing with light) is going to become so impractical that it won't command a professional market anymore. When this happens, form will finally trump function and photography will become an ART! It will be too impractical to be done professionally and will only be done by artists and patrons that support the arts.
Don't you think though that someone is still going to have to direct the lighting, whether behind a camera or a computer? Surely that means photographers will become a prized resource? It doesn't matter whether it's done before or after the shoot, those that can control, shape and understand light are surely going to rise to the top, no? Just a thought
Actually, It seems like you would need to understand and master light even more to be able to do this well. In fact, it looks like it gives you the opportunity to do it even bigger and better than you might otherwise. I'm the type of artist who could get totally carried away with something like this! :) Exploring all sorts of different lighting that doesn't interfere with the ultimate product unless I wanted it to. I like the amount of control you can have with something like this.
I don't think a lazy photographer would have too much fun.
good point
The Oloneo Photo Engine software allows you to turn individual lights in a photo on and off or dim them. I've used that software for changing the light sources in a photo to how I want them. I'm not trying to "sell it" to you by any means, but maybe check it out if you want to.
Hey Lars, Oloneo looks interesting (not used it personally) but i can see how you can change the light intensity and color temp - not sure you can actually change the source lighting in Oloneo though (direction and quality of light). Looks like a fun and interesting tool though, thanks for the insight
I did this manually 6 years ago. In my case I used only the lighting in the space and blended the exposures with meticulously painted layer masks. But I used a new technique at the time that allowed me to re-adjust the color of each layer until they matched. This was necessary because of the color temperature rendering from the different light sources. http://www.johnmaclean.com/leanarch_anatomy/index.htm
This looks exactly like how Mike Kelley does his work. He used to share his work on POTN a while back. Here is a timelapse of his work process.
http://www.youtube.com/watch?v=J-exuHchmSk
To all you people saying to get it right the first time or in camera, this is a perfect example of when it isn't feasible.
Tagnal, it's interesting you reference Mike because he was directly involved in helping develop the software with Cornell (see reference in the article).
Thanks for linking that video, hadn't seen it before!
Heh, I sorta missed that as I didn't know his name off hand. I just remembered someone sharing their work a long time ago on POTN so I went back and searched through the threads and found his name and the video so I could post about it. :P Great coincidence!
I use this all the time... I makes for interesting photos... it makes for photos unskilled photographers cannot copy... it makes me money... http://photobymike.com/light-painting.html
nice LP work Mike!
Nice work Mike, what are you using as your light source?
Purely low tech... Really small things: a small Mag light that takes 2 AA batteries with a fiber optic attachment to make a light source about 3 or 4 mm's. Bigger (like food) take the fiber optic off. Bigger? like the Church organ the Mag Light with 3 D sized batteries or the 5 D size if you wish. Bigger? like the outdoors of buildings boats, cars etc. I use $20 (aprox) Stanley 1M Series hand held spotlight(s) (up to 3 at once) (with or without gels). OR... I use Vivitar 285's with a big 4.5 amp battery on each one... With film you would need a lot bigger lights due to the reciprocity failure... But with digital cameras and a timer to keep running exposures unattended this is relatively easy compared to the old days. You have to plan ahead to make sure you cover the subject. Use gobos to block the light... dark clothes are nice but not necessary... make lots of exposures to blend together with masks. Never use LED lights they are some oddball color.
Wow, love the low tech solutions, thanks for sharing that's awesome!
Bringing some lights and getting it right in camera seems like a lot easier solution to me. And it sounds like being a photographer instead of a software jockey...
Sounds a lot easier to get it right the first time. Well, for photographers it does. I can see why this would be an appealing option to someone who has no clue about photography, or lighting. I would rather screw up my lighting, LEARN from that screw up, and then have that knowledge so I will get it right next time. Man, how'd all you film guys survive without computers to do everything for you???
Multi exposure, multi angle light sources with modifiers and "load files into stack". Thats it. Not a big deal.