January 23, 2006 at 12:20 pm
I was wondering if anyone knew how to turn video to 3D that could be viewed with the old red & green/blue 3D glasses.. I thought maybe color 3 seperate layers of the same footage, each colored red, green or blue with about 50 percent opacity and offset them slightly from each other might work, but maybe someone knows how this effect was traditionally done?
January 23, 2006 at 2:45 pm
Others will correct me if I’m wrong, but you can’t turn video shot with one camera into 3D. There’s no depth information. Normally, this type of 3D (anaglyph) was shot with two cameras side-by-side to represent the viewer’s eyes. The depth is achieved by combining views from the two cameras, similar to the way our two eyes process information. (Lose one eye and you lose depth perception. It’s the same with using only one camera.)
If you want to try offsetting channels, you’d have to offset them by a different amount for each object, depending on the object’s distance from the camera. Personally, I think that’s a lot of work. 🙂
I believe that there’s a 3D glasses effect in AE (from the old Cult Effects bundle?), but I’ve never used it. I’m not sure what it does, so check it out.
January 23, 2006 at 4:24 pm
Of course you’re right – the method used for that style of 3d utilised two cameras to capture the footage.. I suppose I thought that only the new type of really 3d film used specialised technology and the old 3d films just did a cheap effect, like a film version of those 3d images you used to get on the back of cornflakes packets.. maybe thats the kind of effect the plugin you mention does.. if it does, that will do nicely.. I must have a look for that so…
thanks again =)
January 23, 2006 at 4:49 pm
If you have the budget you’d want,
they can turn your 2D film to 3D.
January 24, 2006 at 3:05 am
I did 3d anaglyph print/video professionally for a few years. The short answer is that you cant effortlessly create 3d video from 2d video. The long answer is:
The principles for old anaglyph/polarized 3d are the same as the newfangled lcd-flicker glasses. The 3d is present because there are two different images recorded or created offset from each other. The way to fool your brain into seeing 3d from that is to show a different image to each eye. Then, the same process in your brain that interpolates depth in real life can do the same thing to the two images.
The old method was to process two black and white films, then burn one as red and one as blue/green onto a final plate. when it is projected back as this blurry mess, the 3d glasses filter out all but that color entering that eye, and therefore you only see one image per eye. viola. 3d perception.
The polarizing method used two projectors, each with a polarizer over the lens at 90 degrees from each other, matching the polarization of the two eyes of the 3d glasses you’d watch. Using this method you were able to have full color images. Also, you could take your glasses off and the screen would be a bit blurry, but not a rainbow.
So, the process of creating 3d anaglyph in after effects is pretty simple. After shooting video with 2 cameras side by side, you take one of your cameras footage and replace the red channel with the red channel from the other camera. Viola. as long as you replace the channel to match the glasses you are using, you’ll now have full anaglyph 3d. The anaglyph plugin (3d glasses) for after effects just does that for you. It has nothing to do with actual 3d creation. it just allows you to easily merge the channels from two video sources into one anaglylph image.
NOW… lets talk about creating 3d from 2d when there arent 2 handy stereo sources handy.
Most of anaglyph 3d is broad strokes. You notice the big changes more than small ones. As a result, if you put one or two 3d offsets into an image, the viewer will see a lot more 3d than is really there.
So, if you can take the footage and add an element into the image (a fake foreground image, a cut out window revealing a different layer, text on the screen, etc), then you can create a 3d seperation from the video plate and give the 2d video a 3d “depth”. You could even create the front archway of a theater and have that floating in front of the screen and have the 2d footage “behind” it in 3d space and give a 3d perspective to look at.
The easiest way to do this is to create a full composition with all the elements in it where they belong, then duplicate this composition. Name one “left” and one “right”. The 3d illusion comes from the differences between the images for each eye. When things converge (right moves left, left moves right) then they appear to come out of screen level and towards the audience. When they diverge (left moves left, right moves right) that object is pushed back in 3d space. You dont have to move both the left and right object though. If you move an object in the right composition to the right, that object will be diverged relative to the same object in the left comp and will move “back” in space. keep in mind that it is much easier for the eyes to resolve a deep diverge than converge.
It probably sounds confusing, but it really isnt. As long as this is for a short bit of a video, you can do killer 3d tricks to wow an audience without spending a fortune. The best thing you can do is grab a pair of 3d glasses and the 3d glasses plugin in ae (maybe only in the production version?). Make a layered comp, duplicate it, then bring those two comps into a third comp where you apply the 3d glasses plugin to a setting appropriate for the glasses you have. now, go into one of the comps and move an object left/right (NOT up/down) and go back and see how it has changed in 3d space. You can have moving objects converging on a point and appear to be flying from far away to up close.
1.Dont take an object that is layered on top of another and push it into 3d space behind the other. your brain cant resolve something being in front of and behind at the same time and it will just give you a headache.
2. Go back into the frame or forward. dont try to do both. it seems cool, but it can easily nauseate the audience.
3. 3d anaglyph space as described is unrelated to scale of an object. You can have an object converging and it will appear to be coming towards you in “depth” but since it wont be scaling (unless you scale it), it will also appear to be shrinking as it comes “forward”. In reality it s just staying the same size and its relative depth is moving… much like when they do that trick in movies where you zoom out while dollying in so the person stays the same but the background zooms.
anyways, I hope that helped. i think 3d video is an underused gimmick. every audience we’ve used it on has loved it. We even had a product rollout with all the speaker support in 3d for a laminate company. It turned out great.
ok, one last thought. if what you wanted was the “look” of a 3d movie for a person watching the video without glasses? just offset the red channel a bit to the side. viola. hehe.
January 24, 2006 at 2:20 pm
Thanks for the post, floop. Great information. 🙂
January 24, 2006 at 4:13 pm
Wow, floop.. that is a massive help… I’m looking forward to experimenting with the methods you’ve so generously shared.. Very informative indeed.
Thanks a million man,
All the very best.
May 17, 2009 at 8:56 am
it works great !
i’ve just tested it with 3D glasses after effects plugin
but is it possible to keep some colors after using this method ?
because it’s very green here .. (light green)
Log in to reply.