This content has been marked as final. Show 6 replies
Does the 3D Glasses effect help to some extent? You may need to pre-compose quite a bit, but it would be an easy way to retain the correct perspective information. Beyond that You could possibly get something going with expressions by using stuff similar like that:
eyeSepReal=5; //assumed value in cm
eyeSepPix=(72/2.54*eyeSepReal)*0.5; //assumes 72dpi, value by half for one side
The above would give a simple linear offset. For 3D layers you may need to include additional multipliers to compensate for Z and figure in lens curvature. For the left side you'd simply subtract instead of adding pixels.
I've used the 3D Glasses effect quite successfully in this way. You create 2 identical comps containing the animation, then use expressions to control a camera within each comp, one spaced slightly apart from the other. The result is a comp representing each eye's view, combined in a third comp using 3D Glasses.
I've attached a very basic example project (CS3 only) that demonstrates what I mean.
One thing to keep in mind with your camera rig, is both cameras need to be equal distance to the same focal point. You'll run into problems if you just have 2 cameras spaced apart, with their POI's equal to each other: When the POI is off to the side, both camera's need to rotate around a null to point at the POI, maintaining their distance. My best solution for my 3d camera rig in XSI was to use bones to control the spacing between the 2 cameras (bone lenth), and another bone to control the focal length for the 2 cameras.
thanks a lot guys for your valuable suggestions!
i'ld hereby like to tell you something in brief that i am not making a 3d composition in after effects but importing some video - image sequences instead. infact, i've mad a 2d comp without a camera and there are no 3d layers in it. it's just a simple 2d comp with flat 2d layers containing videos made in 3ds max.
so, first i rendered out my 3d animation in max as tga sequence which i imported in after effects and applied various effects on it. then after finalizing the effects, i rendered out my comp as tga sequence.
but now i want to make this whole "mono" stuff into a "stereo 3d" one. so for that, i already had the scene ready in 3ds max and i just made 2 cameras (left & right) with equal distance bothways in the same scene, with eye separation and rendered out both the camera views as two separate sequences. after that i duplicated my comp in after effects and made 2 separate comps from that (left & right) and just replaced the left and right comp layers with the new left and right image sequences respectively. so now i am manually moving my position points of the effects and masks, with equal distances. thus it becomes a very hectic job to do so, every single time (as needed).
so is there any other way to work with this kind of stuff ?
i hope that i explained my problem to you properly!
anyways thanks again to you guys.
I was listening to Hybride speak at Siggraph last week about creating Journey To The Center Of The Earth. Basically they said rotoscoping really sucked on it, cause you have to do twice the work. And you can't just shift everything over, because it doesn't always line up. Sometimes things are occluded in one eye but not in the other.
means, the bottom line is that there is no solution for this kind of stuff, yet !
dont know how people work with 'stereoscopy' in such situation !!!!!