Chris Smith
Forum Replies Created
-
Yeah, the POST article sums up the whole thing from the horses mouth. In a nutshell they shot tons of stills and moving footage of watercolors and inks till they had quite a library. Then playing with grads, placement, mirroring, and integrating the footage of the performers they have it. It’s really like most great mograph: a display of good artistry as opposed to technical wizardry. It’s all basic stuff, just extremely well done.
-
It doesn’t have to be a situation where you have tons of 3D layers. The opposite could be only a handful of layers, but you have so many effects in the comp that the adaptive slows way down trying to render them.
-
Very cool. Didn’t know that. I figured it was a matter of time.
-
Not for render times. It’s for real time interactivity. When you are using a lot of 3D layers it’s a good idea to switch to OpenGL so that when you move the camera, the layers are converted to OpenGL and the GPU of the render card takes over doing what it does best. So the result is that you will see the layers move with full interactivity. But things like DOF, blurs and a lot of non-supported things in OpenGL gets turned off. It’s just like a video game. Video Games program everything to work whithin OpenGL so it’s fast but you lose a lot of rasterized render effects.
Adaptive resolution shows you every single effect you applied just as you have it, and it tries to speed up interactivity by adaptively lowering the resolution while the playhead scrubs or you move the camera. But it’s slow because it uses the CPU and accounts for every single effect you have creating your comp.
So for example if you are doing every noobies favorite first AE project, putting pictures in 3D space then using DOF, you would set it to OpenGL while placing the pictures and animating the camera, but then switch to adaptive resolution when tweaking the DOF since OpenGL doesn’t support DOF.
But at the end of the day, no matter what setting you have it on, RAM previews and render outs are always rendered the same from the CPU.
-
Search the archives for “HDV”. I’ve seen this discussed many times.
-
Ah, for FX and not correction really. I just use Levels and usually grab one of, or a mix of the color channels. It’s rare that I want the whoel mix of RGB the way it is. I’ll use the channel mixer set to monochrome then mix in the individual color channels till what I want isolated is the strongest. Then just add levels and crunch up the levels to define what I want to keep. This alone does most of the matting I ever need. I do this instead of keying most often. For example, to simulate trapcode’s Starglow, you copy the layer, crunch the hell out of the levels till only the brightest parts are still around, use hue/sat to bring the saturation back down, then add 2 layers of directional blur (for a star shape) then SCREEN it back over the original.
-
Animate a mask around the object to punch it out of the shot. Replace with a black solid that matches the BG. Then simply animate the text over where the object used to be. About as straight forward as it gets.
-
But if I understand correctly you want to isolate Shadows. midtones, and highlits as oppsed to specific colors, yes? Working with Shad, mids, highlits is considered “Primary” color corection then grabbing specific colors (what in Davinci is called a kilovector) is considered “Secondary” color correction and many colorists will tell you to always use Primary color correction for the main look with secondary for touch ups.
The closest thing you will get to good color correction in AE is the free “Color Finesse” plug-in. Look at the “Ranges” section. It is where you can tweak the barriers between the shadows, midtones, and highlits. But many people find the preset ranges to be perfect as is. Point being is that any decent color corrector gives you hue offsets for shadows, mids, and highlits without trying to reinvent the wheel with mattes. Everything you see on TV has been transferred using basic “primary” color correction (manipulation of shadows, midtones, highlights – both levels and hue offsets) adding power windows to isolate large simple feathered areas for overall treatment of an area, with secondary color correction to fix something that pops out (like too much red in a face or whatever).
-
Thanks for doing that. I was curious myself. I have a couple MacPros in the mail and was wondering if I should just keep AE on the G5 for a while. Love to see any other tests you have going.
-
Sure, Yikes. So you tried it on his clip and it worked out well?
I first figured this technique out when I was finishing a job. We had a shot with frosted glass that took up 1/2 the screen which would be the BG plate that we put ending graphics over. But me and the client hated these black flecks all over the glass that just wouldn’t come off on set. They asked if I could paint it out. The clone tool in AE 5 at the time was non-existant or not very easy to use. So I sat and panicked/thought of how I could exploit the darker black flecks and do some overall process because painting out about 40 flecks would have been a nightmare. That’s when I messed around with blurring the dup then using lighten. When I did that and they all “magically” disappeared without breaking out a single brush, I decided it was the coolest trick around 🙂