[Tudor “ted” jelescu]

Flash can play f4v and flv and the quality is really good- just encode with Adobe Media Encoder and not AE. Open your project in Encoder, select the comp you want rendered and choose a preset.

Tudor “Ted” Jelescu
Senior VFX Artist

[Tudor “ted” jelescu]

ProRes 422 (HQ)- it’s the one most folks use between AE and FCP.

Tudor “Ted” Jelescu
Senior VFX Artist

[Tudor “ted” jelescu]

Without seeing all your comps and precomps I can’t say for sure, but either you have a 2d layer in between 3d layers, or the position of the layer forming the cubes is not really as it seems and some layers from the small cube may actually touch the layers of the big cube in 3d space. Instead of precomping the layers for the cubes, I would link them to a Null, name the null for the cube (small cube and big cube) and then hide them so that you’ll only get the null layers in the comp. This way you can position and troubleshoot all in one comp instead of trying to fix things in 3 comps.

Tudor “Ted” Jelescu
Senior VFX Artist

[Tudor “ted” jelescu]

For the glowing silhouette you need to rotoscope the person (cut out the animated shape), comp that on top of the same footage without the mask, and apply glow to the masked layer (silhouette).
The visual echo of the arms can be achieved using the Echo effect.
https://help.adobe.com/en_US/aftereffects/cs/using/WS3878526689cb91655866c1103a9d3c597-7a57a.html

Tudor “Ted” Jelescu
Senior VFX Artist

[Tudor “ted” jelescu]

If I understand correctly what you’re after, I would try to use a AE light in 3D space and Trapcode Lux.
Just to be clear- is there a single light source or does each piece have it’s own internal light source?

Tudor “Ted” Jelescu
Senior VFX Artist

[Tudor “ted” jelescu]

I would go for Green for sure, especially if the screen is that color.

Tudor “Ted” Jelescu
Senior VFX Artist

[Tudor “ted” jelescu]

Did you comp the white shapes on a solid black background and then render, and then use the final?

Anyway, here’s another method I thought of: render the two at 100% separate on black. Use one as the alpha for the other to comp again on black and render. You’ll get the intersection alpha this way.

Tudor “Ted” Jelescu
Senior VFX Artist

[Tudor “ted” jelescu]

SynthEyes does extract textures and rebuild background plates. Maybe you can use that.
If not, my other idea involves 3d camera tracking as well. Camera Tracker from The Foundry is a good tool.
You can paint the whole field in PSD, clean grass and marking lines. Then you would need also the side elements- advertising banners and whatever else the players overlap with. You would need to build the field and the surrounding area in 3D space in AE, match the position in every shot and apply the camera data from that shot to match the virtual set you made to the move in the clip.

Tudor “Ted” Jelescu
Senior VFX Artist

[Tudor “ted” jelescu]

For future projects of this kind and less headaches I recommend using Plane Space from Red Giant.

Tudor “Ted” Jelescu
Senior VFX Artist

[Tudor “ted” jelescu]

I’m not sure I completely understand what you’re trying to achieve, but here’s an idea:
Use the alpha from each clip as matte (channel/set matte) on a white solid layer(one for each clip). Set both white solid layers to 50% opacity and render over black. The result is your alpha, where the two intersected shapes are at 100% and the rest at 50%.

Tudor “Ted” Jelescu
Senior VFX Artist