[Steve Bentley]

You might also want click the movie name in the material and hit the animation tab and then hit calculate. It used to be you had to tell C4D that it had more frames that just the first one.

BTW there is also lateral displacement of the cubes caused by the texture, is that correct or did you just want straight in and out.

[Steve Bentley]

Your file worked for me. Here it is resaved.
Try relinking the texture in the material. It was black when I got it.

13197_demoresaved.c4d.zip

[Steve Bentley]

Or combine our last two posts: Grab frames that have hands off the keyboard in certain areas (if there isn’t one clean frame) then stack these up in Pshop and reveal-paint the layers so that you erase (or mask) the spots with hands. A clean keyboard will emerge. Flatten and use that as your difference matte in AE.
You can also do this in AE. There is a clone stamp you can use “in time” so that you are revealing parts of a frame that is further along the clip to the one you are on. But that means AE is having to calculate this for every frame. It will be simpler in Pshop.

[Steve Bentley]

Gifs can have a unique color space – less of a space really and more of a distinct palette (called indexed color). And AE can have troubles with that. So its always better to turn that gif into a non compressed format and something that is in a similar colorspace to your final comp. It can be converted to a set of frames or an AVI or QT movie (if your AE still supports that).
For AVI try Uncompressed and for QT try animation best.
But there’s the whole problem of changing the gif format into something else.
You can export from the pro version of QT to any of its supported formats, wondershare has a converter too. Do a search for video converter. AE can do it to if that’s all you have in the timline – export the movie and then bring it back in for you comp.

[Steve Bentley]

You might be better to use the mograph selection node after the cloner node and get rid of the python node – its kinda overkill here (unless you have it doing something else in another section of code). I don’t see how you can iterate through the cloner with the set up you have. The mograph selection node has an index port where to can add an iterator node, which if needed can be used as an indexer for the python node too.

[Steve Bentley]

Is there any clean footage of the keyboard without the hands? You might be able to extract a difference matte if you combine the clean footage (or even one frame) with the footage of the keyboard with the hands.

[Steve Bentley]

Don’t forget aerial density/diffusion – this is the effect that humidity and dust in the air imparts to any image (at least on earth or any other planet with an atmosphere and either wind or water). The bigger something is, the more air there is between you and the far reaches of that object and therefore the more aerial density/diffusion will occur.

This tends to make things more blue and lighter the farther they are from camera (water droplets scatter blue light more than they scatter the other colors – its why the sky is blue) so air with water drops in it will impart a blue hue to things the farther away they are. The saturation of colors is less too – this is partly to do with the scattering of the colors in the reflected light and because the more water (in this case vaporized water) the light has to pass through, the more you end up with just the purple end of the spectrum. Go look at pictures of fish under water – the ones in the distance have no red in them, very little yellow, if any, and only some green, but they all have blue.
But this effect is also because there is so much suspended dirt in the air and dirt tends to have a muddying effect on anything – the color of your new car looks less intense with a layer of grime on it. It mutes the color because if its own dull color but it also lightens the color because light reflects off it instead of passing through. You are probably seeing more light reflecting off dust when you look at a distant mountain than you are seeing light reflecting off the mountain. Whatever color that dust is, that’s the color the distant things will lighten to. On earth its paynes gray, because we’ve got both mud and water, on mars it would be a light pinkish orange. (so the chances of there being little “green” men on mars is pretty low, and if there are they will look completely fake because the light they reflect will fight with the above.)

One of the classic things to help making things look big is what we call “signposts”.
There are two kinds of these things but they are both extra objects that get thrown into the scene, objects we know the size of, and when seen in context with or near the big object we can get a sense of scale. It can be as simple as having some birds flying by at great distance but still in front of the big object. We all know how big birds are so our brain says, “if they are that small they must be x far away and if they are that far away and still in front of that thing then…. whoa!”
But what if you are shooting on planet X and they don’t have birds, or their birds are the size of buses? Then use a repeating object – think: a line of fence posts. Show an object in the foreground, say next to a person, so we get a sense of how big this new object is compared to something we do know: the person. Then repeat that object back into the distance – it doesn’t have to be in a line like a line of fenceposts (but that can help), it can just be scattered about. But make sure there are really small ones back in the distance that are both recognizable as the ones we can see in the foreground and in front of the giant object. The brain will make the size association and make the leap.
You can also use “signposts” to create repeating patterns that follow perspective lines leading the eye back. Every time that repeating object gets smaller or the implied lines of perspective get closer your eye knows its going deeper into the scene.

There is something called high frequency and low frequency detail. Things in the distance tend to have less detail because they are far away so we can’t see every little eyelash on the massive head, because those eyelashes are smaller than our eye or the camera can resolve, but also because there is so much water and dust in the way (and air current distortion) it’s actually making a blurry mess of things, so the details gets soft: the noisyness or the frequency of the detail is very low. Compare that to someone standing right next to you where you can see every hair on their head and every pore in their skin. The noisyness or frequency of those details is very high.

Finally, due to aerial diffusion, or some call it the “horizon effect” (the horizon always looks lighter than any other part of a scene because its the farthest away) things that are dark tend to come forward and things that are lighter tend to recede in the eye because we associate a lighter horizon with great distances. If you go and look at concept art for movies or video games you will almost always see a very dark foreground (a character emerging from a cave say, with us looking past him out into the brightly lit world beyond). This is a great way to impart distance. It also creates a comparison – your brain knows that the rock the giant head is made out of is the same color as the rock of the cave mouth we are looking out from, but when you lighten the rock of the head, your brain figures out how far that must be away to be a different color, and compares it to the same colored rock next to us.

[Steve Bentley]

I think what Mike is after is using a background object to hold the original movie that he 3D tracked and have it render behind the C4D generated object so that C4D does the compositing for him.
You can add a background object (create/environment/background) and then make a new material and add a the movie as a texture in the luminance channel (that way you don’t have light the “background” for the movie to show up). As long as the movie and the c4d render size are the same the movie will fit the frame perfectly. Drag that material to the background object.

Depending on your version of C4D and your viewport settings you may or may not see the movie show up or advance frame by frame in the viewport but it will show up when you render.

If the movie in the texture is not lining up to the track or needs to be moved backwards or forwards in time, just click on the name of the file in the luma channel of the material, then select the animation tab in that dialogue. From there you can select what frame of the movie to start on, what frame of the c4D animation timeline that start frame should be attached to and how long the movie should run, and whether you want to assure that its a perfect frame for frame match between movie frame and timeline frame.

[Steve Bentley]

Here’s my guess as to whats happening. In some preference somewhere, there is a radius distance set. I do have additional points along the spline that would be in the way of a rounding on a corner given that “preference” distance, so Ai wont’ give me the radius corner dialogue to change a hard corner. But the rad I’m putting on my points is much smaller than the preference distance and those additional points would never be in the way. So while I get that I can’t make a rad if another point is inside the radius curve distance and it makes sense its not letting me do it, AI will let a path loop back on itself so… give me back control and quit telling me how big my rads should be!

[Steve Bentley]

So that was going along fine (if somewhat tedious, entering the rounding value each time). Then as I picked a new path to work on, the rounding value entry box vanished from control tools bar. Try as I might I can’t get it back. Where did it go?