Steve Bentley
1
After Effects Coin
3
COW Coins
2
Maxon C4D Coins
Forum Replies Created
-
When you did metaballs did you start with geometry or just three, 1 point polygons. that worked perfect for me.
-
Start with an L shape but that has dimension – so a sweep with an L spline and a square profile – then kife out three spots where your spheres would be. (cut “cubes” are standing in for spheres right now). Then cut again halfway between “cubes”. Then scale those spots to get an hourglass shape (but still blocky – think minecraft hourglass). Then apply a subd surface and tweak end cube shapes until you are happy.
-
Ok that’s not quite accurate – the stripes can be done with the gradient tool or the tile shader or the checkerboard shader.
-
From the looks of things that’s probably modelled. Each color seems to have a rad on it and then a deep chasm between it and the next object – even the diagonal half rectangles. And that would keep it all in C4d to animate locally. You could do it with textures with a deformer channel but I think in this case it’s all geometry with selections colored by different materals.
-
If you make a rounded hourglass profile and lathe it and then duplicate that, rotate it, align it, and then do some cutting and joining that should work perfectly.
-
Ok. Its probably the high rate of change of the bright flashes. If you increase the number of keyframes per second that will help. A big change can’t be captured properly in between keyframes. But be aware more keyframes means higher mb.
-
Do the deformers need to be active? or have you just used them to get this shape? You could do a current state to object to “bake” the deformers in, bool them together and then wrap everything into a hypernurbs object (sorry subdivision surface – that’s my age showing) and cut out the cusp as its called and bridge that area with something simple and let the subdivision smooth it out.
If the deformers need to stay intact you could try changing the phong angle in the phong tag – while not a perfect solution the cusp might not show. Adding a bool union to this trick might get rid of it too but you will loose the deformers again.
Or you could start again and bool two capsules and then apply two deformers
-
So a couple of thoughts. Pictures can be deceiving but that looks like a pretty close audience to a pretty big screen. That means that your POV spot is going to be very very small – like two seats. On the plus side the screen geometetry looks simple and clean, which will help expand that spot.
I suppose I’m a POV snob because that’s the only place the illusion truly works, but given how people rave about projection mapping even when they have seen it upside down and the perspective is inside out perhaps I’m the only one who cares. -
As long as you are rendering in an uncompressed format (QT animation best, tiff pshop etc) then the kind of file you render to doesn’t matter. The image will be what it is.
Going from 250mb down to 10 or less you will loose something but if you are only doing grays then you are ahead of the game. Is the compress for the web or are you just trying to reduce the size and the you will work with it as an element? Different output formats from a compressor can make a difference in quality vs mb. But as long as the input movie is clean then the format you render to doesn’t matter.
Also don’t judge the movie by a still – if you judged netflix by a still you would never sign up. The motion of all the compression “noise” often blends together and you don’t really see it. After all few notice film grain and it can be horrible.