Anim8or Community

The movement of the mannequins in 'Escher' is particularly good.  The figures appear to have real 'weight', like it's an actual effort for them to climb the stairs, and the climbing and descending 'gaits' are perfect.  The transitions from 'climbing' to 'walking' to 'descending' could be smoother but sequence transitions are never easy.  Very impressive work!

Since holograms that are animated and aren’t viewed through a photographic plate are still the stuff of science fiction there is no ‘canonical’ way they should look.  It’s pseudo-science so you are free to make it look however you like, ie. there’s no ‘wrong’ way to do it.

A hologram is an effect of light and has no solidity (OK, Star Trek ones do but whatever) so the holographic object and any projection cone you include probably shouldn’t receive or cast shadows.  Also, I’m not sure whether it’s possible in Anim8or but ideally they shouldn’t be affected by lights in the scene other than one positioned near the projector illuminating the holographic object directly.

Including a greyscale white-to-black gradient image mapped in the transparency field in the ‘projection cone’ object will make it appear to fade away at the top.

This quick test incorporates Kevin Gales’ concept and johnar’s flickering idea.  There is a green light illuminating the hologram from below, perhaps a script in the light colour/brightness channel that incorporates randomisation would look interesting, adding a bit of flicker.  I have minimised the effect of scene lights by reducing the specular amount in the textures used for the holographic elements to zero and increased the emissive value to about 0.5 with a green colour to make the holographic elements appear to glow.  I haven’t done it but including a soft green ‘bulb’ light in the centre of the holographic object that casts a bit of a green glow on surrounding objects would be a nice touch too.

So, johnar reminded me that it had been a long time since I had posted anything here (almost 2.5 years!).  I know this thread is old but I figure it's acceptable, etiquette-wise, to resurrect one's own vintage threads.

The Mission Backup Earth project (https://missionbackupearth.com/) is still keeping me busy, as an ongoing web series with CGI in every scene (yes... every, single, scene!) there is an awful lot to keep me occupied.  It's interesting to look back on the early work done in this project to see how things have improved, particularly with CGI/live action integration.  It is common now for renders to align perfectly with the live action, "plug and play" as Alex often says.  Considering that there can often be a year or more between filming the live action and the background plates being modeled and rendered this is quite an achievement.

Below is a sequence I created a few weeks ago for an upcoming episode.  All modeling was done with Anim8or (including Arik's ubiquitous triangular cargo containers) but rendered in Carrara.  Anim8or's renderer is much improved speed-wise and somehow its renders are just more beautiful than Carrara's, but Carrara still has a speed advantage with network rendering (3 x i7 quaddies supplying 24 rendering threads, and this scene still took several hours to render).

Hopefully the Youtube link works, if not please bear with me while I work it out again!

Love your work johnar, and great to see I'm not the only one who regularly falls back on good old pen and paper!  I admit that I am yet to tackle lip syncing, it's one of those "one of these days" things, but I know who to hit up for advice when I do

Before subdividing it is important that all faces have exactly 4 sides.  Triangular or 5+ sided faces will not subdivide cleanly.  End caps on objects extruded in Anim8or are usually comprised of triangles, spending some time removing superfluous lines to ensure all faces are 4-sided will result in smoother subdivisions.

Do you mean setting the morph target in Object mode, or accessing the morph channel in Scene mode?

Love your work Arik, I always loved the original series designs

Welcome to Anim8or!

There's two answers to this question, depending on what you need and how you plan to use the mesh.  The first involves ‘smoothing’ the mesh, and the second involves ‘subdividing’ the mesh.

Smoothing:

Smoothing divides each face of a mesh into four faces (assuming your mesh’s faces are all 4-sided) by bisecting each line with a vertex point, placing a new point in the middle of each face, and joining up the new points with lines.  Importantly, this function changes the original mesh which will now have 4 times as many faces and will subsequently be bigger, from a data point of view, and potentially more difficult to animate later.  This function includes a numeric setting; if its value is 1 each face will simply be divided into 4 with each new face aligned exactly to the original face.  If its setting is 0 the new vertex points will be positioned based on (presumably) Bezier curves, resulting in an overall smoothing of the object.  Settings above 1 can give some interesting results.

Try it out:

• Create a sphere object (under Shapes in the left side menu)
• Convert the sphere to a mesh (Build > Convert to Mesh)
• Smooth the sphere (Build > Smooth Object...).  Enter 0 for an overall smoothing effect.

Subdividing:

Subdividing is similar to smoothing, at least in appearance, however there is a critical difference.  While smoothing changes the original mesh by adding points, lines and faces, subdividing doesn’t change the original mesh.  Instead, it uses the original basic mesh to mathematically define a much finer mesh.  When rendered, the original mesh is ignored, and the calculated fine mesh is rendered instead.  The advantages of subdividing over smoothing are that the overall file size is not increased and, more importantly, the mesh is far simpler to manipulate and animate.

Imagine animating a mesh of a person's face to show expressions.  If the actual mesh was very fine (eg. smoothed) you would need to individually animate a large number of vertex points which would be difficult, painstaking and may ultimately look unrealistic.  With subdivision, your mesh can remain much coarser, and therefore be far simpler to animate, while your render remains smooth and organic in appearance.

Try it out:

• Create a sphere, the same as for smoothing
• Convert to subdivided (Build > Convert to Subdivided)
• Switch to point edit mode (the icon with three dots in the ‘Mode’ section of the left side menu) to see the original course mesh that is controlling the fine subdivision.  Try moving some vertex points and switching back to object mode (the arrow icon under ‘Mode’) to see how the fine mesh is affected by the original mesh.

When defining a mesh that will ultimately be subdivided it is important to ensure that all faces have 4 sides.  3-sided faces and faces with more than 4 sides will subdivide strangely and may look odd in the final render.  This is a fundamental aspect of 'organic modelling'.