[Steve Bentley]

Have you tried the “stick texture” tag from the Cinema 4d Tags list? (right click on object).
Put material how you want it then click “record” in the stick texture tag.
You can also try changing your mapping. That looks like Spatial. Try anything else.

[Steve Bentley]

Ya that’s pretty easy. You can have a soft edged (or blurred) alpha in the alpha channel of a PSD or you can just use an image of your choosing. It needs to be soft in the actual image so you will need to do that in pshop (the alpha should be soft not the image). The image part should extend past the blurry edge.
This psd with an alpha or just a gray scale image that is going to be the alpha needs to be put in the “alpha” channel of the material. If the alpha is part of a pshop file you can go with the default settings, (in this case the same psd would be in the color channel and the alpha channel of the material and c4d sorts out what part of the image it needs) but if the image you want as the alpha is just a separate image you sometimes need to uncheck “image alpha” in the alpha settings so that C4D doesn’t think its a psd and go looking for an embedded alpha.
Make sure to double click the material in the material manager so that it brings up the material editor. Not all the settings show well if you just the use attributes dialogue that comes up when you single click on the material in the material manager.
That “soft” button is more for aliasing issues rather than making a hard edged alpha blurry. You will need to make the alpha blurry in an imaged editing package like Pshop.
Make sure to put the material with the alpha to the right in the list of tags and materials on the actual object in the object manager: things to the right overlap things to the left. You are basically shining the second material through the window that is the alpha, on top of any materials to the left – white is clear and black is opaque in the alpha “window”.
And don’t use the ” mix textures button” in the tag’s attributes. That works for things that can’t really be alpha’d and will mix, say, a bump from one material with the bump of another.

[Steve Bentley]

See also Method 2 below which might be better depending on why you need to render the explosion separately.

So the initial problem is that once you get into 2D and you put that mask of the 3d statue over the explosion (itself now just 2d pixels) there is no way to reveal stuff that is “over top of” the statue in 3D space because that depth info is lost. In 2D, the explosion is behind and below the statue as far as AE (or whatever comper you are using) is concerned.
So the idea with a depth matte is that you can split the explosion element into two parts in the 2D comping stage. The part that is behind the statue and the part in front. But you can’t do that with normal masks because that split point is in depth or Z-space and the comper only knows about x,y space. So instead you can render a depth pass which is in its simplest terms an alpha of the explosion but it has a gradient that travels from the camera to some set distant point as a sort of third axis. There are no object “edges” in this depth axis, just a gradient like a foggy cloud (actually thinking about things emerging from the fog is a great analogy – the fog is a white gradient getting thicker further away from you. When you see something emerging it just has less of that white gradient between the object and you.) By creating a clipping or cut off on that gradient you can cut off all the bits of the object that are either beyond that point or in front of that point. The depth gradient is done at super high bits depths (hence the EXR 32 bit format) otherwise there wouldn’t be enough steps to the gradient to get clean edges all the way along the length of the gradient (or should I say depth of gradient). To be clear, the alpha of the explosion has x,y cooridinates but the depth information goes in and out (sometimes called a Zpass).
So in the end you put one copy of your explosion above the statue in your 2D comp and one copy behind. Then use the 3D depth matte tools in your comper (in AE its in the 3D effects menu) to cut off the Z mask where the statue is in both copies of the explosion, but you just invert the cut off (or the gradient) for one of the copies so that one copy reveals whats behind and the other reveals whats in front with the statue sandwiched between the two.

Method 2 – no depth pass required.
The other way to do this is to render the explosion in 3D with the statue visible, but put a composting tag on the statue. Turn off the receive shadows in the tag and turn on the bottom “matte object” button. Render away. The statue will be self matting in the explosion volume. It will show up as a black silhouette shielding any particles that fall behind it but letting particles in front show up. It will also cast shadows on the explosion (if you want) because its really there in the scene blocking light , but since the statue will have no color information we turn off the receive shadows since there’s no point having those calculated. Make sure “seen by camera” is on in the tag; a lot of people turn this off since they think they don’t want to see the object, but the matte object button looks after this. Oh and you have to render with alpha – since the statue will only really exist there. This can be rendered in any 8bit format that supports alpha – no EXR required.

[Steve Bentley]

I got much better results with a “cylinderYaxis” shape rather than either convex hull or the moving mesh. While the moving mesh is always more accurate, it does always seem to be “stickier”. I also dropped the friction down to 2% and upped the bounce to 25%. For me those simple changes yielded a realistic cellulose pill drop on a countertop type surface. At least from what I remember the last time that damn childproof cap exploded off the bottle.

If you do use the “cylinder”, rotate the transform Pitch in the cloner a little – with no rotation some of the cylinders will land end-on and just stand up. “Box” works too (and doesn’t suffer from the stand up problem due to pointy corners) but the pills won’t roll when they settle. If there is no label on the pills (to giveaway the rotation) then the “box” is the better choice. Plus they simulate so much faster than moving mesh.

You can also add a Friction force as an object from the Thinking Particles menu inside simulation. Its used like an effector but you don’t put it in the effectors field in mograph, it will just work on all the pills. Use a sphere for the Friction object’s fall off then use the inverse button. (see pic below) The yellow area is the area not affecting the pills and the red is the area that will slow them to a stop. Spread out the two areas but make sure the yellow area rises up to encompass where the pills come from otherwise they won’t get to where they need to go. (personally I’d drop the pills from a much lower height – these Sims usually have too much energy in them anyways).
If you are using R20 and above you can use Fields to do this.

[Steve Bentley]

I’m not sure where the statue is – is it part of the C4D scene or is it part of the plate? Is that black silhouette a match moved object in 3D or just the 2d matte from the plate?

But you could render out a multi pass (we use EXR’s) and include a depth pass. I assume you want some of the explosion to be behind the statue and some to be infront/overtop of the statue, yes? The depth matte will allow you to comp the explosion up to a certain depth behind the statue and then overlay the explosion from that depth forward in front.
Does that help?

Here’s a tutorial that not only shows you how to do it but also corrects a long standing problem with the depth pass in C4D.
https://www.youtube.com/watch?v=hej9gSpXKqQ

Some contents or functionalities here are not available due to your cookie preferences!

This happens because the functionality/content marked as “Google Youtube” uses cookies that you choosed to keep disabled. In order to view this content or use this functionality, please enable cookies: click here to open your cookie preferences.

[Steve Bentley]

Is it just that the subD object is set for a higher poly count in the render vs the viewport?
Whoops got that backwards. I mean its set for lower in the render. (swap those numbers in the attached image)

[Steve Bentley]

Whats the latest word on the SSDs for 3D rendering? SSds used to have a limited life for read and writes and 3D rendering does a huge number of reads and writes for every frame as it gathers textures etc from the drive.
It depends on your scene complexity of course but we had file that we clocked doing a million (literally) reads and writes to the drive per frame. If the SSDs still have this gotcha I would think its not going to last too long. SSD for OS is perfect, but for your 3D assets, an HDD drive might still be the way to go.

[Steve Bentley]

The dynamics engine can do this quite easily.

You will have to make all the text “editable” (as in make it polys and not leave it as extruded text – make sure to set the text to “Create Single Object” in the Extrude object so the caps get glued to the extrusion. – the collision engine can only work on polygonal objects)

Turn off gravity (project settings). This way the text won’t fall when you start the simulation.

Then you need objects to bang into the text. These can be hand animated or they can be unleashed in a particle system or even have the wind push them along.

You can have different wind objects – one for the collider objects and one for the text and you can ramp that up and down as needed so the wind doesn’t affect the text until after they have been disturbed by the colliding objects.

All the geometry that needs to collide will need Collider Body tags (Right click/simulation tags).
Depending on complexity of text and collider objects you may need to make stunt objects. These are lower poly versions of each object that are not seen by the camera (rightclick/Cinema4DTags/composting tag/uncheck “seen by camera”) and do all the banging around and carry the hero objects around with them – the stunt objects get the collider tags not the hero objects. There’s a lot of math going on in the collision detection algorithm so the lower the poly count for the smashing calcs the faster it will calculate.

Once you are happy with the simulation you can cache it so as you tweak textures or other elements it doesn’t have to do the heavy math every time you make a change and you can play the animation real time (or as real time as your system will allow)

[Steve Bentley]

Well so much for that. Using the exclusion tab does seem to work some times but as in this example not always.
This is in R19 (its a legacy project so it has to stay in this rev for now). We’re using GI with the standard renderer since the other methods have too high a noise to render time ratio (and we’re not liking the after market denoisers for this).
This of course is a simplified scene but all the main players are here.

There are a bunch of GI emitting planes and disks, but at the bottom of the objects list is the “odd man out” emitter. Currently it is set to only affect the small primitive grey sphere in its composting tag exclusion tab. But if you turn the oddManOut emitter off, you can see it is affecting the skin colored sphere. Is the “exclude” option the only one that works in the tag? “Include” doesn’t seem to isolate that emitter to only the objects included.

I’ve even put all the emitters in all the other emitter’s compositing tag exclusion tags thinking that they might be lighting each other up and then acting as bounce cards.
But if you turn off the emitter called “eyelight” (which has the same exclusion parameters as OddManOut) it doesn’t affect the skin colored object. Inconsistent!

What I’m after here is to add some emitters and reflectors to add little kisses on a much larger and more complex model, but I want to only affect certain objects. But some of the emitters, even with composting tag exclusion sets, are affecting all the objects and not just the one I’m trying to target.

Am I not thinking about this correctly?

13950_balltest.c4d.zip

[Steve Bentley]

Hmmm. Not sure how to think about this. That did work, but only for the first bounce.

Regarding an “odd man out” reflection object (the object that should only be reflected in one object instead of all objects): Any object that is included in the exceptions list to receive the reflection will also lend the odd man out reflection to other objects even though other objects are not included in the in the exceptions list. So the other objects can’t see the reflection of the odd man out object directly, but do see it indirectly through the reflection of the object that can see the odd man out reflection.

As well, if you use the pre R16 method of materials (without everything jammed into the Reflectance channel, but using a Beckman et al in the Reflectance channel) and map those materials onto the receiving objects, anything that receives luminance from the odd man out object (whether that receiver is reflective or not), that light will also be reflected in objects that have not been included in the exceptions list, and this will happen on the first bounce as well as subsequent ones.

I guess I figured the traced ray would be tagged with objects it could intersect with – otherwise the system is burning tracing runs when they aren’t needed. Why calculate a path to an object the light isn’t supposed to be reaching?