View Full Version : Modifying the look of HyperVoxel blending?

07-03-2018, 01:39 AM

Is it possible to better control the look of blending between HV nodes (or the new Volume/Shape/etc Primitives) apart from the Blending Scale value? I'm after a solid looking surface with larger attraction distance but with a thinner "bridge" between metaballs. So less like glue blobs and more like those visualisations of connections between brain cells (but regular, not distorted). Any ideas will be much appreciated. Thanks.


07-04-2018, 12:53 AM
You can change the shape of the HVs or Volumetric primitives, but I don't think there's any way to change the behavior of the Blending mode. Are HVs / volumetrics required in this case, or could you model them geometrically?


07-04-2018, 02:37 AM
The HVs are not required but I'd like to have an option of animating them, similarly to this 2D example (https://codepen.io/znak/details/jKRWxy/).

07-04-2018, 05:19 AM
You could as monroepoteet says use geometry, and use morphs to control the blend/ pulling apart.
If you are making it flat and 2d then there won't be any edges showing between the 2 blobs when they pull/ morph apart.

If you have After Effects have a look at this tutorial: https://www.schoolofmotion.com/tutorials/toon-shaded-after-effects

07-04-2018, 05:29 AM
So the idea with morphs would be to have a regular ball and a morphed ball with a pulled side? OK, that could possibly work. But when animating, is it possible to target two objects to themselves, so they rotate towards each other correctly and most importantly activate morph below certain distance?

07-04-2018, 06:21 AM
This (https://www.youtube.com/watch?v=HPKiiSKTvoo&frags=pl%2Cwn) looks like one solution I need to try.

07-04-2018, 10:56 AM
This (https://www.youtube.com/watch?v=HPKiiSKTvoo&frags=pl%2Cwn) looks like one solution I need to try.

and Part 2: https://www.youtube.com/watch?v=hsxPAUGz4YY

07-04-2018, 01:36 PM
Here is a sample scene demonstrating the shape-changing Volumetrics in LW2018.0.5. It may not be the effect you're looking for, but I think it's pretty cool.


Note that this is a very pair-oriented solution - i.e. it cannot work on the multiple targets as shown in the sample motion-graphics video you posted.

It is a fairly complex setup. The Volumetrics (and classic HVs) are spherical shaped (other than Cube or Cone Volumetrics in LW2018), so the only way to shape them is to change the Texture Amplitude. What I did in this case was three basic things:

Use Motion=>Target Item to target each Volumetric Null to the other
Use Motion=>Modifier=>Nodal Motion to change the Z Scale based upon the distance to the other Null
Create a Node network which shapes the Volumetric Texture Amplitude based upon three Gradients

The Target Item setup means the Nulls "point" at each other down the Z axis. The Nodal Motion node graph calculates the Distance to the other Null and scales up the Null on Z. A nice feature of LW2018 Volumetrics is the base Null can be scaled separately in X, Y and Z, stretching the volumetric shape. Classic HV's are scaled "locked" in X, Y and Z.


The Node Network basically has three Gradients ranging from -1 to +1, one for each of X Coordinate, Y Coordinate and Z Coordinate, which are the coordinates of the Volumetric "cloud" before being stretched by the Nodal Motion. The Gradients each have a relatively complex shape to shape the Volumetric to the "stretched" shape, and only the Z Coordinate Gradient is shown below. These are all added together and fed into the Volumetric Amplitude. A LOT of trial and error is required to get the shape I wanted - the Gradients all interact with each other and the value ranges are not intuitive, IMO.

The spider web of node setups using the second set of Gradients is to specify a plain constant for the Texture Amplitude when the Z Scaling is "low" (i.e. around 1.0). This causes the Volumetrics to be spherical until they actually start stretching due to distance, and then the Gradient-based Amplitude takes precedence.


Good luck!

P.S. For another example of creating a complex shape for the volumetric using X, Y and Z gradients, see the cumulonimbus cloud entry posted here: