Woa, how did you do it, i’m beginning to be taught this however i do not know the place to start.
Do you thoughts displaying the way you made that one? 😮 I at all times questioned how Plexus is made
In a base materials (on this case it simply has a base shade, smoothness and normals of the bricks), make a wetness masks much like how you’d do dissolve (Smoothstep and so on…)
Offset the wetness masks by the peak of the bottom materials.
Then use the consequence to extend the smoothness and reduce the bottom shade.
There are tons of the way to make this type of impact, however they usually contain inserting meshes/ribbons between a particle and it’s subsequent closest neighbor.
It may be pretty costly to iterate over each particle to discovered these closest neighbors, so I desire setting some max size and eradicating connections which are too lengthy. Not the identical however usually shut sufficient.
As for the connection itself, you’ll be able to spawn meshes, and set their location to one thing like
Point1 + (Point2-Point1)/2 (midway between point1 and point2)
then set the orientation to normalize (Point2-Point1) and scale it by size (Point2-Point1)
Alternatively you should use ribbons and spawn twice as many, then set RibbonID to UniqueID/3 to get pairs of two, then choose some random particles. Notice this makes it much more tough to test the size.
Additionally there may be the sprite final analysis module, which can be utilized to set 2 factors to span a sprite between (would actually advocate this one, it’s very easy to make use of)
1 Like
Thanks lots! That’s truly actually useful 😮 will do some experiments with that later
Experiment, not tremendous completely satisfied, however needed to indicate it so it feels completed and I can transfer on to different issues.
6 Likes
I believe it’s cool, thanks for sharing 
Did some checks utilizing the brand new decal renderer
5 Likes
Did some checks utilizing Niagara baker to bake out outdated bursts with giant numbers of gpu particles.
4 Likes
I’ve a query right here, between baking them out right into a flipbook and simply spawn like 1000’s of mini particles, which is extra optimized right here? Afaik this flipbook sprite will take over more room on the display.
it’s non comparable.
making an attempt to ask if 1(overdraw) is dearer then 1vertex or 1byte of texture reminiscence; these are completely different bottlenecks within the rendering pipeline
it solely takes 1 a part of the pipeline to sluggish the complete recreation, at that time that space of the rendering throughout usually many belongings must be re-factored to make use of a unique course of
if overdraw was the difficulty, (or a recognized to be a difficulty) usually many belongings must be refactored to make use of much less overdraw and begin leaning extra on GPU,CPU processing,vertex data or node rely and so on… all only for the sake of retaining overdraw from reaching bottleneck.
is that extra clear?
with out context there isn’t any solution to evaluate an overdraw to vertex place replace or to texture reminiscence
