SD5 - FX Map - Pattern scatter by Mask Input?

polycounter lvl 2
Offline / Send Message
Kid.in.the.Dark polycounter lvl 2
Hey there,

I'm currently trying to build a grass generator using an FX-Map ... I'm curious to know how I go about building a mask input node where I can plug in a pattern of some sort which will mask out the areas which my grass will be placed? Exactly like how the Tile Sampler Node's mask function works.

I've had a look through the Tile Sampler Node's FX Map, etc already and it's just really complex and I can't seem to find where some of these variables are placed (mask_map_sampling_technique to be exact) and I've tried to reverse engineer it but I've pretty much hit the end of my experimentation distance with this and I've just hit a brick wall with it now.

I guess what I'd want is the most cleanest/least long winded approach to this... from what I've gathered from looking at the Tile Sampler Node is that it's an FX Graph compiled of input nodes which pipes out into an Output node with exposed parameters which becomes a graph object you drag and drop into your working substance graph.

I could just build this Substance using a Tile Sampler node from the get-go but I'd like to learn more about FX Maps as I'm getting more and more deep into SD now :)

Any help is much appreciated, Thanks!

Replies

  • kurt_hectic
    Offline / Send Message
    kurt_hectic polycounter lvl 6
    Hello,

    You may seen my topic on allegortmic forum, because I tried to do exactly same thing. I spent on this a few days; me thoughts:

    -it's almost impossible to do it- have you seen a good looking grass made in SD? Maybe one or two, right?
    I wish SD could operate on 3d models...without it it's not possible to create complex layering, avoid intersections, create nice AO of the whole thing

    -I was able to create some kind of pipeline but it was slow and very imperfect

    -rather than messing with SD I recommend to check:
    http://www.carbonscatter.com/  plus this: http://www.evermotion.org/shop/show_product/archmodels-vol-124/9371

    Imagine, 3d, high poly, realistic grass, scattered via masks with basic logic. It seems it's all we need....but I was so exhausted after SD that I didn't try it.

    And yeah, 3d scans appeared. Probably it's not worth the effort right now, don't you thing?

    https://megascans.se/assets/pjwbe0

  • m4dcow
    Offline / Send Message
    m4dcow polycounter lvl 7
    Hey there,

    I'm currently trying to build a grass generator using an FX-Map ... I'm curious to know how I go about building a mask input node where I can plug in a pattern of some sort which will mask out the areas which my grass will be placed? Exactly like how the Tile Sampler Node's mask function works.

    I've had a look through the Tile Sampler Node's FX Map, etc already and it's just really complex and I can't seem to find where some of these variables are placed (mask_map_sampling_technique to be exact) and I've tried to reverse engineer it but I've pretty much hit the end of my experimentation distance with this and I've just hit a brick wall with it now.

    I guess what I'd want is the most cleanest/least long winded approach to this... from what I've gathered from looking at the Tile Sampler Node is that it's an FX Graph compiled of input nodes which pipes out into an Output node with exposed parameters which becomes a graph object you drag and drop into your working substance graph.

    I could just build this Substance using a Tile Sampler node from the get-go but I'd like to learn more about FX Maps as I'm getting more and more deep into SD now :)

    Any help is much appreciated, Thanks!
    The way I tend to do it, is to sample the mask texture at the position where a blade is generated, and then use that to modulate the opacity or size of the blade. I prefer changing the size, because then I get a small blade of grass instead of some intermediate opacity value grass blade.

    So basically you generate a whole bunch of blades, but the ones in the dark parts of the mask are still technically generated but they aren't actually visible due to opacity or pattern size.

    This is in the pattern size parameter of one of my graphs

    Looking at the forest ground graph on substance share, specifically the grass fx map helped alot (I think besides a few different values my graph is identical).
    https://share.allegorithmic.com/libraries/702

    FX maps are really powerful if you know what you're doing and worth delving into for all sorts of other substance effects.

Sign In or Register to comment.