New Grass material in LumenRT update 14!

Hi Ron,

I looked at your grass material, and here is what I have found:
Your density map contains normalized values 0.62 (158/255) for brighter stripes and 0.4 (103/255) for darker stripes.
I recommend scaling the values so that max is 1.0 (or 255).
So the map would look like the one that I have attached (VERG_grey_stripes2.png).
Then, you can use the slider to scale down the entire density (eg. set slider to 60%).

Indeed, low values in the density map (values below 0.5) will cause the blades to be smaller and a bit darker.
This is to simulate transitions between high-density patches (eg. tufts) and areas with no grass.
This yields realistic results for density maps representing "natural" (random) distribution.
But for density maps containing patterns with "hard" transitions like your map, it can produce unintended effects.
So, for this kind of density map I suggest using values above 0.5.

Moreover, there is a bug in the algorithm that reduces the height of blades when the density map value is between 0.25 and 0.5 (like in the darker stripes of your map).
This makes the blades even darker and smaller than what they should be.
So this amplifies the issue in your material.

Hope this helps.

Regards,
Alex

Hi Ron,

I looked at your grass material, and here is what I have found:
Your density map contains normalized values 0.62 (158/255) for brighter stripes and 0.4 (103/255) for darker stripes.
I recommend scaling the values so that max is 1.0 (or 255).
So the map would look like the one that I have attached (VERG_grey_stripes2.png).
Then, you can use the slider to scale down the entire density (eg. set slider to 60%).

Indeed, low values in the density map (values below 0.5) will cause the blades to be smaller and a bit darker.
This is to simulate transitions between high-density patches (eg. tufts) and areas with no grass.
This yields realistic results for density maps representing "natural" (random) distribution.
But for density maps containing patterns with "hard" transitions like your map, it can produce unintended effects.
So, for this kind of density map I suggest using values above 0.5.

Moreover, there is a bug in the algorithm that reduces the height of blades when the density map value is between 0.25 and 0.5 (like in the darker stripes of your map).
This makes the blades even darker and smaller than what they should be.
So this amplifies the issue in your material.

Hope this helps.

Regards,
Alex

Hi Alex,

Thanks for examining at my material. This was just a test to investigate how the new grass materials work, and not necessarily a realistic example.

I don't really see why the density map should affect the height and color when there are height maps and color maps to handle each of those grass characteristics. Each grass characteristic map should be independent unless the user enables an override to allow one characteristic to affect another. Otherwise the maps will potentially be fighting/conflicting with each other. Maybe you should have characteristic check boxes that allows multiple characteristics to be controlled by a map. 

The way the grass material currently works. there is no way to have tall, low density grass because the density map will also shorten the grass.

Regards,
Ron

Hi Ron,

The idea behind this automatic height modulation depending on density map, was to give the user an easy way of getting convincing, "natural" results by just setting a density map, without having to set a height map that matches the density map.
But this behavior should be optional indeed.
I have added an enhancement request to our backlog for that.

Regards,
Alex

Glad to hear this. It's always nice to have more options.