Grassy area - scatter test

[percydaman]

nice work. grass blades look a little on the beefy side though.

[beestee]

nice work. grass blades look a little on the beefy side though.

I agree on the closeup, but it looks great at a little distance.

Duke2, your latest is very photographic!

Now I will never be satisfied with my displacement grass...this subject makes a very strong case for VRayScatter!

[PIXELBOX_SRO]

shame we dont have in Max something like Paint Effects in Maya ((

Have a look here:
http://www.evermotion.org/vbulletin/...ad.php?t=73894

All the foliage was doen with Paint Effects

[duke2]

Now I will never be satisfied with my displacement grass...this subject makes a very strong case for VRayScatter!

The problem with this method still remains as it did before though - it relies on patches rather than blades of grass, so you're going to have difficulty where it interfaces with a path, etc.

[simmsimaging]

The problem with this method still remains as it did before though - it relies on patches rather than blades of grass, so you're going to have difficulty where it interfaces with a path, etc.

That depends. I'm still working on the parts, but here is the current state of my scatter version: It's made with 15 single grass blade variants, scattered about 125K times each. We're making some more weeds, and odds and end bits to create more variation, but it's working fine and would follow a path no problem IMO.

(rendered in about 1:30hr at 2500X1200 approx. BF + LC, VrayCamera depth of field.)

[skogskalle]

That looks amazing simmsimaging!

But what if the grass-area was a little bit bigger and you needed lets say 2 million copys of each type of straw? Would vrayscatter be able to handle that? Or is there an uppe limit for how many proxycopies it can create?

[Karba]

limit is about 10-20 millions

[jacksc02]

very nice Brett.
I am going to be very critical here.
I think there needs to be some SSS variation per blade. A blade of grass is usually darker(less transparent) in the centre.
I think if you added that it would be perfect!

[rivoli]

well Brett, it's really getting out of hand now.. that really look amazing.
even though I share the same doubts as others about the actual usefulness of this method when doing something more complicated than a lawn, I must say you're achieving impressive results.

[peterguthrie]

looks really really good Brett!

[simmsimaging]

Thanks all

skogskalle: you got the answer straight from the man, but my opinion is that cases where you need that many instances could probably equally be done with displacement or even a 2Dmap, or a mix. For me, we have remade the grass blades entirely and made a low poly and hi-poly version. I think we could easily cover most of our needs for larger areas with the lo-poly without trouble. Anyway, there would be workarounds.

Chris: agreed about the materials. I'm not really liking the results with the 2sided material so I may go to an actual SSS mat and try that. I think this grass also suffers for having no thickness, so I may shell the geometry too. It's fine as is for more distant stuff though.

Rivoli: thanks man, but what would be the type of thing that would be covered in grass and not be a lawn? Just trying to figure out what this method would not work for?

b

[rivoli]

Rivoli: thanks man, but what would be the type of thing that would be covered in grass and not be a lawn? Just trying to figure out what this method would not work for?

I too am thinking about big areas, paths and uneven terrains. I've recently been trying to use scattered geometry on a steep terrain, a kind of slope with a bunch of houses facing a lake, and eventually I had to give up. the patches of grass wouldn't conform to the terrain, and on a big scale displacement seemed to work better. at the end of the day handling the scene was easier, and the look of the grass was ok. but of course that might be just me being a punter and really capable of managing the scene. I guess in the right hands this method it's just ace.

[simmsimaging]

I too am thinking about big areas, paths and uneven terrains. I've recently been trying to use scattered geometry on a steep terrain, a kind of slope with a bunch of houses facing a lake, and eventually I had to give up. the patches of grass wouldn't conform to the terrain, and on a big scale displacement seemed to work better. at the end of the day handling the scene was easier, and the look of the grass was ok. but of course that might be just me being a punter and really capable of managing the scene. I guess in the right hands this method it's just ace.

I think this could work to follow paths, but not sure till I really try. I've run out of play time though, but when I get a chance I will give it a go.


b

[Neilg]

I think this grass also suffers for having no thickness, so I may shell the geometry too.

This is insane, listen to yourself man!


That being said, I cant wait to see the results (and rendertimes)

[simmsimaging]

Actually, just talked about it with my modeler and he did build these with thickness, which explains why the 2sided mat is not looking very good Kinda dumb thing to miss, but it is hard to tell when I'm just loading VrayScatter proxies - they just show up as cubes.

I will to re-render this geometry as is but with a different mat and see how it looks, but now I'm going to be busy on another job so I probably won't get to it for a couple of weeks.

I will post where we net out though.

b