[zparrish]

I'm running Maxwell version 3.2.1.4. To test this, I setup the Grass modifier directly in Maxwell Studio to eliminate the possibility of corruption by an external plugin.

This may be intentional, but it doesn't appear to be noted anywhere in the docs. I think the initial angle map in the Maxwell Grass extension bases it's orientation on World Space coordinates instead of local normals. It uses the UVW mapping correctly, but I'm trying to use the Grass extension to render stylistic fibers that are part of a curtain fabric. For me to blend the initial angle map with a world space normal map is a bit tricky, especially since I'll have to use another engine to extract the flattened, World Space normal map of the curtain. I suppose it's possible, but it will take a really complicated Photoshop filter that I'll need to run against prerendered assets so it can't be done in a single pass. It will also require me to squeeze the entire texture space into a single tile, making my map resolutions huge! My UVW tile is designed to be a 4K image on a 48.5" X 48.5" space for a curtain that is ~82" tall and in excess of 120" in length. That would be a texture that is ~7,000 pixels X 10,000+ pixels.

This could all be simplified if the initial angle map were based on an object's local normals and UVW mapping. It would also let me use tilable images for my initial angle map.

I've included a quick scene with a cylinder to show this. I've only included 1/16th of the original texture sizes in this example so you won't need to play around with 4K images. There's also 2 pics I took of the fabric with my iPhone, so you can see the fibers I'm referring to. They are a light green color. Thanks Next Limit!

http://csimages.c-sgroup.com/external_f ... le_map.zip

[zparrish]

I'm assuming this is a bug because IF it were intentional, then the color values available for the initial angle map would not clamp the Green channel from 129 to 255. You would have full 3 dimensional range if it were to be based on World Space coordinates.

Also, I did indirectly see a reference in the docs indicating that it's intended to be based on local normals right below the supporting graphic:

*note that for the Green channel the lowest value considered is 128, since the blade can't start growing at a negative angle to the surface normal.

http://support.nextlimit.com/display/maxwelldocs/Grass

Thanks Next Limit!

[Mihai]

Yes, I suppose this is a bug as it should work in the same way when not using an angle map, with the "Grow towards Y" being turned off.

[zparrish]

I just wanted to confirm that this hasn't been patched in MR 3.2.1.5. The initial angle map still ignores local normals. I have a collection of patterns for a client that are actually on hold pending the fix to this bug. It's not like anybody is banging down my door or anything, but it would be really nice to get this fixed soon so I can post their patterns for them. Plus I know you guys have your hands full making changes to core functionality (like all the MXI issues that were just fixed, as well as any R&D you guys are doing on the performance side), so I totally understand. I just want to keep this one in the air. Thanks Next Limit!