|
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
A 100% procedural dandelion.
I've been working on this on and off for a while now, writing some macros to
build simple plants. I'm using a multipurpose mesh circle arc plotter, with a
creased or circular cross-section and a set of functions to control diameter
along the leaf length - the code used for the stalks, leaves and petals of this
plant is the same as for my grass. It's been quite fun playing with functions to
generate leaf shapes, but quite tricky to get them to look realistic. The
function for the dandelion leaf is a monster!
I have no idea how to get a decent translucent-yet-opaque texture for the
leaves, however. Does anyone have any pointers in this regard?
Bill
Post a reply to this message
Attachments:
Download 'dandelion1.jpg' (56 KB)
Preview of image 'dandelion1.jpg'
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Bill Pragnell" <bil### [at] hotmailcom> wrote:
> A 100% procedural dandelion.
>
man, these things really grow EVERYWHERE, even on checkered planes...
Very impressive. I'm looking forward for more plants.
Have a nice day
Karl
--
.... back to lurking mode :-)
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
He he he, my leaf macro now has uv coordinates. This is a big improvement!
Post a reply to this message
Attachments:
Download 'dandelion2.jpg' (56 KB)
Preview of image 'dandelion2.jpg'
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Bill Pragnell" <bil### [at] hotmailcom> wrote:
> He he he, my leaf macro now has uv coordinates. This is a big improvement!
Okay, this is really cool Bill, keep it coming!
The UV coords will make it much more fun at the texturing stage. That said, I've
been playing around with ways to make lighting look more realistic on leaves and
such, but it's given me quite a headache - the old double_illuminate trick just
doesn't cut it for stuff like this.
It's a really complex problem because typically the front side can range from
velvet-diffuse to highly specular, but the back side is usually highly diffuse,
with a lot of subsurface scattering. It really requires a two-sided texture to
be believable, with variable translucency based on the angle of light to the
surface normal. Reflectance when viewed from the front side (convex),
transmissive SSS when viewed from the rear (concave)...
Hey Sam, you have this figured out yet?
-Rob :)
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Robert McGregor" <rob### [at] mcgregorfineartcom> wrote:
>
> It's a really complex problem because typically the front side can range from
> velvet-diffuse to highly specular, but the back side is usually highly diffuse,
> with a lot of subsurface scattering. It really requires a two-sided texture to
> be believable, with variable translucency based on the angle of light to the
> surface normal. Reflectance when viewed from the front side (convex),
> transmissive SSS when viewed from the rear (concave)...
>
> Hey Sam, you have this figured out yet?
>
> -Rob :)
Yes, I was also thinking of Sam Benge's SSS stuff. But wouldn't that require a
3D leaf, rather than a 2D one? (I don't know that answer myself, just
guessing.)
As an alternative, I was thinking that a (complex?) pigment_map of some kind
might look cool--with a partly-translucent cellular appearance, or some such,
with veins and whatnot. (Kind of an amorphous suggestion, eh?) :-)
Nice plant, Bill! I'd *really like* to see a CU of the flower and its green
bulb...
Ken W.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Robert McGregor" <rob### [at] mcgregorfineartcom> wrote:
> The UV coords will make it much more fun at the texturing stage.
I never really thought about it but UV mapping for mesh leaves is actually
rather simple - once I've specified the coords in the macro, any of the leaves
can be given any texture I want, procedural, images, anything. This will go a
long way in the realism stakes.
> That said, I've been playing around with ways to make lighting look
> more realistic on leaves and such, but it's given me quite a headache
> - the old double_illuminate trick just doesn't cut it for stuff like this.
> It's a really complex problem because typically the front side can range from
> velvet-diffuse to highly specular, but the back side is usually highly
> diffuse, with a lot of subsurface scattering.
Agreed. I'm using double_illuminate at the moment but it's a stopgap really.
> It really requires a two-sided texture to
> be believable,
That's not a problem - a single-sided mesh can be given an inside_texture
(although that might scotch the interior for a 3D media-containing leaf).
> with variable translucency based on the angle of light to the
> surface normal. Reflectance when viewed from the front side (convex),
> transmissive SSS when viewed from the rear (concave)...
I'd settle for translucency from both sides, some leaves don't have
significanctly different backsides.
Thanks for the comments!
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Karl Anders" <kar### [at] webde> wrote:
> man, these things really grow EVERYWHERE, even on checkered planes...
....where the cracks are REALLY small!
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Kenneth" <kdw### [at] earthlinknet> wrote:
> Yes, I was also thinking of Sam Benge's SSS stuff. But wouldn't that require a
> 3D leaf, rather than a 2D one? (I don't know that answer myself, just
> guessing.)
I think it would, but that's not necessarily a problem. Actually, I just thought
of a method (with my current macro) to give a 3D leaf underside a separate
surface texture too... hmmm... It's not just the leaves, of course. If you
compare this to a real dandelion, the leaves actually look OK, but there's far
too much shading on the petals. In this case at least, I think they would
benefit from SSS more than the leaves.
I'm sure Sam can correct me, but I thought his recent SSS stuff wasn't actually
object media? I think for the simple case of a leaf, media could well be
sufficient.
> As an alternative, I was thinking that a (complex?) pigment_map of some kind
> might look cool--with a partly-translucent cellular appearance, or some such,
> with veins and whatnot. (Kind of an amorphous suggestion, eh?) :-)
Might be worth thinking about if all else fails.
> Nice plant, Bill! I'd *really like* to see a CU of the flower and its green
> bulb...
Cheers, I'll post one later. Bear in mind that this isn't really intended for
such close scrutiny - I'm hoping to end up with a selection of different preset
plants that I can use to populate larger scenes. The triangle count on this
dandelion isn't actually that high!
Bill
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On Mon, 13 Oct 2008 17:58:50 EDT, "Bill Pragnell" <bil### [at] hotmailcom>
wrote:
Mary Queen of Scots, had her head chopped off!
Nice one Bill all that is missing is the burdock :)
--
Regards
Stephen
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Kenneth" <kdw### [at] earthlinknet> wrote:
> Nice plant, Bill! I'd *really like* to see a CU of the flower and its green
> bulb...
Here it is. As I said, it's not meant to stand up to this sort of closeup!
I've been playing with solid leaves and media... not great results so far. I can
just about get a translucent effect on the unlit side, and from certain angles
it looks quite good, but to be honest the effect is very subtle, and the render
time shoots up. 3D leaves = easy, media settings = hard. As per usual.
The solid leaves are only a relatively simple addition to the macro, so I think
I'll shelve it for now so I can do some different plant types...
B
Post a reply to this message
Attachments:
Download 'dandhead.jpg' (31 KB)
Preview of image 'dandhead.jpg'
|
|
| |
| |
|
|
|
|
| |