 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
As you all know I speend alot of time creating landscapes. Trying to speed
up grass has been one of my major preoccupations. After using an inc file
which was a union of triangles and getting fed up with the time it took to
parse, I changed the union to mesh. To my astonishment the file parsed in
less than a second. At this point I took a look at my grass file to see if
I could speed it up - lots of work. Coincidentaly Gilles put up his new web
site and whilst enjoying this I came across his grass file and Gilles had
got there first. His file outputs a grass patch as a mesh and to give you
some idea of of it's capabilities and to pass on the lesson I have learned I
rendered this picture.
Here's some stats:
Peak Memory 1323748 bytes
Parse Time 33 secs
trace time 4 min 15 secs
Number of Grass blades ............................4,297,500 !!!!
Wow
Mick
Post a reply to this message
Attachments:
Download 'grass.jpg' (151 KB)
Preview of image 'grass.jpg'
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Mick Hazelgrove" <mic### [at] mhazelgrove fsnetcouk> wrote in message
news:395df76f@news.povray.org...
|
| Here's some stats:
| Peak Memory 1323748 bytes
| Parse Time 33 secs
| trace time 4 min 15 secs
|
| Number of Grass blades ............................4,297,500 !!!!
All very well and good, but how about some grass blades protruding up into
that large white-space ;-)
Oh, ditto that "wow" too.
Bob
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Mick Hazelgrove wrote:
> His file outputs a grass patch as a mesh and to give you
> some idea of of it's capabilities and to pass on the lesson I have learned I
> rendered this picture.
Actually the more blades you put in a patch the least it's going to cost
memorywise. I've covered whole prairies this way. The patch method has a big
drawback, though : it only works well on flat ground. On round objects, the
borders of the patches can be seen.
Another difficulty is to make it work with the eval_pigment trick (which gives a
more realistic grass "movement"). In my prairie macro, the blades are quite
randomly curved and rotated, so that when you "sew" several patches, the seams
are quite invisible. If the curve, size and rotation are determined by
eval_pigment, it's likely that the seams will show up because of the lack of
continuity between patches. This won't be less of a problem if you work with
very large patches, though.
G.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
One way round the edges problem, which I havn't suffered yet, is to create
spiral
patches with a lot of randomness in the placing. BTW I've just rendered some
grass with 36 million blades of grass in less than 10 mins!
Mick
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"I am going to let go of the wall..." -- Buzz Lightyear
Do you have any idea the amount of time I've dedicated to improving Samuel's
grass macro? I was just about to release it. You just managed to flush it
straight down the toilet. I'm going to go cry...
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Sorry :(
"Tony[B]" <ben### [at] panamac-comnet> wrote in message
news:395e12e6@news.povray.org...
> "I am going to let go of the wall..." -- Buzz Lightyear
>
>
> Do you have any idea the amount of time I've dedicated to improving
Samuel's
> grass macro? I was just about to release it. You just managed to flush it
> straight down the toilet. I'm going to go cry...
>
>
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> Sorry :(
It's OK. I feel strange. I'm very sad that all my work was in vain, yet I'm
happy that someone found a better way. Know what I mean? Perhaps I can add
some of the features I've worked on to improve this one. Do you think I
should post the modified country_grass anyway?
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Thanks Gilles, it's his code, his idea, we should have all noticed earlier!
Yes, post the code cause I'm not sure this will work with the eval_pigment
method very well.
Mick
"Tony[B]" <ben### [at] panamac-comnet> wrote in message
news:395e1933@news.povray.org...
> > Sorry :(
>
> It's OK. I feel strange. I'm very sad that all my work was in vain, yet
I'm
> happy that someone found a better way. Know what I mean? Perhaps I can add
> some of the features I've worked on to improve this one. Do you think I
> should post the modified country_grass anyway?
>
>
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Thank you, Gilles. :) Is this a newer version of your grass, or is it the
same old one (pre-trace())?
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Your work is not in vain, you've probably learned some new things.
"Tony[B]" wrote:
> > Sorry :(
>
> It's OK. I feel strange. I'm very sad that all my work was in vain, yet I'm
> happy that someone found a better way. Know what I mean? Perhaps I can add
> some of the features I've worked on to improve this one. Do you think I
> should post the modified country_grass anyway?
--
Samuel Benge
E-Mail: STB### [at] aolcom
Visit my isosurface tutorial at http://members.aol.com/stbenge
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
