POV-Ray : Newsgroups : povray.binaries.images : Comparing mesh grass with faked grass (32+20 kbu) Server Time
16 Aug 2024 06:11:29 EDT (-0400)
  Comparing mesh grass with faked grass (32+20 kbu) (Message 9 to 18 of 18)  
<<< Previous 8 Messages Goto Initial 10 Messages
From: Norbert Kern
Subject: Re: Comparing mesh grass with faked grass (32+20 kbu)
Date: 3 Apr 2002 11:32:40
Message: <3cab2ea8@news.povray.org>
> Your grass is very nice however I wonder how it works in real scene - with
> radosity, more light_sources, fog, media, etc. Does applying of your grass
> require complete rewrite of scene or just replacement of one type of grass
with
> another (and everything is 3d) ?
>
> ABX


for yourself. There are some drawbacks but overall it is a good effect
(perhaps it is also useful as fur texture?).
If I would have used mesh grass, 1 GB peak memory would not have been enough
because of the large area.
In my scene there is no radiosity, but photons worked as "global
illumination". The scene uses fog and absorbing media too. And as Rune
stated, it is "only" a texture.

Norbert


Post a reply to this message

From: Shay
Subject: Re: Comparing mesh grass with faked grass (32+20 kbu)
Date: 3 Apr 2002 11:46:32
Message: <3cab31e8@news.povray.org>
Rune <run### [at] mobilixnetdk> wrote in message
>
> Ok, I really don't know what to conclude, so I'll just let it be up to
> you...
>

I believe that each has it's place. Your macro is a very powerful tool which
will be more effective than mesh grass in many instances. Perhaps for closer
shots, a combination of the two techniques would work well. Mesh grass could
be used to give a distinct blade shape, and your grass could be used to add
density.

 -Shay


Post a reply to this message

From: Tony[B]
Subject: Re: Comparing mesh grass with faked grass (32+20 kbu)
Date: 3 Apr 2002 13:25:57
Message: <3cab4935@news.povray.org>
The mesh grass looks much thicker than the texture grass. There's also less
variation of color in the above image. It's interesting how they both look
bare in the same spots.


Post a reply to this message

From: Hugo
Subject: Re: Comparing mesh grass with faked grass (32+20 kbu)
Date: 3 Apr 2002 14:44:39
Message: <3cab5ba7$1@news.povray.org>
It's interesting to compare these two images. I like the grass texture best,
and 15 MB counts for me in any case. But I assume you used a single, strong
lightsource for the mesh grass, which is why the shadows really show up.
This is the biggest difference in the two images. For a more natural
appearance of the mesh, the shadows need to be much more subtle. This will
increase render time. But I assume multiple lightsources or radiosity won't
change the look a lot of your grass texture ... so all in all, there's a
decision to make for every scene.

Oh and I should say "great work!" once again because I'm surprised this is
possible with a texture - even for animation!

Regards,
Hugo


Post a reply to this message

From: Rune
Subject: Re: Comparing mesh grass with faked grass (32+20 kbu)
Date: 3 Apr 2002 14:59:16
Message: <3cab5f14@news.povray.org>
"Tony[B]" wrote:
> The mesh grass looks much thicker than the texture grass.

That's probably right.

> There's also less variation of color in the above image.

Err, which one of them?

> It's interesting how they both look bare in the same spots.

Glad you find my density maps interesting!

Rune
--
3D images and anims, include files, tutorials and more:
Rune's World:  http://rsj.mobilixnet.dk (updated Mar 19)
POV-Ray Users: http://rsj.mobilixnet.dk/povrayusers/
POV-Ray Ring:  http://webring.povray.co.uk


Post a reply to this message

From: Rune
Subject: UPDATE
Date: 3 Apr 2002 16:22:54
Message: <3cab72ae@news.povray.org>
The mesh grass in my last post didn't really look very good IMO, so I tried
to make it denser and to add another light source, and I think it really
helped. However, it also increased the render time quite a bit, so that my
faked grass texture is faster now after all. I guess one can draw any
conclusions by now, but as many have said, each method will have its own
use.

Faked texture grass:
~~~~~~~~~~~~~~~~~~~~
Time for parse: 0 min.  0 sec.
Time for trace: 6 min.  3 sec.
Time total:     6 min.  3 sec.
Peak memory used:  0.36 MB


Real mesh grass:
~~~~~~~~~~~~~~~~
Time for parse:  0 min. 15 sec.
Time for trace: 14 min. 39 sec.
Time total:     14 min. 54 sec.
Peak memory used: 23.30 MB

Grass blades: 75^2*25^2 = 3515625


The grass texture looks a bit more boring with the second light source
added, because there's less contrast, but such things will always vary from
scene to scene.

Rune
--
3D images and anims, include files, tutorials and more:
Rune's World:  http://rsj.mobilixnet.dk (updated Mar 19)
POV-Ray Users: http://rsj.mobilixnet.dk/povrayusers/
POV-Ray Ring:  http://webring.povray.co.uk


Post a reply to this message


Attachments:
Download 'grassreal2.jpg' (29 KB) Download 'grassfaked2.jpg' (21 KB)

Preview of image 'grassreal2.jpg'
grassreal2.jpg

Preview of image 'grassfaked2.jpg'
grassfaked2.jpg


 

From: Sir Charles W  Shults III
Subject: Re: UPDATE
Date: 3 Apr 2002 17:19:50
Message: <3cab8006$1@news.povray.org>
Ah, the upper picture is St. Augustine grass, and the lower is Bahia or
Zoysia...

Cheers!

Chip Shults
My robotics, space and CGI web page - http://home.cfl.rr.com/aichip


Post a reply to this message

From: Daniel Matthews
Subject: Re: Comparing mesh grass with faked grass (32+20 kbu)
Date: 3 Apr 2002 19:50:56
Message: <72719949.rKQ5TDuZKl@3-e.net>
Your fake looks good/best, but can you combine the two methods so that 
there is a little more shadow variation in the final image?

Rune wrote:

> I used TonyB's modified version of Gilles Tran's mesh grass to compare how
> mesh grass renders and looks compared to my faked grass.
> 
> Both images were rendered at 800x600 AA0.3, then scaled to half size
> before posting here.
> 
> Faked texture grass:
> ~~~~~~~~~~~~~~~~~~~~
> Time for parse: 0 min.  0 sec.
> Time for trace: 5 min. 58 sec.
> Time total:     5 min. 58 sec.
> Peak memory used:  0.36 MB
> 
> 
> Real mesh grass:
> ~~~~~~~~~~~~~~~~
> Time for parse: 0 min. 10 sec.
> Time for trace: 4 min.  8 sec.
> Time total:     4 min. 18 sec.
> Peak memory used: 15.09 MB
> 
> As you can see, my faked grass renders a bit slower. The real mesh grass
> renders quicker, it doesn't use particularly much memory, and it parses in
> almost no time. I expect that the fields of grass must be extremely huge
> before my faked grass becomes more efficient than mesh grass...
> 
> However, my grass looks rather different than the mesh grass. It has a
> softer look, which I rather like. I may be of course, that I just haven't
> been able to tweak the mesh grass to look good in my scene...
> 
> Ok, I really don't know what to conclude, so I'll just let it be up to
> you...
> 
> Rune
> --
> 3D images and anims, include files, tutorials and more:
> Rune's World:  http://rsj.mobilixnet.dk (updated Mar 19)
> POV-Ray Users: http://rsj.mobilixnet.dk/povrayusers/
> POV-Ray Ring:  http://webring.povray.co.uk


Post a reply to this message

From: Tony[B]
Subject: Re: UPDATE
Date: 3 Apr 2002 21:23:50
Message: <3cabb936@news.povray.org>
I suppose that, if anything, one can conclude that the mesh grass allows for
a greater variety of grass types and is actually 3D (casting shadows, etc.),
while the texture grass saves memory and rendertime. I would like to see
them combined somehow. Like a distance-based grass macro, that flows from
mesh to texture as distance increases. It'd be hard to get them to match and
figure a proper distance algorithm, though...


Post a reply to this message

From: Dave Dunn
Subject: Re: Comparing mesh grass with faked grass (32+20 kbu)
Date: 4 Apr 2002 10:48:07
Message: <3CAC7578.D6B4F9F8@aol.com>
Rune wrote:

> I used TonyB's modified version of Gilles Tran's mesh grass to compare how
> mesh grass renders and looks compared to my faked grass.

What struck me was the similarity of the mesh grass to the PDI "style" as seen
in Shrek, and the texture-based "faked" grass to Pixr's style for A Bug's Life.


Post a reply to this message

<<< Previous 8 Messages Goto Initial 10 Messages

Copyright 2003-2023 Persistence of Vision Raytracer Pty. Ltd.