This time old boards. http://www.pp.htv.fi/kkivisal/woody.jpg

The bump map isn't just a channel/combination from the original. I had
to actually write down what I did because it wasn't simple. High-pass
filtering, histogram tweaking and layers in multiply mode were involved.
The final texture is, as before, a combination of multiple offset layers
in lighten mode. With aa it renders very slowly.


_____________
Kari Kivisalo

Post a reply to this message

You really have to write a proper tutorial with pictures and screenshots
from whatever graphics program you're using. This is outstanding! :)

Post a reply to this message

Kari Kivisalo wrote:
> 
> This time old boards. http://www.pp.htv.fi/kkivisal/woody.jpg
> 
> The bump map isn't just a channel/combination from the original. I had
> to actually write down what I did because it wasn't simple. High-pass
> filtering, histogram tweaking and layers in multiply mode were involved.
> The final texture is, as before, a combination of multiple offset layers
> in lighten mode. With aa it renders very slowly.

The HF looks almost photographic. (Well, it WAS to start with, but
still, excellent job.)

-Xplo

Post a reply to this message

Just curious:

  If you use a bump_mapped box instead of a heightfield, what is the render
speed difference and how does the image look like?

-- 
#macro N(D,I)#if(I<6)cylinder{M()#local D[I]=div(D[I],104);M().5,2pigment{
rgb M()}}N(D,(D[I]>99?I:I+1))#end#end#macro M()<mod(D[I],13)-6,mod(div(D[I
],13),8)-3,10>#end blob{N(array[6]{11117333955,
7382340,3358,3900569407,970,4254934330},0)}//                     - Warp -

Post a reply to this message

Warp wrote:
>
>   If you use a bump_mapped box instead of a heightfield, what is the render
> speed difference and how does the image look like?

From this angle and light source position the bump_map looks quite
good and when applied to a low polycount mesh would look just as good
as the HF. I could propably use mesh with bump_map for most of the
wooden board objects. I had to use bump_size 1000 which is quite odd
since all the other times I have used bump_maps bump_size higher
than 10 didn't do any good.

http://www.pp.htv.fi/kkivisal/bump.jpg

HF:                  17 min, 15 MB
bump_map:             5 min,  5 MB 
bump_map, normal_on: 38 min,  5 MB (takes less blue color from the sky)

+am2 +a0.5

global_settings{
  assumed_gamma 1.0 
  ini_option "+qr"
  radiosity{
    pretrace_start 0.04
    pretrace_end 0.02
    count 100
    recursion_limit 2 
    nearest_count 2
    error_bound 1
    //normal on
  }
}

_____________
Kari Kivisalo

Post a reply to this message

I thought that it would look that good. Bump maps can sometimes give
surprisingly good results with a lot less overhead than a real surface
displacement.

-- 
#macro N(D,I)#if(I<6)cylinder{M()#local D[I]=div(D[I],104);M().5,2pigment{
rgb M()}}N(D,(D[I]>99?I:I+1))#end#end#macro M()<mod(D[I],13)-6,mod(div(D[I
],13),8)-3,10>#end blob{N(array[6]{11117333955,
7382340,3358,3900569407,970,4254934330},0)}//                     - Warp -

Post a reply to this message

It looks to me like you loose some detail when the bump_map method is used.
The HF image seems more realistic.  The grain seems to be deeper and more 3D
in the HF.  The differences are subtle, though.

Cris

"Kari Kivisalo" <kki### [at] pphtvfi> wrote in message
news:3B626137.C600D2AC@pp.htv.fi...
> Warp wrote:
> >
> >   If you use a bump_mapped box instead of a heightfield, what is the
render
> > speed difference and how does the image look like?
>
> From this angle and light source position the bump_map looks quite
> good and when applied to a low polycount mesh would look just as good
> as the HF. I could propably use mesh with bump_map for most of the
> wooden board objects. I had to use bump_size 1000 which is quite odd
> since all the other times I have used bump_maps bump_size higher
> than 10 didn't do any good.
>
> http://www.pp.htv.fi/kkivisal/bump.jpg
>
> HF:                  17 min, 15 MB
> bump_map:             5 min,  5 MB
> bump_map, normal_on: 38 min,  5 MB (takes less blue color from the sky)
>
> +am2 +a0.5
>
> global_settings{
>   assumed_gamma 1.0
>   ini_option "+qr"
>   radiosity{
>     pretrace_start 0.04
>     pretrace_end 0.02
>     count 100
>     recursion_limit 2
>     nearest_count 2
>     error_bound 1
>     //normal on
>   }
> }
>
> _____________
> Kari Kivisalo

Post a reply to this message

Cris Williams <wor### [at] netscapenet> wrote:
: It looks to me like you loose some detail when the bump_map method is used.
: The HF image seems more realistic.  The grain seems to be deeper and more 3D
: in the HF.  The differences are subtle, though.

  Are you aware of what exactly does a bump_map do? Your text sounds a bit
like you may not be completely sure. I (or someone else) can explain it if
you want (it's rather interesting).

-- 
#macro N(D,I)#if(I<6)cylinder{M()#local D[I]=div(D[I],104);M().5,2pigment{
rgb M()}}N(D,(D[I]>99?I:I+1))#end#end#macro M()<mod(D[I],13)-6,mod(div(D[I
],13),8)-3,10>#end blob{N(array[6]{11117333955,
7382340,3358,3900569407,970,4254934330},0)}//                     - Warp -

Post a reply to this message

Warp wrote:
> 
> Cris Williams <wor### [at] netscapenet> wrote:
> : It looks to me like you loose some detail when the bump_map method is used.
> : The HF image seems more realistic.  The grain seems to be deeper and more 3D
> : in the HF.  The differences are subtle, though.
> 
>   Are you aware of what exactly does a bump_map do? Your text sounds a bit
> like you may not be completely sure. I (or someone else) can explain it if
> you want (it's rather interesting).
> 

Apart from the obvious differences, it's also important to adapt normal
accuracy or heightfield resolution so results are comparable. 

Christoph

-- 
Christoph Hormann <chr### [at] gmxde>
IsoWood include, radiosity tutorial, TransSkin and other 
things on: http://www.schunter.etc.tu-bs.de/~chris/

Post a reply to this message

It is my understanding that the bump_map simulates the grain and cracks in
the wood on a flat surface using lighting tricks, and the height field
actually creates a true 3D surface (virtually, of course).  It seems as if
both methods should yield similar results, but the HF image just looked
sharper (more detailed) to me.  And I would be very happy to receive any
additional information, clarifications, observations, tips, etc. that you
(or anyone else) would care to send my way.  :-)
.
"Warp" <war### [at] tagpovrayorg> wrote in message
news:3b6603b4@news.povray.org...
> Cris Williams <wor### [at] netscapenet> wrote:
> : It looks to me like you loose some detail when the bump_map method is
used.
> : The HF image seems more realistic.  The grain seems to be deeper and
more 3D
> : in the HF.  The differences are subtle, though.
>
>   Are you aware of what exactly does a bump_map do? Your text sounds a bit
> like you may not be completely sure. I (or someone else) can explain it if
> you want (it's rather interesting).
>
> --
> #macro N(D,I)#if(I<6)cylinder{M()#local D[I]=div(D[I],104);M().5,2pigment{
> rgb M()}}N(D,(D[I]>99?I:I+1))#end#end#macro M()<mod(D[I],13)-6,mod(div(D[I
> ],13),8)-3,10>#end blob{N(array[6]{11117333955,
> 7382340,3358,3900569407,970,4254934330},0)}//                     - Warp -

Post a reply to this message