 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
I haven't followed too closely this HDRI discussion.
Looking at the few pages and images posted to p.b.i I see that I haven't
the slightest clue about what is it all about. The images look very nice,
but I can't see what is the HDRI thing in those images.
So could someone explain exactly what is HDRI and how it affects
rendering? If it's an algorithm to make something better, could someone
make images where the regular way is used and the same scene with HDRI
is used so that they can be compared side by side? If it's something
completely new, not comparable to anything in POV-Ray, could someone
explain how it affects the images?
--
#macro M(A,N,D,L)plane{-z,-9pigment{mandel L*9translate N color_map{[0rgb x]
[1rgb 9]}scale<D,D*3D>*1e3}rotate y*A*8}#end M(-3<1.206434.28623>70,7)M(
-1<.7438.1795>1,20)M(1<.77595.13699>30,20)M(3<.75923.07145>80,99)// - Warp -
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Hey Warp,
An HDRI image is an image map that stores brightness values greater than
the typical [0,1] (it uses floating point numbers to represent [0,
infinity]). With this, you can use a HDRI environment map as a light source
to realistically light a scene by simulating the lighting coming from the
environment.
What I'm not sure on is how POV-Ray or mlpov deals with the extended
range, when POV-Ray's lighting engine only handles the [0,1] range. The
pictures do look good though, so it seems to be working.
Regards,
George Pantazopoulos
"Warp" <war### [at] tag povrayorg> wrote in message
news:3e0ce0a8@news.povray.org...
> I haven't followed too closely this HDRI discussion.
> Looking at the few pages and images posted to p.b.i I see that I haven't
> the slightest clue about what is it all about. The images look very nice,
> but I can't see what is the HDRI thing in those images.
>
> So could someone explain exactly what is HDRI and how it affects
> rendering? If it's an algorithm to make something better, could someone
> make images where the regular way is used and the same scene with HDRI
> is used so that they can be compared side by side? If it's something
> completely new, not comparable to anything in POV-Ray, could someone
> explain how it affects the images?
>
> --
> #macro M(A,N,D,L)plane{-z,-9pigment{mandel L*9translate N color_map{[0rgb
x]
> [1rgb 9]}scale<D,D*3D>*1e3}rotate y*A*8}#end M(-3<1.206434.28623>70,7)M(
> -1<.7438.1795>1,20)M(1<.77595.13699>30,20)M(3<.75923.07145>80,99)// -
Warp -
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <3e0cef3e$1@news.povray.org>,
"George Pantazopoulos" <the### [at] attbicom*KILLSPAM*> wrote:
> What I'm not sure on is how POV-Ray or mlpov deals with the extended
> range, when POV-Ray's lighting engine only handles the [0,1] range.
This isn't true. Values only get clamped to the [0, 1] range after being
computed. An rgb 2 light is twice as bright as an rgb 1 light, and half
as bright as an rgb 4 light.
--
Christopher James Huff <cja### [at] earthlinknet>
http://home.earthlink.net/~cjameshuff/
POV-Ray TAG: chr### [at] tagpovrayorg
http://tag.povray.org/
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Thanks for the correction Christopher, thats great!
George
"Christopher James Huff" <chr### [at] maccom> wrote in message
news:chr### [at] netplexaussieorg...
> In article <3e0cef3e$1@news.povray.org>,
> "George Pantazopoulos" <the### [at] attbicom*KILLSPAM*> wrote:
>
> > What I'm not sure on is how POV-Ray or mlpov deals with the extended
> > range, when POV-Ray's lighting engine only handles the [0,1] range.
>
> This isn't true. Values only get clamped to the [0, 1] range after being
> computed. An rgb 2 light is twice as bright as an rgb 1 light, and half
> as bright as an rgb 4 light.
>
> --
> Christopher James Huff <cja### [at] earthlinknet>
> http://home.earthlink.net/~cjameshuff/
> POV-Ray TAG: chr### [at] tagpovrayorg
> http://tag.povray.org/
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
George Pantazopoulos <the### [at] attbicom*killspam*> wrote:
> With this, you can use a HDRI environment map as a light source
> to realistically light a scene by simulating the lighting coming from the
> environment.
How this is done? Simply using the regular radiosity algorithm in POV-Ray?
Or is there a new algorithm for this as well?
--
#macro N(D)#if(D>99)cylinder{M()#local D=div(D,104);M().5,2pigment{rgb M()}}
N(D)#end#end#macro M()<mod(D,13)-6mod(div(D,13)8)-3,10>#end blob{
N(11117333955)N(4254934330)N(3900569407)N(7382340)N(3358)N(970)}// - Warp -
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Warp" <war### [at] tagpovrayorg> a écrit dans le message de news:
3e0d620d@news.povray.org...
> How this is done? Simply using the regular radiosity algorithm in
POV-Ray?
> Or is there a new algorithm for this as well?
From what I've understood, there's nothing really new in the way POV-Ray
handles radiosity itself, it's the map that does all the work and that's the
real beauty of it. The radiosity itself is traditional but because light
intensity has a much wider range in the HDR map than in a usual map (like in
the real world) it makes the results much more natural-looking, even in the
absence of specular highlights. This becomes clearer if you load HDR maps in
HDRShop (at Paul Debevec's page http://www.debevec.org) and play with the
exposure settings. The maps then reveal themselves in their 2 dimensions
(color and light intensity).
G.
--
**********************
http://www.oyonale.com
**********************
- Graphic experiments
- POV-Ray and Poser computer images
- Posters
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <3e0d9909@news.povray.org>,
"Gilles Tran" <git### [at] wanadoofr> wrote:
> From what I've understood, there's nothing really new in the way POV-Ray
> handles radiosity itself, it's the map that does all the work and that's the
> real beauty of it. The radiosity itself is traditional but because light
> intensity has a much wider range in the HDR map than in a usual map (like in
> the real world) it makes the results much more natural-looking, even in the
> absence of specular highlights. This becomes clearer if you load HDR maps in
> HDRShop (at Paul Debevec's page http://www.debevec.org) and play with the
> exposure settings. The maps then reveal themselves in their 2 dimensions
> (color and light intensity).
So, instead of taking samples from the actual scene, it takes them from
a separately calculated image? The precision might be better than 8 or
16 bpc images, but it will still almost definitely be worse than float
precision. And if it is applied like an environment map, it has the same
problems: flat surfaces will have no shading or shadows. UV mapping
would make more sense, you could take the object geometry into account,
though you still have to make sure the map is at a high enough
resolution.
I could understand it having an advantage over using more common image
formats for light mapping, but what use does it have in POV? Is the
patch for output of HDRI images? That way, POV could generate them for
other, more limited programs...
--
Christopher James Huff <cja### [at] earthlinknet>
http://home.earthlink.net/~cjameshuff/
POV-Ray TAG: chr### [at] tagpovrayorg
http://tag.povray.org/
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Christopher James Huff" <chr### [at] maccom> a écrit dans le message de
news: chr### [at] netplexaussieorg...
> In article <3e0d9909@news.povray.org>,
> "Gilles Tran" <git### [at] wanadoofr> wrote:
> I could understand it having an advantage over using more common image
> formats for light mapping, but what use does it have in POV?
Hmm, it looks that you haven't seen what's going on in p.b.i these days.
Come on, give it a try, you'll see the advantage by yourself :-) (and I
didn't believe it before I was given the patch). With HDR, we now have
POV-Ray images virtually impossible to tell from actual photographs. See
JRG's watch pic for instance.
Actually, HDR effects can be done in POV-Ray without a patch, by using
radiosity with a texture_map with a very high range of ambient values (this
is what Ive did for his Vermeer images) but it's not exactly easy to set up.
Setting up a simple HDR scene is a piece of cake by comparison.
G.
--
**********************
http://www.oyonale.com
**********************
- Graphic experiments
- POV-Ray and Poser computer images
- Posters
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Gilles Tran wrote:
>
> "Christopher James Huff" <chr### [at] maccom> a écrit dans le message de
> news: chr### [at] netplexaussieorg...
>> In article <3e0d9909@news.povray.org>,
>> "Gilles Tran" <git### [at] wanadoofr> wrote:
>> I could understand it having an advantage over using more common image
>> formats for light mapping, but what use does it have in POV?
>
> Hmm, it looks that you haven't seen what's going on in p.b.i these days.
> Come on, give it a try, you'll see the advantage by yourself :-) (and I
> didn't believe it before I was given the patch). With HDR, we now have
> POV-Ray images virtually impossible to tell from actual photographs. See
> JRG's watch pic for instance.
In my opinion HDRI is the next step closer to reality and POV-Ray will
become a high end renderering system. The images in p.b.i are - more ore
less - impressive because of the natural way of illumination.
Regards,
Andreas
--
http://www.render-zone.com
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <3e0e4bc9$2@news.povray.org>,
"Gilles Tran" <git### [at] wanadoofr> wrote:
> Hmm, it looks that you haven't seen what's going on in p.b.i these days.
I'm on a modem connection, so I haven't kept caught up.
> Come on, give it a try, you'll see the advantage by yourself :-) (and I
> didn't believe it before I was given the patch). With HDR, we now have
> POV-Ray images virtually impossible to tell from actual photographs. See
> JRG's watch pic for instance.
But it is still less accurate than radiosity could be, as far as I can
tell. Why use a high dynamic range environment map instead of sampling
the actual environment? Speed? The ability to use real-world samples
instead of coding a background?
--
Christopher James Huff <cja### [at] earthlinknet>
http://home.earthlink.net/~cjameshuff/
POV-Ray TAG: chr### [at] tagpovrayorg
http://tag.povray.org/
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Christopher James Huff" <chr### [at] maccom> a écrit dans le message de
news: chr### [at] netplexaussieorg...
> But it is still less accurate than radiosity could be, as far as I can
> tell.
Accuracy here is really a non-issue given the benefits...
>Why use a high dynamic range environment map instead of sampling
> the actual environment?
> Speed? The ability to use real-world samples
> instead of coding a background?
Creating complex environments just to make them look nice in reflections and
project the right lights is rarely done :
for practical purposes, you just don't code the whole St Peters Basilic to
make a background (or you've really been raytracing too long...). There's a
example by Jaime who did a fake HDRI that way, but he used a complete scene
he had done before.
Speed is of course a big issue there too, since maps render very fast.
Also, natural light environments taken from photographs have a complexity
and a richness that is unmatched by usual 3D lighting, which is why people
go ooh and aah in front of these pics.
Now HDR isn't the solution to all lighting problems, and is only valuable in
certain circumstances, like "packshot" images (close ups of objects), but
then it is very good and extremly simple to set up using regular radiosity
parameters.
G.
--
**********************
http://www.oyonale.com
**********************
- Graphic experiments
- POV-Ray and Poser computer images
- Posters
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
I'm starting to get interested...
--
#macro M(A,N,D,L)plane{-z,-9pigment{mandel L*9translate N color_map{[0rgb x]
[1rgb 9]}scale<D,D*3D>*1e3}rotate y*A*8}#end M(-3<1.206434.28623>70,7)M(
-1<.7438.1795>1,20)M(1<.77595.13699>30,20)M(3<.75923.07145>80,99)// - Warp -
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On Mon, 30 Dec 2002 13:02:54 -0500
Christopher James Huff <chr### [at] maccom> wrote:
> The ability to use real-world samples
> instead of coding a background?
Yes, I think it's main purpouse is to create "augmented reality", but
it's good also for some other interesting things, as Gilles said.
--
Jaime Vives Piqueres
La Persistencia de la Ignorancia
http://www.ignorancia.org
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Can you tell me a bit more about the implementation here? For example, I
have rendered a few scenes using an HDR file and I would like to "mask out"
the actual original scene that the HDR file contains. Is there a way to do
that? If I put a plane or perhaps a wall in front of the "background" image
that appears in my scene, will that block out the light coming from the HDR
file? In other words, does the placement of the HDR file in the scene act
exactly like a light-emitting source? That's the part I don't quite get.
In effect, the probes I have used are of a campus, a beach, a kitchen, etc.
and you can always see some part of that original scene in the POV scene.
I'd like to "sample" the illumination of the HDR file but leave the kitchen
(or the beach) out of my own picture.
Thanks much,
Dennis
"Gilles Tran" <git### [at] wanadoofr> wrote in message
news:3e10983b@news.povray.org...
>
> "Christopher James Huff" <chr### [at] maccom> a écrit dans le message de
> news: chr### [at] netplexaussieorg...
> > But it is still less accurate than radiosity could be, as far as I can
> > tell.
> Accuracy here is really a non-issue given the benefits...
>
> >Why use a high dynamic range environment map instead of sampling
> > the actual environment?
> > Speed? The ability to use real-world samples
> > instead of coding a background?
>
> Creating complex environments just to make them look nice in reflections
and
> project the right lights is rarely done :
> for practical purposes, you just don't code the whole St Peters Basilic to
> make a background (or you've really been raytracing too long...). There's
a
> example by Jaime who did a fake HDRI that way, but he used a complete
scene
> he had done before.
> Speed is of course a big issue there too, since maps render very fast.
> Also, natural light environments taken from photographs have a complexity
> and a richness that is unmatched by usual 3D lighting, which is why people
> go ooh and aah in front of these pics.
> Now HDR isn't the solution to all lighting problems, and is only valuable
in
> certain circumstances, like "packshot" images (close ups of objects), but
> then it is very good and extremly simple to set up using regular radiosity
> parameters.
>
> G.
>
> --
> **********************
> http://www.oyonale.com
> **********************
> - Graphic experiments
> - POV-Ray and Poser computer images
> - Posters
>
>
>
>
>
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <3e10983b@news.povray.org>,
"Gilles Tran" <git### [at] wanadoofr> wrote:
> "Christopher James Huff" <chr### [at] maccom> a écrit dans le message de
> news: chr### [at] netplexaussieorg...
> > But it is still less accurate than radiosity could be, as far as I can
> > tell.
> Accuracy here is really a non-issue given the benefits...
Well, since my interests are mainly in simulation, "better" == "more
accurate", "less accurate" == "worse". Speed is only part of my equation
for judging a simulation if the quality is a constant. Like glows or
fog, this is a compromise of quality vs. speed...it isn't "better than
radiosity", just faster without being a lot worse, like what ambient was
originally for.
Since you are an artist, you have different criteria about which is
"better".
--
Christopher James Huff <cja### [at] earthlinknet>
http://home.earthlink.net/~cjameshuff/
POV-Ray TAG: chr### [at] tagpovrayorg
http://tag.povray.org/
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Dennis Miller" <dhm### [at] attbicom> a écrit dans le message de news:
3e15d35a$1@news.povray.org...
> file? In other words, does the placement of the HDR file in the scene act
> exactly like a light-emitting source? That's the part I don't quite get.
> In effect, the probes I have used are of a campus, a beach, a kitchen,
etc.
> and you can always see some part of that original scene in the POV scene.
> I'd like to "sample" the illumination of the HDR file but leave the
kitchen
> (or the beach) out of my own picture.
Well, HDR maps really behave like any other map apart the high range stuff
(mildy reflective objects will show proper highlights, radiosity effects
will be more natural-looking etc.). I don't see a way to use only the
illumination but not the image itself, apart carefully setting up the scene
so that the HDR map doesn't show up in the frame and the non-HDR parts shown
are consistent with the rest of the lighting (the sort of trick we use with
regular maps). And no_image only works for the reflective parts, not the
illumination itself (just checked...). Possibly the best would be to use
Jaime's trick, setting up a complete scene and then using it as a map with
the "fake HDR" method(s) he describes.
G.
--
**********************
http://www.oyonale.com
**********************
- Graphic experiments
- POV-Ray and Poser computer images
- Posters
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Thank for your reply. Where is the message in which Jamie describes this
technique> I must have missed it.
But to be clear, if I put a plane on the Y axis beyond my object, I should
still be able to use the illumination from the map even if the map image is
not showing in my scene?
thanks again,
D.
"Gilles Tran" <git### [at] wanadoofr> wrote in message
news:3e162bfa@news.povray.org...
> "Dennis Miller" <dhm### [at] attbicom> a écrit dans le message de news:
> 3e15d35a$1@news.povray.org...
> > file? In other words, does the placement of the HDR file in the scene
act
> > exactly like a light-emitting source? That's the part I don't quite get.
> > In effect, the probes I have used are of a campus, a beach, a kitchen,
> etc.
> > and you can always see some part of that original scene in the POV
scene.
> > I'd like to "sample" the illumination of the HDR file but leave the
> kitchen
> > (or the beach) out of my own picture.
>
> Well, HDR maps really behave like any other map apart the high range stuff
> (mildy reflective objects will show proper highlights, radiosity effects
> will be more natural-looking etc.). I don't see a way to use only the
> illumination but not the image itself, apart carefully setting up the
scene
> so that the HDR map doesn't show up in the frame and the non-HDR parts
shown
> are consistent with the rest of the lighting (the sort of trick we use
with
> regular maps). And no_image only works for the reflective parts, not the
> illumination itself (just checked...). Possibly the best would be to use
> Jaime's trick, setting up a complete scene and then using it as a map with
> the "fake HDR" method(s) he describes.
>
> G.
>
> --
> **********************
> http://www.oyonale.com
> **********************
> - Graphic experiments
> - POV-Ray and Poser computer images
> - Posters
>
>
>
>
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Program ended abnormally on 1/3/03 8:02 PM, Due to a catastrophic Dennis
Miller error:
> Thank for your reply. Where is the message in which Jamie describes this
> technique> I must have missed it.
I think it was in povray.binaries.images.
> But to be clear, if I put a plane on the Y axis beyond my object, I should
> still be able to use the illumination from the map even if the map image is
> not showing in my scene?
The parts of the HDR map behind the camera, yes.
The parts hidden by the plane, no.
> thanks again,
> D.
>
--
/*Francois Labreque*/#local a=x+y;#local b=x+a;#local c=a+b;#macro P(F//
/* flabreque */L)polygon{5,F,F+z,L+z,L,F pigment{rgb 9}}#end union
/* @ */{P(0,a)P(a,b)P(b,c)P(2*a,2*b)P(2*b,b+c)P(b+c,<2,3>)
/* videotron.ca */}camera{orthographic location<6,1.25,-6>look_at a }
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On Fri, 3 Jan 2003 20:02:03 -0500
"Dennis Miller" <dhm### [at] attbicom> wrote:
> Thank for your reply. Where is the message in which Jamie describes
> this technique> I must have missed it.
The basic trick consists on using a surrounding sphere with a
multilayer texture like this one (more or less, I'm still
experimenting with it):
--
// "back-lighting" layer
texture{
pigment{
image_map{
jpeg "test.jpg"
map_type 1
interpolate 2
}
}
finish{
diffuse 0
ambient light_ambient
}
}
// 254 layers to fake the HDR behavoir
texture{
pigment{
image_map{
gif "test.gif" // grayscale version of the original map
map_type 1
#while (il<254)
// transmit 255-il max_intensity/(pow(il+1,2))
filter 255-il max_intensity/(pow(il+1,2))
#local il=il+1;
#end
}
}
finish{
diffuse 0
ambient 0
}
}
// final layer to hide the grayscale maps
texture{
pigment{
image_map{
jpeg "test.jpg"
map_type 1
interpolate 2
// filter all .5
filter all .9
}
}
finish{
diffuse 1
ambient 1
}
}
--
The problem is that each map needs diferent settings that must be
found by trial and error. The advantage is that it can work with any
image where the brightness levels are somewhat proportional to real
luminance. Aditionally, you need a program to convert the image to a
grayscale one with ordered palette (255=white, 0=black). I've done it
with The Gimp, simply saving the grayscale version as gif.
--
Jaime Vives Piqueres
La Persistencia de la Ignorancia
http://www.ignorancia.org
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
|
|
