|
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
After reading the thread in p.a-u on are lights, I decided to experiment.
In the attached images I tried three different light schemes, a single point
light, an area light and an array of point lights.
The lights are all in the same location, 50 units above a 50 unit sphere
(with phong highlighting)
The arealight and array of point lights have the same number of lights and
are the same size, 8x8 lights, over a 150x150 unit area (area light w/o
jitter)
The white balls show the positions of the lights in each
The images are:
PLight.png - single point light
ALight.png - area light
APLight.png - array of area lights
As the images show, area lights do not work quite as one would expect (but
that is not to say they do not work as intended*).
The point light (for comparison) shows clearly the hard shadow of the
sphere, the hard shadowline (separating the dark and illuminated sides of
the sphere) and the phong highlight.
The area light shows the correct 'soft shadowing' of the sphere, however,
the phong highlight is identical to that of the point light and the
shadowline is similar save for a bit of softening. (*This is how it is
intended to work though, according to the docs)
The array of area lights, on the other hand, shows closer to what would be
expected in real life, the soft shadows (similar to that of the area light),
but also the phong highlight shape (perhaps not as bright as it should be,
due to light intesities used) and most importantly, the shadowline.
For objects that are relatively far/large compared to the arealight, the
shadowline effect is not significantly noticeable, but in a situation such
as this, where the object is near/small compared to the area light, the
difference is quite noticeable.
Any other comments on this?
-tgq
Post a reply to this message
Attachments:
Download 'APLight.png' (17 KB)
Download 'ALight.png' (15 KB)
Download 'PLight.png' (9 KB)
Preview of image 'APLight.png'
Preview of image 'ALight.png'
Preview of image 'PLight.png'
|
|
| |
| |
|
|
|
|
| |
| |
|
|
dammit, my images always post in the wrong order.
-tgq
Post a reply to this message
|
|
| |
| |
|
|
From: Thomas Willhalm
Subject: Re: True Area Light Sources (from p.a-u)
Date: 5 Dec 2002 04:03:23
Message: <3def165b@news.povray.org>
|
|
|
| |
| |
|
|
TinCanMan wrote:
>
> The area light shows the correct 'soft shadowing' of the sphere, however,
> the phong highlight is identical to that of the point light and the
> shadowline is similar save for a bit of softening. (*This is how it is
> intended to work though, according to the docs)
> Any other comments on this?
IMO it would be nice to have "real" area lights. By this I mean an adaptive
array of point light sources including jittering. It would be faster than
an array of point lights, but I know that it would also be slower than
the current "area lights". However, isn't raytracing supposed to be slow?
Thomas
P.S.:Some years ago, I tried to hack the code to change this behavior.
Since my knowledge of the povray sources was (and still is) small, I failed
and lost interest later on.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Hey,
>Any other comments on this?
Well, in August of this year, I did pretty much the same set of experiments,
and wrote up a short article, which you can read at:
http://www.redbrick.dcu.ie/~element/areaLights/
Now, I didn't have all my facts straight when I wrote it, and the subsequent
flame war that was targeted at me on povray.general quickly pointed out
that Povray's arealights are *not* designed to give area illumination, but
simply to simulate area *shadows*. I agree though, a source of area
illumination in Povray would be most welcome...
John
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"John Mellerick" <ele### [at] redbrickdcuie> wrote in message
news:web.3def2a228cef6e97dbca1a080@news.povray.org...
> Now, I didn't have all my facts straight when I wrote it, and the
subsequent
> flame war that was targeted at me on povray.general quickly pointed
out
> that Povray's arealights are *not* designed to give area illumination,
but
> simply to simulate area *shadows*. I agree though, a source of area
> illumination in Povray would be most welcome...
How about surfaces emitting light and interacting accordingly?
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Hey,
>How about surfaces emitting light and interacting accordingly?
Hmm, not sure exactly what you mean here, but if you mean some kind of
native "object light" (like the kind of thing that is available in various
add-on renderers for 3DSMax), then yeah, that would be cool too ;)
John
Post a reply to this message
|
|
| |
| |
|
|
From: Marc-Hendrik Bremer
Subject: Re: True Area Light Sources (from p.a-u)
Date: 5 Dec 2002 11:30:46
Message: <3def7f36@news.povray.org>
|
|
|
| |
| |
|
|
news:3def762d@news.povray.org...
> How about surfaces emitting light and interacting accordingly?
>
>
Radiosity?
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
With radiosity you can do that! Just give an object a very high ambience and
maybe also color rgb > 1.
John Mellerick wrote:
>Hey,
>
>>How about surfaces emitting light and interacting accordingly?
>
>Hmm, not sure exactly what you mean here, but if you mean some kind of
>native "object light" (like the kind of thing that is available in various
>add-on renderers for 3DSMax), then yeah, that would be cool too ;)
>
>
>John
>
Post a reply to this message
|
|
| |
| |
|
|
From: Xplo Eristotle
Subject: Re: True Area Light Sources (from p.a-u)
Date: 5 Dec 2002 12:26:10
Message: <3def8c32@news.povray.org>
|
|
|
| |
| |
|
|
John Mellerick wrote:
>
> Well, in August of this year, I did pretty much the same set of experiments,
> and wrote up a short article, which you can read at:
>
> http://www.redbrick.dcu.ie/~element/areaLights/
>
> Now, I didn't have all my facts straight when I wrote it, and the subsequent
> flame war that was targeted at me on povray.general quickly pointed out
> that Povray's arealights are *not* designed to give area illumination, but
> simply to simulate area *shadows*. I agree though, a source of area
> illumination in Povray would be most welcome...
And slow. Which, as I recall, was half the reason for the "flame war";
you claimed reasonable speeds for your point light arrays, but you were
using very simple scenes for your tests, and anyone used to creating
complex scenes in POV-Ray (or any other raytracer I've seen, for that
matter) knows that adding more lights can have a major impact on render
times.. a fact which you stubbornly refused to accept, based on your
limited testing.
Generally speaking, I don't think a point light array object is needed;
they are easily written with loops, which would also provide more
control over the individual lights in the array than an object's syntax
probably would. The only real benefit I can see to having such an object
would be if POV-Ray could make reasonable assumptions about a predefined
light array that would make rendering faster than an improvised one, but
I have no idea whether this would be the case.
-Xplo
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Apache" <apa### [at] yahoocom> wrote in message
news:web.3def81ce8cef6e97863e7dd0@news.povray.org...
> With radiosity you can do that! Just give an object a very high
ambience and
> maybe also color rgb > 1.
Does it work with photons?
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |