 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Severi Salminen" <sev### [at] NOT_THISsiba fi> wrote in message
news:4028fff1$1@news.povray.org...
> Phil Cook wrote:
>
> > The camera is inside a hollowed out box, there are no openings to the
> > outside world. I'm pointing a set of parallel lights from outside the
> > box to a point inside.
> >
> > So. where's the light coming from?
> >
> > Am I doing something fundementally wrong (most likely) or is this some
> > property of parallel light's I don't know about showing up?
>
> Read this section in the manual:
>
> 6.5.7.4 Parallel Lights
Hopefully Phil will see that you are probably referring to the plane of
light projection. So, Phil, you'll need to watch out for that plane of light
and any intersecting with objects it might do. I realize the doc says this,
basically, but perhaps not in a way everyone will understand-- which is
pictures like Phil's. :-)
--
Bob H.
http://www.3digitaleyes.com
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On Tue, 10 Feb 2004 10:26:46 -0600, Hughes, B. <omn### [at] charternet>
wrote:
> "Severi Salminen" <sev### [at] NOT_THISsibafi> wrote in message
> news:4028fff1$1@news.povray.org...
>> Phil Cook wrote:
>>
>> > The camera is inside a hollowed out box, there are no openings to the
>> > outside world. I'm pointing a set of parallel lights from outside the
>> > box to a point inside.
>> >
>> > So. where's the light coming from?
>> >
>> > Am I doing something fundementally wrong (most likely) or is this some
>> > property of parallel light's I don't know about showing up?
>>
>> Read this section in the manual:
>>
>> 6.5.7.4 Parallel Lights
>
> Hopefully Phil will see that you are probably referring to the plane of
> light projection. So, Phil, you'll need to watch out for that plane of
> light
> and any intersecting with objects it might do. I realize the doc says
> this,
> basically, but perhaps not in a way everyone will understand-- which is
> pictures like Phil's. :-)
>
You caught me in the middle of replying to Severi; so I'll start again
here. Yep I guessed it had something to do with the light plane, but can't
figure out how it works. I'm producing a plane presumely based at location
perpendicular to the point_at so if my location was at 0 and point_at y
then then it would be equivalent a plane{0,y} I assume?
I've just read what I wrote and worked it out: the plane I'm producing is
passing through the box; as the light is taken from the closet point of
the plane I'm getting the equivalent of an internal light source. Sheesh
Thanks for all the help
--
Phil
--
All thoughts and comments are my own unless otherwise stated and I am
happy to be proven wrong.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <opr25yj0lyp4ukzs@news.povray.org>,
Phil Cook <phi### [at] nospamdeckingdealscouk> wrote:
> You caught me in the middle of replying to Severi; so I'll start again
> here. Yep I guessed it had something to do with the light plane, but can't
> figure out how it works. I'm producing a plane presumely based at location
> perpendicular to the point_at so if my location was at 0 and point_at y
> then then it would be equivalent a plane{0,y} I assume?
No. The point_at vector is a location, the plane vector is a direction
(the plane normal). Thought the two are the same when the location is <
0, 0, 0>, as in your example.
> I've just read what I wrote and worked it out: the plane I'm producing is
> passing through the box; as the light is taken from the closet point of
> the plane I'm getting the equivalent of an internal light source. Sheesh
Right.
--
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
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On Tue, 10 Feb 2004 12:44:21 -0500, Christopher James Huff
<cja### [at] earthlinknet> wrote:
> In article <opr25yj0lyp4ukzs@news.povray.org>,
> Phil Cook <phi### [at] nospamdeckingdealscouk> wrote:
>
>> You caught me in the middle of replying to Severi; so I'll start again
>> here. Yep I guessed it had something to do with the light plane, but
>> can't
>> figure out how it works. I'm producing a plane presumely based at
>> location
>> perpendicular to the point_at so if my location was at 0 and point_at y
>> then then it would be equivalent a plane{0,y} I assume?
>
> No. The point_at vector is a location, the plane vector is a direction
> (the plane normal). Thought the two are the same when the location is <
> 0, 0, 0>, as in your example.
Errr. If light-source{0 rgb 1 parallel point_at y} then surely if I
connect a line between location and point_at than a plane perpendicular to
this would be lying at 0 in the XZ plane ie. plane {0,y} and not
plane{y,y} or even plane{y,-y}; in the same way that light_source{y rgb 1
parallel point_at <1,1,0>} would produce the equivalent of a plane{0,x} or
plane{y,x} to be more precise: otherwise the scene I've just produced
wouldn't work; unless I'm just lucky :)
>> I've just read what I wrote and worked it out: the plane I'm producing
>> is
>> passing through the box; as the light is taken from the closet point of
>> the plane I'm getting the equivalent of an internal light source. Sheesh
>
> Right.
So I moved the light location to:
light_source{<-5,20,WindowGap*WindowCount> rgb 1 parallel point_at
<10,0,WindowGap*WindowCount>} which produces just the right results. Now
I've figured it out I assume that I don't need to reproduce the lights at
multiple points as the light plane will cover the entire scene. Though
just doing a quick render I prefer the extra amount of light; I'll just
hike the rgb up. Hmm much faster.
--
Phil
--
All thoughts and comments are my own unless otherwise stated and I am
happy to be proven wrong.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <opr27akwh2p4ukzs@news.povray.org>,
Phil Cook <phi### [at] nospamdeckingdealscouk> wrote:
> > No. The point_at vector is a location, the plane vector is a direction
> > (the plane normal). Thought the two are the same when the location is <
> > 0, 0, 0>, as in your example.
>
> Errr. If light-source{0 rgb 1 parallel point_at y} then surely if I
> connect a line between location and point_at than a plane perpendicular to
> this would be lying at 0 in the XZ plane ie.
Right. The surface normal direction and the point_at location match up,
because the light location is < 0, 0, 0>. You could use a non-normalized
vector as well, it doesn't matter as long as they point in the same
direction.
> plane {0,y} and not plane{y,y} or even plane{y,-y};
Er, no. It would be "plane {y, 0}". The last two aren't even valid
planes...the POV primitive takes a normal vector and displacement
distance.
> in the same way that light_source{y rgb 1
> parallel point_at <1,1,0>} would produce the equivalent of a plane{0,x} or
> plane{y,x} to be more precise: otherwise the scene I've just produced
> wouldn't work; unless I'm just lucky :)
A location of < 0, 1, 0> (y) and point_at of < 1, 1, 0> would lead to a
direction of < 1, 0, 0>, or x. The equivalent plane would be "plane {x,
0}". Again, "plane {y, x}" isn't even a valid plane.
--
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
|
|
| |
| |
|
|
|
|
| |
| |
|
|
You guys are starting to confuse me... Only joking.
I had forgotten how parallel lights treat fading so I gave it a look in a
few renders by building on Phil's example scene (see image). Maybe this will
help a few people understand parallel lights a little better.
http://www.3digitaleyes.com/parallel_light.jpg
Maybe this picture is worth a thousand words-- or at least a paragraph. :-)
I know the documention tells how fading (attenuation) is same as for
ordinary point lights but I wanted to be sure if that meant for all
directions and not just the point_at one. I probably knew already but
couldn't remember. The blue cones are where the light sources are and where
they point. Inside the box is a yellow cone to show that 'parallel'
direction and shadow; no shadow in opposite direction, as per the
documentation (unidirectional).
Bob H.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On Wed, 11 Feb 2004 08:06:22 -0500, Christopher James Huff
<cja### [at] earthlinknet> wrote:
> In article <opr27akwh2p4ukzs@news.povray.org>,
> Phil Cook <phi### [at] nospamdeckingdealscouk> wrote:
>
>> > No. The point_at vector is a location, the plane vector is a direction
>> > (the plane normal). Thought the two are the same when the location is
>> <
>> > 0, 0, 0>, as in your example.
>>
>> Errr. If light-source{0 rgb 1 parallel point_at y} then surely if I
>> connect a line between location and point_at than a plane perpendicular
>> to
>> this would be lying at 0 in the XZ plane ie.
>
> Right. The surface normal direction and the point_at location match up,
> because the light location is < 0, 0, 0>. You could use a non-normalized
> vector as well, it doesn't matter as long as they point in the same
> direction.
>
>
>> plane {0,y} and not plane{y,y} or even plane{y,-y};
>
> Er, no. It would be "plane {y, 0}". The last two aren't even valid
> planes...the POV primitive takes a normal vector and displacement
> distance.
>
>
>> in the same way that light_source{y rgb 1
>> parallel point_at <1,1,0>} would produce the equivalent of a plane{0,x}
>> or
>> plane{y,x} to be more precise: otherwise the scene I've just produced
>> wouldn't work; unless I'm just lucky :)
>
> A location of < 0, 1, 0> (y) and point_at of < 1, 1, 0> would lead to a
> direction of < 1, 0, 0>, or x. The equivalent plane would be "plane {x,
> 0}". Again, "plane {y, x}" isn't even a valid plane.
>
Sorry my bad I don't use planes that often and was confusing my terms;
again sorry, of course plane{y,y} should have been plane{y,1} etc. Odd
that when I code I don't tend to get it wrong but do when I try to post it
in a newsgroup for the whole world to see :)
--
Phil
--
All thoughts and comments are my own unless otherwise stated and I am
happy to be proven wrong.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On Wed, 11 Feb 2004 08:18:51 -0600, Hughes, B. <omn### [at] charternet>
wrote:
> You guys are starting to confuse me... Only joking.
>
> I had forgotten how parallel lights treat fading so I gave it a look in a
> few renders by building on Phil's example scene (see image). Maybe this
> will
> help a few people understand parallel lights a little better.
>
> http://www.3digitaleyes.com/parallel_light.jpg
>
> Maybe this picture is worth a thousand words-- or at least a paragraph.
> :-)
>
> I know the documention tells how fading (attenuation) is same as for
> ordinary point lights but I wanted to be sure if that meant for all
> directions and not just the point_at one. I probably knew already but
> couldn't remember. The blue cones are where the light sources are and
> where
> they point. Inside the box is a yellow cone to show that 'parallel'
> direction and shadow; no shadow in opposite direction, as per the
> documentation (unidirectional).
Thanks that's interesting, and I'm sure will come in useful for some
interesting lighting techniques. Out of curiosity as parallel can be used
with any light_source do you know the effects of combining it with area
lights I'm imagining an area light in the XY plane with a 45 degree
point_at would a plane be created for each light?
--
Phil
--
All thoughts and comments are my own unless otherwise stated and I am
happy to be proven wrong.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> Sorry my bad I don't use planes that often and was confusing my terms;
> again sorry, of course plane{y,y} should have been plane{y,1} etc. Odd
> that when I code I don't tend to get it wrong but do when I try to post
> it in a newsgroup for the whole world to see :)
It is because there are 5+ billion people watching every single letter
you write to this newsgroup - you just get a little nervous ;-) But
you'll get used to it...
Severi
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Phil Cook" <phi### [at] nospamdeckingdealscouk> wrote in message
news:opr27syqj0p4ukzs@news.povray.org...
> interesting lighting techniques. Out of curiosity as parallel can be used
> with any light_source do you know the effects of combining it with area
> lights I'm imagining an area light in the XY plane with a 45 degree
> point_at would a plane be created for each light?
Hmmm. I don't really know the answer to that. Common sense would be yes to
that, since each point in the array ought to look toward the point_at
instead of a central point source... except you know how area_light behaves
for specular hilights, right? Single point. So it could be that it becomes
singular in the same way.
Only shadows are treated to the multiplicity of the arrays, which makes it
all about shadows. Maybe it actually skews along lines between array
rectangle and point_at vector. Or could be that it makes the array behave as
though 'orient' were used.
Something to check into. Now you've got me rethinking, even moreso, what I
thought I already knew. Like so many others, I typically go about finding
out about such things by experimentation. I.e. scene creation first, then
discovery of the results. I'm not always working from the documentation.
Bob H.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
|
|
