POV-Ray : Newsgroups : povray.binaries.images : scanned blobs Server Time
26 Dec 2024 00:50:05 EST (-0500)
  scanned blobs (Message 1 to 6 of 6)  
From: IMBJR
Subject: scanned blobs
Date: 12 Apr 2004 10:53:00
Message: <407aad4c@news.povray.org>
3-light system used, probably incorrectly tho.


Post a reply to this message


Attachments:
Download 'scan_3_lights.png' (153 KB)

Preview of image 'scan_3_lights.png'
scan_3_lights.png


 

From: JC (Exether)
Subject: Re: scanned blobs
Date: 12 Apr 2004 12:46:44
Message: <407ac7f4$1@news.povray.org>
IMBJR wrote:
> 3-light system used, probably incorrectly tho.

Not so bad. :-)
For some reason, the position for the back light that is given in 
tutorials does not quite work in my scenes, it seems I'm not the only 


lit areas that will let you see the shape of your spheres better.

Try to render your image with only one of the three lights (trying with 
key, fill and back) and you'll see better what happens and which brings 
what.

How many spheres do you have in this render ?

JC

-- 
http://exether.free.fr/irtc (more IRTC stats !)


Post a reply to this message

From: IMBJR
Subject: Re: scanned blobs
Date: 12 Apr 2004 16:14:34
Message: <407af8aa$1@news.povray.org>
"JC (Exether)" <no### [at] spamfr> wrote in message
news:407ac7f4$1@news.povray.org...
> IMBJR wrote:
> > 3-light system used, probably incorrectly tho.
>
> Not so bad. :-)
> For some reason, the position for the back light that is given in
> tutorials does not quite work in my scenes, it seems I'm not the only


> lit areas that will let you see the shape of your spheres better.

If I understand correctly, the angular relationship between the key and
backlight is already at 135 degrees (viewed down the y-axis), since the key
light is 45 degrees off  to the left and the backlight is behind the
subject. It might help if I position the lights by putting them at a
standard position and rotating them around the subject instead of my rather
rough positioning by offsets.

>
> Try to render your image with only one of the three lights (trying with
> key, fill and back) and you'll see better what happens and which brings
> what.

The key light is certainly key. I've set it up so I have a test render of
the target object available, and the other 2 lights provide little light,
but more than the recommended light mind - I thought that they needed more
of a presence.

>
> How many spheres do you have in this render ?

Durn, you've made me re-parse this monster to find out: There are ... mmm,
POV-Ray does not say how many blob components there are!
>
> JC
>
> -- 
> http://exether.free.fr/irtc (more IRTC stats !)


Post a reply to this message

From: JC (Exether)
Subject: Re: scanned blobs
Date: 12 Apr 2004 16:28:42
Message: <407afbfa@news.povray.org>
IMBJR wrote:

> "JC (Exether)" <no### [at] spamfr> wrote in message
> news:407ac7f4$1@news.povray.org...
> 
>>IMBJR wrote:
>>
>>>3-light system used, probably incorrectly tho.
>>
>>Not so bad. :-)
>>For some reason, the position for the back light that is given in
>>tutorials does not quite work in my scenes, it seems I'm not the only


>>lit areas that will let you see the shape of your spheres better.
> 
> 
> If I understand correctly, the angular relationship between the key and
> backlight is already at 135 degrees (viewed down the y-axis), since the key
> light is 45 degrees off  to the left and the backlight is behind the
> subject. It might help if I position the lights by putting them at a
> standard position and rotating them around the subject instead of my rather
> rough positioning by offsets.


with the >camera<, you should turn it around so that it lights the right 
part of the object. It would be easier working with angles, but you can 
set it up with offsets too. I posted the source of the menger sponge in 
p.t.scene-files on 18/08/2003 if you want to have a look at the 3 lights 
code.

>>Try to render your image with only one of the three lights (trying with
>>key, fill and back) and you'll see better what happens and which brings
>>what.
> The key light is certainly key. I've set it up so I have a test render of
> the target object available, and the other 2 lights provide little light,
> but more than the recommended light mind - I thought that they needed more
> of a presence.

Back light should provide as much light as the key light, because it's 
in the back it won't provide too much light.

>>How many spheres do you have in this render ?
> Durn, you've made me re-parse this monster to find out: There are ... mmm,
> POV-Ray does not say how many blob components there are!

Sorry, I was just beeing curious as your algorithm provides a lot of 
spheres for the outside, but fewer for the inside, so I thought the 
result might be interesting. You can use #debug traces to know.

JC

-- 
http://exether.free.fr/irtc (more IRTC stats !)


Post a reply to this message

From: IMBJR
Subject: Re: scanned blobs
Date: 12 Apr 2004 17:40:31
Message: <407b0ccf$1@news.povray.org>
"JC (Exether)" <no### [at] spamfr> wrote in message
news:407afbfa@news.povray.org...
> IMBJR wrote:
>
> > "JC (Exether)" <no### [at] spamfr> wrote in message
> > news:407ac7f4$1@news.povray.org...
> >
> >>IMBJR wrote:
> >>
> >>>3-light system used, probably incorrectly tho.
> >>
> >>Not so bad. :-)
> >>For some reason, the position for the back light that is given in
> >>tutorials does not quite work in my scenes, it seems I'm not the only


> >>lit areas that will let you see the shape of your spheres better.
> >
> >
> > If I understand correctly, the angular relationship between the key and
> > backlight is already at 135 degrees (viewed down the y-axis), since the
key
> > light is 45 degrees off  to the left and the backlight is behind the
> > subject. It might help if I position the lights by putting them at a
> > standard position and rotating them around the subject instead of my
rather
> > rough positioning by offsets.
>

> with the >camera<, you should turn it around so that it lights the right
> part of the object. It would be easier working with angles, but you can
> set it up with offsets too. I posted the source of the menger sponge in
> p.t.scene-files on 18/08/2003 if you want to have a look at the 3 lights
> code.

http://www.mavart.com/art-film-three-point.html
is just one of a number of sites that give yet another slight variation in
the rules.
This one agrees with my original interpretation that the back light is 180
degrees to the camera, behind the subject.

>
> >>Try to render your image with only one of the three lights (trying with
> >>key, fill and back) and you'll see better what happens and which brings
> >>what.
> > The key light is certainly key. I've set it up so I have a test render
of
> > the target object available, and the other 2 lights provide little
light,
> > but more than the recommended light mind - I thought that they needed
more
> > of a presence.
>
> Back light should provide as much light as the key light, because it's
> in the back it won't provide too much light.

The site above indicates a smaller light (than the key light) is to be used.

Blimey. Well I suppose it's not a bad thing not to lay down hard and fast
rules for first-time lighting technicians etc.

>
> >>How many spheres do you have in this render ?
> > Durn, you've made me re-parse this monster to find out: There are ...
mmm,
> > POV-Ray does not say how many blob components there are!
>
> Sorry, I was just beeing curious as your algorithm provides a lot of
> spheres for the outside, but fewer for the inside, so I thought the
> result might be interesting. You can use #debug traces to know.

I think that's probably for the future. I'm gurding up my parts to work on
something that can produce DXF output.

>
> JC
>
> -- 
> http://exether.free.fr/irtc (more IRTC stats !)


Post a reply to this message

From: JC (Exether)
Subject: Re: scanned blobs
Date: 13 Apr 2004 01:09:13
Message: <407b75f9$1@news.povray.org>
IMBJR wrote:
> "JC (Exether)" <no### [at] spamfr> wrote in message
> news:407afbfa@news.povray.org...
> 
>>IMBJR wrote:
>>
>>
>>>"JC (Exether)" <no### [at] spamfr> wrote in message
>>>news:407ac7f4$1@news.povray.org...
>>>
>>>
>>>>IMBJR wrote:
>>>>
>>>>
>>>>>3-light system used, probably incorrectly tho.
>>>>
>>>>Not so bad. :-)
>>>>For some reason, the position for the back light that is given in
>>>>tutorials does not quite work in my scenes, it seems I'm not the only


>>>>lit areas that will let you see the shape of your spheres better.
>>>
>>>
>>>If I understand correctly, the angular relationship between the key and
>>>backlight is already at 135 degrees (viewed down the y-axis), since the
> 
> key
> 
>>>light is 45 degrees off  to the left and the backlight is behind the
>>>subject. It might help if I position the lights by putting them at a
>>>standard position and rotating them around the subject instead of my
> 
> rather
> 
>>>rough positioning by offsets.
>>

>>with the >camera<, you should turn it around so that it lights the right
>>part of the object. It would be easier working with angles, but you can
>>set it up with offsets too. I posted the source of the menger sponge in
>>p.t.scene-files on 18/08/2003 if you want to have a look at the 3 lights
>>code.
> 
> 
> http://www.mavart.com/art-film-three-point.html
> is just one of a number of sites that give yet another slight variation in
> the rules.
> This one agrees with my original interpretation that the back light is 180
> degrees to the camera, behind the subject.
> 
> 
>>>>Try to render your image with only one of the three lights (trying with
>>>>key, fill and back) and you'll see better what happens and which brings
>>>>what.
>>>
>>>The key light is certainly key. I've set it up so I have a test render
> 
> of
> 
>>>the target object available, and the other 2 lights provide little
> 
> light,
> 
>>>but more than the recommended light mind - I thought that they needed
> 
> more
> 
>>>of a presence.
>>
>>Back light should provide as much light as the key light, because it's
>>in the back it won't provide too much light.
> 
> 
> The site above indicates a smaller light (than the key light) is to be used.
> 
> Blimey. Well I suppose it's not a bad thing not to lay down hard and fast
> rules for first-time lighting technicians etc.

I didn't get my best results by sticking to the rules and you're right 
that my setting are not quite academic. But well, that's what worked on 
my scene. :-)

JC

-- 
http://exether.free.fr/irtc (more IRTC stats !)


Post a reply to this message

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