POV-Ray : Newsgroups : povray.general : Light amplification and artifact. Server Time
26 Nov 2024 20:53:54 EST (-0500)
  Light amplification and artifact. (Message 1 to 3 of 3)  
From: John Coppens
Subject: Light amplification and artifact.
Date: 28 Dec 2009 07:28:15
Message: <20091228092814.7e3a9811.john@johncoppens.com>
Hello guys and gals,

I was playing around with a simple scene for a Holiday card, and have two
problems. This years' focus was on trying glasses and contents as
believable as possible.

The scene is a PNG background image of some fireworks, and two glasses in
the front. The PNG is defined with an ambient light of 1, there is no
other light source in the scene.

Strangely, the image of the PNG coming through the glass seems to be
amplifying: Without the glasses, the image shows fine, through the
glass, the image saturates completely!

And the second problem is an artifact appearing in the middle line (row)
of pixels. This artifact seems to depend on the size of the image. In the
image I uploaded, it's not very visible (right side of each glass). In a
1200x900 version, it runs over the entire width (I'll render another one
of these later).

The output images are:

http://jcoppens.com/misc/with_glasses.png
http://jcoppens.com/misc/no_glasses.png

I'm running POV-ray 3.7beta34, on Linux.

John

PS: I still have to set ambient of the liquid to 0...


Post a reply to this message

From: Alain
Subject: Re: Light amplification and artifact.
Date: 28 Dec 2009 12:59:15
Message: <4b38f1f3$1@news.povray.org>

> Hello guys and gals,
>
> I was playing around with a simple scene for a Holiday card, and have two
> problems. This years' focus was on trying glasses and contents as
> believable as possible.
>
> The scene is a PNG background image of some fireworks, and two glasses in
> the front. The PNG is defined with an ambient light of 1, there is no
> other light source in the scene.
>
> Strangely, the image of the PNG coming through the glass seems to be
> amplifying: Without the glasses, the image shows fine, through the
> glass, the image saturates completely!
>
> And the second problem is an artifact appearing in the middle line (row)
> of pixels. This artifact seems to depend on the size of the image. In the
> image I uploaded, it's not very visible (right side of each glass). In a
> 1200x900 version, it runs over the entire width (I'll render another one
> of these later).
>
> The output images are:
>
> http://jcoppens.com/misc/with_glasses.png
> http://jcoppens.com/misc/no_glasses.png
>
> I'm running POV-ray 3.7beta34, on Linux.
>
> John
>
> PS: I still have to set ambient of the liquid to 0...

Do your glass pigment have both filter and transmit?
Do they add up to more than 1?

Normaly, the sum of transmit and filter must total at most 1, usualy to 
a value slightly less than that, like filter + transmit <= 0.999. If it 
adds up to more than 1, it will have an amplification effect. As you 
cross 4 surfaces as you trverse the glass, the amplification will appens 
4 times. So, even a value just slightly over 1 can have a very 
niticeable effect.

Is the colour part larger than 1?

If it's larger than 1, it will also have an amplification effect.

Try a pigment of rgbt 1 or rgbf 1 for your glass and see what it does.
You can also try rgbft<1,1,1,0.5,0.5>

The ambient part for a transparent object is usualy barely dicernible, 
if at all.



Alain


Post a reply to this message

From: John Coppens
Subject: Re: Light amplification and artifact.
Date: 28 Dec 2009 21:01:56
Message: <20091228230155.8235a138.john@johncoppens.com>
On Mon, 28 Dec 2009 12:59:15 -0500
Alain <aze### [at] qwertyorg> wrote:

> Do your glass pigment have both filter and transmit?
> Do they add up to more than 1?

No, transmit is 0, filter is 1
 
> Normaly, the sum of transmit and filter must total at most 1, usualy to 
> a value slightly less than that, like filter + transmit <= 0.999. If it 
> adds up to more than 1, it will have an amplification effect. As you 
> cross 4 surfaces as you trverse the glass, the amplification will
> appens 4 times. So, even a value just slightly over 1 can have a very 
> niticeable effect.
> 
> Is the colour part larger than 1?

No. Color is 1
 
> If it's larger than 1, it will also have an amplification effect.
> 
> Try a pigment of rgbt 1 or rgbf 1 for your glass and see what it does.
> You can also try rgbft<1,1,1,0.5,0.5>
> 
> The ambient part for a transparent object is usualy barely dicernible, 
> if at all.

Yes, I set the ambient value to 0. The amplification was still there.

I just found a solution, but I don't really know why it is one. I'll
investigate a bit more:

I have enabled Fresnel reflection now, and the amplification disappeared
completely! The image is now quite acceptable.

Thanks for the suggestions, Alain! And best wishes for the holidays.
John


Post a reply to this message

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