Hi again,

It would appear from the online doc
(http://www.povray.org/documentation/view/3.6.1/223/) that jittering is not
used when antialiasing is turned off: "If anti-aliasing is not used one
sample per pixel is taken regardless of the super-sampling method
specified."

Is this a correct interpretation of the text?

Are there any random sampling algorithms used when antialiasing is turned
off?

cheers

Post a reply to this message

dragonmage wrote:
> Hi again,
> 
> It would appear from the online doc
> (http://www.povray.org/documentation/view/3.6.1/223/) that jittering is not
> used when antialiasing is turned off: "If anti-aliasing is not used one
> sample per pixel is taken regardless of the super-sampling method
> specified."
> 
> Is this a correct interpretation of the text?

I don't believe so; rather, that single ray should be randomly perturbed 
due to the jitter.

...Chambers

Post a reply to this message

dragonmage <dra### [at] gmailcom> wrote:
> It would appear from the online doc
> (http://www.povray.org/documentation/view/3.6.1/223/) that jittering is not
> used when antialiasing is turned off: "If anti-aliasing is not used one
> sample per pixel is taken regardless of the super-sampling method
> specified."

> Is this a correct interpretation of the text?

> Are there any random sampling algorithms used when antialiasing is turned
> off?

  Just to make sure that you don't have a confusion about this:
Antialiasing jittering is only done inside a pixel for the extra rays
for the antialiasing. It's not randomized raytracing as in "send rays
in random directions".

  If you randomize the rays more than a pixel, you'll just get a noisy
image (in fact, even if you randomize inside a pixel but without antialising
you will also get noise).
  You can actually simulate what it looks like by adding something like
this at the end of your camera definition block:

normal { bumps .01 scale 1e-5 }

-- 
                                                          - Warp

Post a reply to this message

Warp <war### [at] tagpovrayorg> wrote:
> dragonmage <dra### [at] gmailcom> wrote:
> > It would appear from the online doc
> > (http://www.povray.org/documentation/view/3.6.1/223/) that jittering is not
> > used when antialiasing is turned off: "If anti-aliasing is not used one
> > sample per pixel is taken regardless of the super-sampling method
> > specified."
>
> > Is this a correct interpretation of the text?
>
> > Are there any random sampling algorithms used when antialiasing is turned
> > off?
>
>   Just to make sure that you don't have a confusion about this:
> Antialiasing jittering is only done inside a pixel for the extra rays
> for the antialiasing. It's not randomized raytracing as in "send rays
> in random directions".
>
>   If you randomize the rays more than a pixel, you'll just get a noisy
> image (in fact, even if you randomize inside a pixel but without antialising
> you will also get noise).
>   You can actually simulate what it looks like by adding something like
> this at the end of your camera definition block:
>
> normal { bumps .01 scale 1e-5 }
>
> --
>                                                           - Warp

ok, so jitter won't "move" a pixel but may change its rendered colour?
I have "    jitter 0" in the global settings, photons anyway so, according
to the manual there will be no jitter subsampling.

thanks

Post a reply to this message

dragonmage wrote:

> Are there any random sampling algorithms used when antialiasing is turned
> off?

Pragmatically, try rendering your scene twice and flip between the
two renders in your favorite viewer. Even a slightly changed pixel
will attract your attention due to the way human vision works ;)

Post a reply to this message

dragonmage nous apporta ses lumieres en ce 19-03-2007 22:20:
> Warp <war### [at] tagpovrayorg> wrote:
>> dragonmage <dra### [at] gmailcom> wrote:
>>> It would appear from the online doc
>>> (http://www.povray.org/documentation/view/3.6.1/223/) that jittering is not
>>> used when antialiasing is turned off: "If anti-aliasing is not used one
>>> sample per pixel is taken regardless of the super-sampling method
>>> specified."
>>> Is this a correct interpretation of the text?
>>> Are there any random sampling algorithms used when antialiasing is turned
>>> off?
>>   Just to make sure that you don't have a confusion about this:
>> Antialiasing jittering is only done inside a pixel for the extra rays
>> for the antialiasing. It's not randomized raytracing as in "send rays
>> in random directions".

>>   If you randomize the rays more than a pixel, you'll just get a noisy
>> image (in fact, even if you randomize inside a pixel but without antialising
>> you will also get noise).
>>   You can actually simulate what it looks like by adding something like
>> this at the end of your camera definition block:

>> normal { bumps .01 scale 1e-5 }

>> --
>>                                                           - Warp

> ok, so jitter won't "move" a pixel but may change its rendered colour?
> I have "    jitter 0" in the global settings, photons anyway so, according
> to the manual there will be no jitter subsampling.

> thanks



There will be no jittering in the subsampling. All subsamples will be taken in a 
nice, regular array of points.

-- 
Alain
-------------------------------------------------
I'm interested in the fact that the less secure a man is, the more likely he is 
to have extreme prejudice. (Clint Eastwood)

Post a reply to this message

....
>
> >> --
> >>                                                           - Warp
>
> > ok, so jitter won't "move" a pixel but may change its rendered colour?
> > I have "    jitter 0" in the global settings, photons anyway so, according
> > to the manual there will be no jitter subsampling.
>
> > thanks
>
>
>
> There will be no jittering in the subsampling. All subsamples will be taken in a
> nice, regular array of points.
ok, thanks,

that at least explains how jittering works - I'll see if the programming ppl
can point me at the code that does it,

chris
>
> --
> Alain
> -------------------------------------------------
> I'm interested in the fact that the less secure a man is, the more likely he is
> to have extreme prejudice. (Clint Eastwood)

Post a reply to this message