See thread:

http://news.povray.org/povray.advanced-users/thread/%3C5a7f9271%241%40news.povray.org%3E/


I think it would be more intuitive than the current behavior if 
focal_point were to scale along with the rest of the camera. E.g., if 
the original focal_point is <3,4,5>, then scaling the camera by 3 should 
move the focal_point to <9,12,15>. The current existing behavior is a 
bit weird IMO.


Mike

Post a reply to this message

Am 12.02.2018 um 06:59 schrieb Mike Horvath:
> See thread:
> 
>
http://news.povray.org/povray.advanced-users/thread/%3C5a7f9271%241%40news.povray.org%3E/
> 
> 
> 
> I think it would be more intuitive than the current behavior if
> focal_point were to scale along with the rest of the camera. E.g., if
> the original focal_point is <3,4,5>, then scaling the camera by 3 should
> move the focal_point to <9,12,15>. The current existing behavior is a
> bit weird IMO.

I agree.

We should also consider scaling the aperture size along the way.

Maybe all depending on the order in which camera statements are done. E.g.

    camera_location <0,0,0>
    focal_point <1,2,3>
    scale 2
    aperture 10

should scale the focal point but not the aperture size.


Just brainstorming here ATM.

Post a reply to this message

On 2/12/2018 5:29 AM, clipka wrote:
> Am 12.02.2018 um 06:59 schrieb Mike Horvath:
>> See thread:
>>
>>
http://news.povray.org/povray.advanced-users/thread/%3C5a7f9271%241%40news.povray.org%3E/
>>
>>
>>
>> I think it would be more intuitive than the current behavior if
>> focal_point were to scale along with the rest of the camera. E.g., if
>> the original focal_point is <3,4,5>, then scaling the camera by 3 should
>> move the focal_point to <9,12,15>. The current existing behavior is a
>> bit weird IMO.
> 
> I agree.
> 
> We should also consider scaling the aperture size along the way.
> 
> Maybe all depending on the order in which camera statements are done. E.g.
> 
>      camera_location <0,0,0>
>      focal_point <1,2,3>
>      scale 2
>      aperture 10
> 
> should scale the focal point but not the aperture size.
> 
> 
> Just brainstorming here ATM.
> 

I'm not sure aperture needs to be scaled, depending on what you're 
trying to achieve.

This is hard for me to explain.

Let's say you have an aperture value of 4, and (I'm making this up) 50% 
of the image is blurred. If you scale the camera by 3 but leave the 
aperture at 4, then then the resulting image will show a larger area, 
but still be 50% blurred.

If you instead also scale the aperture by 3, then the image will be more 
% blurred, but those portions of the image that were blurred before you 
scaled the camera will remain as blurred as before.

E.g. if a box in the original image was 25% blurred, then the box will 
shrink in area in the second image, but will itself remain 25% blurred.

So it depends on the effect you want to achieve.


Mike

Post a reply to this message

I also tested camera translation, and it seems the focal point is not 
translated along with the camera. The focal point stays at the location 
it was originally set to occupy.

Mike


On 2/12/2018 5:29 AM, clipka wrote:
> Am 12.02.2018 um 06:59 schrieb Mike Horvath:
>> See thread:
>>
>>
http://news.povray.org/povray.advanced-users/thread/%3C5a7f9271%241%40news.povray.org%3E/
>>
>>
>>
>> I think it would be more intuitive than the current behavior if
>> focal_point were to scale along with the rest of the camera. E.g., if
>> the original focal_point is <3,4,5>, then scaling the camera by 3 should
>> move the focal_point to <9,12,15>. The current existing behavior is a
>> bit weird IMO.
> 
> I agree.
> 
> We should also consider scaling the aperture size along the way.
> 
> Maybe all depending on the order in which camera statements are done. E.g.
> 
>      camera_location <0,0,0>
>      focal_point <1,2,3>
>      scale 2
>      aperture 10
> 
> should scale the focal point but not the aperture size.
> 
> 
> Just brainstorming here ATM.
>

Post a reply to this message

I did not test what happens when the camera is rotated. Using just my 
eyeballs alone, I don't think I would be able to catch an error if it 
occurred.

Mike

Post a reply to this message