Hi,

I was playing around with the output dimensions in QUICKRES.INI and ran
a few tests which brought to light a few questions:

1)   Since it looks as if the same amount of information (as far as
camera angle of the scene) is being put into the output file no matter
what pixel dimensions you specify (as in, this has no effect on the
cropping or changing the camera angle in your scene), if you change from
the defaults to a more vertical or horizontal shape, why is my image
being skewed?

2)  Using the answer to question 1) as a rule of thumb, what pixel
dimension ratios should I use for the most accurate sampling (not at all
skewed from what it should look like) of the scene I have input to
POV-Ray?  Would it be the defaults, 512x384, 640x480, 800x600, etc.
(which are based on standard screen resolution which, of course, is
messed up anyway [a square on your computer screen does not look like a
square]) or a square output from POV-Ray, which wasn't all that bad when
I tried it.  Would this be at all accurate?

ANY input or explanation on the math would be useful.

Thank you.

- CK

Post a reply to this message

Short answer would be to look into using the aspect ratio keywords, 'right'
and 'up'.  When you tie those in with your resolutions you will get
non-skewed images.
Far as a long answer goes, I couldn't say.  However, the math is as easy as
dividing the width by the height, i.e. 640/480=1.33333333 (or 4/3), so that
right 4/3*x and up 1*y make it match the resolution chosen.  Any changes to
resolution would need the ratios of right to up changed in the camera as
well.

Bob H.

"Chad Kruger" <ckr### [at] telusplanetnet> wrote in message
news:3B5782D9.424CC774@telusplanet.net...
>
> I was playing around with the output dimensions in QUICKRES.INI and ran
> a few tests which brought to light a few questions:
>
> 1)   Since it looks as if the same amount of information (as far as
> camera angle of the scene) is being put into the output file no matter
> what pixel dimensions you specify (as in, this has no effect on the
> cropping or changing the camera angle in your scene), if you change from
> the defaults to a more vertical or horizontal shape, why is my image
> being skewed?
>
> 2)  Using the answer to question 1) as a rule of thumb, what pixel
> dimension ratios should I use for the most accurate sampling (not at all
> skewed from what it should look like) of the scene I have input to
> POV-Ray?  Would it be the defaults, 512x384, 640x480, 800x600, etc.
> (which are based on standard screen resolution which, of course, is
> messed up anyway [a square on your computer screen does not look like a
> square]) or a square output from POV-Ray, which wasn't all that bad when
> I tried it.  Would this be at all accurate?
>
> ANY input or explanation on the math would be useful.

Post a reply to this message

POV's default is to make an image (camera-angle-wise) in the 4:3 ratio. 
Actually in the standard resolutions the pixels are pretty close to
square, or at least they should be.  The only non-4:3 "standard"
resolution is 1280x1024, which has much worse distortion.

POV doesn't have to render square pixels though.  You can set whatever
image size you want in the command-line (Alt+C in Windows) with "+w..."
and "+h...", and you change the actual camera angles with "up <x,y,z>",
"right <x,y,z>" and "angle n" in the camera block.  "angle" specifies
the angle between the camera's axis and any point one unit away from
that axis on the projection plane.  "up" and "right" specify the vectors
up and to the right of the camera's axis (these three axes must be
perpendicular), and their length specifies how much of the projection
plane is visible in the rendered image.  There is also a "direction
<x,y,z>" keyword that sets the camera's axis and moves the projection
plane depending on the vector length, but I just use look_at.  See the
docs for a more clear explanation.

-- 
David Fontaine  <dav### [at] faricynet>  ICQ 55354965
My raytracing gallery:  http://davidf.faricy.net/

Post a reply to this message

(Sometimes I feel like making work in vain by maintining the VFAQ because
people seldom use it when answering these kind of questions even when the
answer is directly there, but they write their own answers... *sigh*)

http://www.students.tut.fi/~warp/povVFAQ/languageVFAQ.html#aspectratio

-- 
#macro N(D,I)#if(I<6)cylinder{M()#local D[I]=div(D[I],104);M().5,2pigment{
rgb M()}}N(D,(D[I]>99?I:I+1))#end#end#macro M()<mod(D[I],13)-6,mod(div(D[I
],13),8)-3,10>#end blob{N(array[6]{11117333955,
7382340,3358,3900569407,970,4254934330},0)}//                     - Warp -

Post a reply to this message

Warp wrote:
> 
>   (Sometimes I feel like making work in vain by maintining the VFAQ because
> people seldom use it when answering these kind of questions even when the
> answer is directly there, but they write their own answers... *sigh*)
> 
> http://www.students.tut.fi/~warp/povVFAQ/languageVFAQ.html#aspectratio

Oops.  Well, I know that stuff, and I kinda forget about it.  It's in
the bookmarks though!  Hey, I just went and added it to my links page
too.

-- 
David Fontaine  <dav### [at] faricynet>  ICQ 55354965
My raytracing gallery:  http://davidf.faricy.net/

Post a reply to this message