|
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Hello !
I understand why systematically fixing the camera aspect ratio to the image
pixel ratio is bad, because one might want to render images for non-squares
pixel display (or one might just want to experiment).
But shouldn't the default value be set to the ratio of the image's pixels ?
Squared pixels, if not omnipresent, are at least more common than non-suared
pixels (especially with non-cathodics screens), and the current "4/3" default
ratio pretty much sounds like a magic number to me anyway.
Then again, there might be reasons I am not aware of ; please be indulgent :)
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Gyscos <gys### [at] gmailcom> wrote:
> But shouldn't the default value be set to the ratio of the image's pixels ?
No, because changing the image aspect ratio while keeping the pixel aspect
ratio changes the framing, and hence the composition of the image, in other
words, part of the image will be cropped, or previously unseen parts (which
might be unmodeled!) might come into view.
The ultimate decision of the images framing should be with the author,
not the person who renders the image. (Please skip nitpicking about the
latter being able to modify the scene if he wants. You know what I mean.
And by "you" I mean people reading this in general.)
Possible solutions to this have been presented here:
http://bugs.povray.org/task/85
> Squared pixels, if not omnipresent, are at least more common than non-suared
> pixels (especially with non-cathodics screens), and the current "4/3" default
> ratio pretty much sounds like a magic number to me anyway.
4/3 is not a magic number because it's the most common and ubiquitous
aspect ratio.
--
- Warp
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> Hello !
>
> I understand why systematically fixing the camera aspect ratio to the image
> pixel ratio is bad, because one might want to render images for non-squares
> pixel display (or one might just want to experiment).
>
> But shouldn't the default value be set to the ratio of the image's pixels ?
> Squared pixels, if not omnipresent, are at least more common than non-suared
> pixels (especially with non-cathodics screens), and the current "4/3" default
> ratio pretty much sounds like a magic number to me anyway.
>
> Then again, there might be reasons I am not aware of ; please be indulgent :)
>
>
>
4/3 is not magic. It's the effective aspect ratio of just about all CRT
monitors. It have been almost universal since the mid 70's, at least for
the PC.
Only with the introduction of LCDs this has changed. You now commonly
have displays with 5:4 (1280 x 1024) and 16:9 aspect ratio.
Alain
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Alain <aze### [at] qwertyorg> wrote:
> > Hello !
> >
> > I understand why systematically fixing the camera aspect ratio to the image
> > pixel ratio is bad, because one might want to render images for non-squares
> > pixel display (or one might just want to experiment).
> >
> > But shouldn't the default value be set to the ratio of the image's pixels ?
> > Squared pixels, if not omnipresent, are at least more common than non-suared
> > pixels (especially with non-cathodics screens), and the current "4/3" default
> > ratio pretty much sounds like a magic number to me anyway.
> >
> > Then again, there might be reasons I am not aware of ; please be indulgent :)
> >
> >
> >
> 4/3 is not magic. It's the effective aspect ratio of just about all CRT
> monitors. It have been almost universal since the mid 70's, at least for
> the PC.
>
> Only with the introduction of LCDs this has changed. You now commonly
> have displays with 5:4 (1280 x 1024) and 16:9 aspect ratio.
4:3 is the oldest film standard (from silent films), and it pretty close to
Academy Ratio (which was the standard for talking pictures until the widescreen
formats started showing up in the 50s).
Cheers,
Edouard.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Alain <aze### [at] qwertyorg> wrote:
> >
> > pixels (especially with non-cathodics screens), and the current "4/3" default
> > ratio pretty much sounds like a magic number to me anyway.
> >
> > Then again, there might be reasons I am not aware of ; please be indulgent :)
> >
> >
> >
> 4/3 is not magic. It's the effective aspect ratio of just about all CRT
> monitors. It have been almost universal since the mid 70's, at least for
> the PC.
>
> Only with the introduction of LCDs this has changed. You now commonly
> have displays with 5:4 (1280 x 1024) and 16:9 aspect ratio.
Indeed. Not just on tv or pc monitors, but also cell phones. 4x3 is very old
hat.
not unlike povray, of course... :p
this posted from a cell phone, BTW...
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> 4/3 is not a magic number because it's the most common and ubiquitous
> aspect ratio.
If that's true, it won't be true for very much longer...
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
I know, and understand that the ability to control the aspect ratio is
important, but I'm talking about the default behavior here :
If someone creates a scene, and wants only a small part of it to be visible,
then explicitly entering the desired aspect ratio in the camera makes more sense
than relying on the default settings. On the other hand, for quick and short
scene, it is much more practical to have the default settings adapt to the image
size.
Also, the 4:3 ratio is really not so frequent, especially when using POV-Ray to
render images to be used by another program (which is its main use in the lab
I'm working for).
I completely agree with the link about aspect ratio issue, and I think a
command-line argument for pixel ratio could be an elegant idea to easily fix
anamorphic issues. I just don't understand the default setting.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Am 17.01.2011 21:14, schrieb Gyscos:
> I know, and understand that the ability to control the aspect ratio is
> important, but I'm talking about the default behavior here :
> If someone creates a scene, and wants only a small part of it to be visible,
> then explicitly entering the desired aspect ratio in the camera makes more sense
> than relying on the default settings. On the other hand, for quick and short
> scene, it is much more practical to have the default settings adapt to the image
> size.
>
> Also, the 4:3 ratio is really not so frequent, especially when using POV-Ray to
> render images to be used by another program (which is its main use in the lab
> I'm working for).
>
> I completely agree with the link about aspect ratio issue, and I think a
> command-line argument for pixel ratio could be an elegant idea to easily fix
> anamorphic issues. I just don't understand the default setting.
The default setting has been around probably since the days POV-Ray was
still named DKB-trace, and back then the setting was just taken for
granted because /every/ computer display (except for a few extremely
exotic ones) used a 4:3 aspect ratio - even when the number of pixels
would nowadays seem to imply otherwise (e.g. 320x200 CGA, 640x350 EGA,
320x200 MCGA, or even 1280x1024 SXGA). Back then, non-square pixels were
actually quite common, whereas non-4:3 aspect ratios were rare exceptions.
The reason that this default has never been changed is pretty simple:
Backward compatibility.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Am 18.01.2011 09:23, schrieb clipka:
> The default setting has been around probably since the days POV-Ray was
> still named DKB-trace, and back then the setting was just taken for
> granted because /every/ computer display (except for a few extremely
> exotic ones) used a 4:3 aspect ratio - even when the number of pixels
> would nowadays seem to imply otherwise (e.g. 320x200 CGA, 640x350 EGA,
> 320x200 MCGA, or even 1280x1024 SXGA). Back then, non-square pixels were
> actually quite common, whereas non-4:3 aspect ratios were rare exceptions.
Ah, and of course not to forget all the various home computers used in
those days; after all, DKB-trace didn't even start as a DOS program.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
|
|