 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Everyone acknowledges the benefits of AA in a still picture, but what
about when I'm making an animation and its going to be put into a very
low bitrate MPEG1? Is there any benefit at all at still using AA? Do
the smoother edges compress better? Or does it depend on the scene?
The scene in question would be the roller coaster I've posted to
povray.binaries.animations.
-DJ
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
DJ Wiza wrote:
> Everyone acknowledges the benefits of AA in a still picture, but what
> about when I'm making an animation and its going to be put into a very
> low bitrate MPEG1? Is there any benefit at all at still using AA? Do
> the smoother edges compress better?
As a mater of fact, they do! the math behind this effect that causes the
"better" compression is of course rather complicated. Yet, the simpler
explanation for this "better" compression is that there is less detail in
the picture because high contrast edges simply have a lot of geometry detail
hardly found in real-world images. Yet, real world images are what lossy
compression algorithms such as MPEG and JPEG have been designed for. They
specifically build on the perception of the human eye by eliminating what
the viewer is less likely to "miss".
Thorsten
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
It depends on the scene. For instance I've found that in an animation with
a star-field background the stars flicker as the aa from frame to frame
either allows them to appear or blanks them out.
--
#####-----#####-----#####
POV Tips and Hints at ...
http://povman.blogspot.com/
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Thorsten Froehlich <tho### [at] trf de> wrote:
> As a mater of fact, they do! the math behind this effect that causes the
> "better" compression is of course rather complicated. Yet, the simpler
> explanation for this "better" compression is that there is less detail in
> the picture because high contrast edges simply have a lot of geometry detail
> hardly found in real-world images.
Curiously, it seems that at least in MPEG-4 (and perhaps even MPEG-2) they
dedicated more data to the higher frequencies of the spectrum (iow to the
parts of the image with higher contrast) than to the lower frequencies.
Thus in a low-bitrate MPEG-4 you are more likely to see artifacts in areas
where there are smooth color transitions than in areas with sharp edges...
This makes sense, though. With a proper bitrate you get a higher image
quality because sharp edges are sharp and mostly artifact-free.
--
- Warp
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Btw, forgot to mention that this doesn't mean that AA shouldn't be used
if encoding to MPEG-4. Of course it should be used. The quality of the
result will only improve. (After all, antialiasing is not really a "smooth
color transition".)
--
- Warp
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"POVMAN" <s### [at] acom> wrote:
> It depends on the scene. For instance I've found that in an animation with
> a star-field background the stars flicker as the aa from frame to frame
> either allows them to appear or blanks them out.
>
The best solution I have found to mimic a star-field for animations. Is to
create a lot of spheres distributed randomly on the surface of a Great
Sphere. The Star spheres should have a randomised radius and brightness to
my laptop with me.
Stephen
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
DJ Wiza wrote:
> Everyone acknowledges the benefits of AA in a still picture, but what
> about when I'm making an animation and its going to be put into a very
> low bitrate MPEG1? Is there any benefit at all at still using AA? Do
> the smoother edges compress better?
Absofrigginglutely, smooth edges compress better than non-antialiased.
> The scene in question would be the roller coaster I've posted to
> povray.binaries.animations.
AA, all the way.
Regards,
John
(who speaks from experience)
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
