Hi Juan,
Those samples look quite amazing! At first I found it a little hard to
believe that current anti-aliasing (AA) techniques looked so poorly, so
I took a quick look at the aliasing in a bunch of different programs:
First, I opened your sample .svg picture of a tiger in Firefox,
Illustrator CS4, Adobe Reader 9 and Corel Paint Shop Pro X; none of them
seem to come close to your Morphic 3 solution. However, to my knowledge,
none of these programs are hardware-accelerated, meaning the CPU is
doing all the hard work. I wanted to see whether my graphics card could
do any better, so I booted up a couple of games (Crysis and Supreme
Commander :) ). You can still see some aliasing at 2x and 4x
supersampling AA, but once you crank it up to 8x or higher, there is
very little aliasing left, if anything at all. The same goes for the two
3D modeling tools I've tried (3ds Max 2010 and Blender 2.49b), which are
hardware-accelerated as well. I don't know much about anti-aliasing
(well, pretty much nothing until after I read your e-mail :) ), but I
don't think these results are very surprising: You just use the GPU's
immense horsepower to supersample your image at such a high resolution
until you surpass, or come close to, the image's Nyquist frequency.
For this reason, I'm mainly interested in how Morphic 3 compares to
current AA techniques performance-wise. I think all graphics-related
tasks, not just 3D rendering tasks, will be handled by the GPU in the
(hopefully) near-future anyway. (As an example of this future direction,
Photoshop already supports hardware acceleration as of version CS4.)
However, even with all of the GPU's power, rendering images at 8x
supersampling or higher is an immense performance-drain, so it would be
very nice if Morphic 3's AA has much better performance. If so, it would
be even better if Morphic 3 eventually catches the industry's attention,
because I started paying attention to the aliasing in every image I see
since I saw your samples and it's becoming quite bothersome. :)
Cheers,
Tim Molderez
PS: To demonstrate how powerful a GPU really is, the image processing
research lab at our university has turned a plain old desktop computer
into a full-blown supercomputer by simply throwing in 4 graphics cards:
http://fastra.ua.ac.be I'm just mentioning it because their
demonstration video is such a fun watch. :)
At 2/06 16:15, Juan Vuletich wrote:
Hi Folks,
I am developing a novel way to do anti aliased 2d graphics that breaks
away from pixel coverage and super sampling, while being simpler and
providing higher quality. Please take a look at http:www.jvuletich.org
, especially at the samples.
I believe this work fits nicely in the FONC project, as a Nile
implementation of it would yield better quality results than Gezira.
Comments welcome.
Cheers,
Juan Vuletich
_______________________________________________
fonc mailing list
[email protected]
http://vpri.org/mailman/listinfo/fonc
