What's the difference between "normal" and "superfine" antialias (I use Bryce4)? I think the result look the same whichever option I use.

If you use it with a very sharped object or a complex tree, you will notice de difference, anyway, the mayor diferrence is the render time.-

In B5 using superfine you can set the "rays per pixel" from 4 up to 256 in the render options. Normal will give you default setting of 1. Not sure if version 4 is the same, as I started with version 5. It depends on the scene and materials used to notice a differance between the two. I remember there was a good explanation given to this quite a while back with some screen shots etc, try doing a search in the Forum, I believe it was Agent Smith who offered up the explanation.

There was a great image posted here a few weeks ago that demoed superfine AA at different levels with complex textures dropping into the background it made a huge difference.

Try it with a plaid type texture and a cube fading off into the distance (ie one end close to camera other end off into the distance) and render with the different settings. You will see the normal AA to be jagged or even some lines missing in places while the superfine will be more detailed.

See objects with many paralel(I dont know if it is correct English) lines, like a fence into the distance, or anything like tghat.

With Anti-Aliasing Bryce raytraces the scene to the pixel level then runs a 2D process on the resultant image which blurs sharp color transitions (kinda like Photoshop's Unsharp Mask). This is a way to speed up the rendering process but it also can blur detail and fine lines out of existence. Superfine is a different form of ray tracing called Distributed Ray Tracing. Since the scene has more internal data then screen/pixel resolution, DRT ray traces a block of data around the pixel and does a type of averaging to come up with the final pixel value. Since it's entirely 3D scene data (no 2D AA), it retains detail and fine lines much more accurately but at the cost of a slower render. It also improves the rendering of distant objects. HTH; Dan

dan whiteside , wow.

double wow.

A huge downside of using standard AA lies in artifacts. On objects that are thin (such as my grass, etc.) you will notice "breaks" in the model, where it seemingly misses portions of the actual model. This is caused by the per pixel preimise that dan described. If the object does not appear in two concurrent pixels, it is basically averaged out of the image. This is a problem, and often leads to work that smacks of 'bryce' or 'low quality.' I'll take the longer renders anyday.

Yup, lol, what Dan says. The example I spew out is say, distant blades of grass, or also the multitude of ropes seen on a Pirate-type ship, these are the types of smaller/fine objects, (that could be in the midground to background), that you would want to make sure they showed up well. With a superfine aa these kinds of objects will be more readily distinguishable. A Superfine aa will also help distinguish some textures as they are farther away. But, really the best attribute of the Superfine aa is to bring out those smaller/finer mesh parts in the distance. *If you can...just plop render the parts of your render that would benefit from Superfine aa. This way you can render as normal (faster), then just plop render the smaller areas that may need the superfine aa. AgentSmith

OK! I've got it! thanks!

Thanks AS that was the pic I was thinking of :-)

One other aspect of the extra rays is when using premium effects such as soft shadows and blurry relections or DOF. The effects of lower ray numbers (or oversample rate) are a serious graininess in what should be a smooth visual effect. 36 (or 6x6 oversample) may work depending on the individual image but 64 (8x8) is definitely better for premium effects as a rule of thumb.