i finished some tv(pal) stuff recently and found shading at 4-6 and
pixels at 6-8 sufficient for my kind of work. it was mainly detailed
textures on maps, that neede to be animated with slow camera moves.
so the biggest issue is to get crawling and flickering in the fields
under control.
these numbers up there and a gauss filter did the job very nicely
and MUCH faster than the normal renderer...
ciao
franky
Thursday, March 1, 2007, 10:58:09 PM, you wrote:
>> start with the shading samples around 2 or 3. Then I render out a test.
>>I then raise it up to around 6 and do a test.
gl> Rasterizer render is very interesting for objects with lot of sharp
gl> geometric details (like long vertical or horizontal lines create by shadows
gl> on a building for example).
gl> In this case you must use more samples like Joe said.
gl> In my project, I used a very low sampling because it was only an organic
gl> subdivided character with a very simple color and bump map.
gl> On 3/1/07, Joe Laffey <joe(at)laffey.tv> wrote:
>>
>> >>
>> >> I'm very new to the rasterizer. Lots of hair to render, so I wanted to
>> >> get familiar with it. Now I was wondering what pixel samples value
>> >> people usually do for good old TV output (for now it's only going to a
>> >> turn-table, no light movement and no character movement).
>> >>
>>
>> I start with the shading samples around 2 or 3. Then I render out a test.
>> I then raise it up to around 6 and do a test. If I see a big difference,
>> and I prefer the higher level I use it (or in between). But shading
>> samples take a lot more time to render. So you want this as low as
>> possible. But if you use detailed shaders, or bumps you will see a big
>> difference in detail between low and high samples. Rarely do I go over 8
>> or so for shading samples.
>>
>>
>> Pixel samples have much less of an effect on render time. I frequently
>> have these at 10-30. It depends on the fine details of the model. A finely
>> detailed moel, like a ship with rope rigging, or a tree with leaves can
>> require fairly high pixel samples. With simpler geometry you could keep
>> pixel samples pretty low like 5-10.
>>
>> Note that I am stickler for seeing NO crawlies or jaggies... So my numbers
>> may be biased on the high side...
>>
>> --
>> Joe Laffey | Visual Effects for Film and Video
>> LAFFEY Computer Imaging | -------------------------------------
>> St. Louis, MO | Show Reel http://LAFFEY.tv/?e04853
>> USA | -------------------------------------
>> . | -*- Digital Fusion Plugins -*-
>> --------------------------------------------------------------------------
>> Mail here will be rejected --> "Real Trap" <r_trap(at)laffeycomputer.com>
>> ---
>> Unsubscribe? Mail Majordomo(at)Softimage.COM with the following text in body:
>> unsubscribe xsi
>>
---
Unsubscribe? Mail Majordomo(at)Softimage.COM with the following text in body:
unsubscribe xsi
