Hair and FG, any point trying? 2005-03-16 - By Oz Adi
Back
yes, I am testing non casting shadow, arealights, right now... it looks
good!
A friend who worked in the lighting & texturing dept. on a certain studio,
told me, they had a tool, so they could have non casting shadows lights,
light the scene,
and another some kind of directional object, that only "emits" shadows..
so the shadow casting direction is not "connected" to the direction the
lights hit the object....
is it possible to make something similar in XSI? (they were using
maya&render man)
regarding raytrace... I got good results... but it flickered pretty bad..
even with very high AA..
volumic shadow maps+Rapid motion and AA of 0,2,3 gaussian or triangle, are
almost
flicker free! (tested a sphere covered with 600,00 hairs, and some wiggle
animation)
but they are memory hogs.. and cause my XP systems to crash alot..
On linux, it's rendering fine.
cheers,
Oz
-- -- Original Message -- --
From: "kim aldis" <kim@(protected)>
To: <XSI@(protected)>
Sent: Wednesday, March 16, 2005 7:23 PM
Subject: RE: Hair and FG, any point trying?
> Actually, there's not much difference in timing between Point raytraced
> and
> volumic shadow maps in the tests I've been doing here. Area lights only
> really kill if you cast shadows from them. If you use them without shadows
> they're not too bad.
>
>> -- --Original Message-- --
>> From: owner-xsi@(protected)
>> [mailto:owner-xsi@(protected)] On Behalf Of
>> softimage@(protected)
>> Sent: 15 March 2005 17:02
>> To: XSI@(protected)
>> Subject: Re: Hair and FG, any point trying?
>>
>> Raytraced will be quite slower than shadow maps. Even worse
>> with area lights.
>> Unless you have access to Pixar's farm, you don't have any
>> other option than detailed shadow maps if you ever want to
>> get the job done... nicely.
>>
>> Also, I'd consider using dirtmap on hairs instead of final
>> gathering. You will get a lot less flickering than with FG.
>> We did that a lot on Magic Roundabout.
>> It did flickered, but rendering the dirtmap pass on a single
>> computer can help.
>>
>>
>> My few cents
>> Bernard
>>
>>
>> Quoting Oz Adi <oz@(protected)>:
>>
>> > I hope the list didnt get this mail 3 times...
>> > the list server bounced my replies... last try..
>> >
>> > :) add me to the failure list..
>> >
>> > do you think area lights might be better and/or quicker
>> than detailed
>> > = shadow maps?
>> >
>> >
>> > back to traditional lighting...
>> >
>> > Thanks!
>> >
>> > Oz.
>> > www.coolcalb.com
>> >
>> > -- -- Original Message -- --
>> > From: Raffaele "ThE_JacO" Fragapane
>> > To: XSI@(protected)
>> > Sent: Tuesday, March 15, 2005 1:03 PM
>> > Subject: RE: Hair and FG, any point trying?
>> >
>> >
>> > Seen quite a few people trying it, seen all of them failing.
>> >
>> > Giving some thought at how FG works I can see why all hell breaks
>> > loose, not to mention that MI_Hair primitives are not tubular
>> > geometry, so the way landmark samples happen on them is radically
>> > different everytime the camera moves, making caching (and
>> consequent
>> > flickering reduction) absolutely impossible.
>> >
>> >
>> >
>> > Maybe somebody made it work in ways I don't know of, but
>> as far I'm
>> > concerned it's not something I'm going to try anytime soon.
>> >
>> >
>> >
>> > ~Raffaele Fragapane
>> > ~Lead "I'm sure we can make it work"
>> > ~Peerless Camera Company
>> >
>> >
>> >
>> -- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- --
>> > -- -----
>> >
>> > From: owner-xsi@(protected) [mailto:owner-xsi@(protected)] On
>> > Behalf Of Oz Adi
>> > Sent: Tuesday, March 15, 2005 11:31 AM
>> > To: XSI@(protected)
>> > Subject: Re: Hair and FG, any point trying?
>> >
>> >
>> >
>> > I got great results using 3 spotlights, with detailed
>> shadow maps,
>> > and rendering with rapid motion..
>> >
>> > but I am checking the FG thing, as it gives even better
>> results on
>> > one frame basis, and renders faster with considerably
>> >
>> > lower memory consumption..
>> >
>> >
>> >
>> > just wanted to see if I am missing something regarding
>> fur and FG...
>> >
>> >
>> >
>> > Oz.
>> >
>> >
>> >
>> > -- -- Original Message -- --
>> >
>> > From: Axel Akesson
>> >
>> > To: XSI@(protected)
>> >
>> > Sent: Tuesday, March 15, 2005 12:21 PM
>> >
>> > Subject: Re: Hair and FG, any point trying?
>> >
>> >
>> >
>> > Why not use a lightrig or something like that?
>> >
>> >
>> >
>> > Axel
>> >
>> > VR
>> >
>> >
>> >
>> > -- --Urspr�ngliche Nachricht-- --
>> > Von: owner-xsi@(protected)
>> [mailto:owner-xsi@(protected)] Im
>> > Auftrag von Oz Adi
>> > Bereitgestellt: Dienstag, 15. M�rz 2005 10:32
>> > Bereitgestellt in: XSI
>> > Unterhaltung: Hair and FG, any point trying?
>> > Betreff: Hair and FG, any point trying?
>> >
>> >
>> >
>> > Hi,
>> >
>> >
>> >
>> > I am trying to render alot of fur with FG (no shadow maps and
>> > Rapid motion, just raytrace).
>> >
>> > I am getting great results, on one frame basis.. but on
>> animation,
>> > all artifacts are set loose!
>> >
>> >
>> >
>> > I've tried playing with all of the FG parameters... but
>> it seems,
>> > like no matter what I do,
>> >
>> > I cant get the slightest improvment, artifacts wise..
>> >
>> >
>> >
>> > any point keep on trying? any big trick, on how to do it?
>> >
>> >
>> >
>> > cheers,
>> >
>> >
>> >
>> > Oz.
>> >
>> >
>> >
>> >
>> >
>> > ---
>> > Incoming mail is certified Virus Free.
>> > Checked by AVG anti-virus system (http://www.grisoft.com).
>> > Version: 6.0.799 / Virus Database: 543 - Release Date: 19-Nov-04 (See http://Nov-04.ora-code.com)
>> >
>> >
>> >
>> >
>> > ---
>> > Outgoing mail is certified Virus Free.
>> > Checked by AVG anti-virus system (http://www.grisoft.com).
>> > Version: 6.0.799 / Virus Database: 543 - Release Date: 19-Nov-04 (See http://Nov-04.ora-code.com)
>> >
>> >
>>
>>
>>
>>
>> ---
>> Unsubscribe? Mail Majordomo@(protected) with the following
>> text in body:
>> unsubscribe xsi
>>
>>
>
>
> ---
> Unsubscribe? Mail Majordomo@(protected) with the following text in body:
> unsubscribe xsi
>
>
---
Unsubscribe? Mail Majordomo@(protected) with the following text in body:
unsubscribe xsi
|
|
