Re: Material, seamless_circular_noise-cloudy
Jyrki did some experiments some time ago - see RS web gallery/Animations/Experimental. Mark , your 'Mill.avi' is just above Jyrki's cloud anim . Nice Mill ! I didn't know you ever finished such a polished collision detection anim . Some nice anims on the gallery page to download . Seems something's wrong with the big 19MB Divx 'Seasons Greetings' . The smaller mpeg1 on the experimental page looks great though . studio www.niagara.com/~studio www.studiodynamics.net
Re: Material, seamless_circular_noise-cloudy
On Mon, 20 Feb 2006 09:15:44 +0100, studio [EMAIL PROTECTED] wrote: Jyrki did some experiments some time ago - see RS web gallery/Animations/Experimental. Mark , your 'Mill.avi' is just above Jyrki's cloud anim . Nice Mill ! I didn't know you ever finished such a polished collision detection anim . I rendered several versions back then, I think this wasn't the best one: some stray balls rolling to the right out of themselves ;) This was 2004, I haven't tested RS native collision since then, who knows improvements were added? BTW all credit to Jyrki's cloud animation, it looks really fluffy! Some nice anims on the gallery page to download . Seems something's wrong with the big 19MB Divx 'Seasons Greetings' . The smaller mpeg1 on the experimental page looks great though . studio It's playing well here (a bit blocky)... It's a must: some really promising new V6 features that could revolutionize landscape rendering in RS! -Mark H
Re: Material, seamless_circular_noise-cloudy
The first attached image is showing the outside. The other has camera put inside the cloud (using different settings also). I used 100 volume samples, so rendertimes? Yes, there are rendertimes... So although it looks good it's pretty unusable due to insane rendertimes Remember Mark's Marble ? : Summer of '05 ? No ? Basically it was the usage of Mark's cloud Mat (I trimmed the six thousand lines of trailing text from the bottom of your mail so that I could include it below) . ** Mark : Here is a really simple rendering , of a sphere with an Earth mapping (thanks edmond) using your Volumetric Cloud shader . I don't think I did too much there , except change your Distant Light out for a Spotlight . Anyway , a 36 Minute render , Ray Traced shadows ... http://www.studiodynamics.net/images/marks_marble_shad.jpg and then shadows turned off Four minute render ; http://www.studiodynamics.net/images/marks_marble.jpg Cheers Mark ! studio www.niagara.com/~studio www.studiodynamics.net
Re: Material, seamless_circular_noise-cloudy
Jyrki did some experiments some time ago - see RS web gallery/Animations/Experimental. Vesa Yes (hi Vesa) : As I make mention in the Planetary Tutorial ... http://rswiki.matthias-kappenberg.de/index.php/Planetary_Project_Tutorial Jyrki did get the ball rolling with his volumetric cloud animation tests using Mark Heuymans shader (someone please correct me if I have my facts mixed up) . I did some more work in this area and the 'Mark's Marble' renderings were a kind of result (below) . http://www.studiodynamics.net/images/marks_marble_shad.jpg Really should do a tutorial about that whole era of RS procedural experiments , since they appear to be a bit forgotten already . studio www.niagara.com/~studio www.studiodynamics.net
Re: Material, seamless_circular_noise-cloudy
Hello, You say _brute-force_. Are there other methods that can be used? I know the game industry is looking into approaches made available with DX10. Then again, not sure if it deals with selfshadowing or just simple volumetrics, haven't really looked into it. I guess if you have any knowledge - in advance - of the nature of the fog, you can save majority of the computations. For example, if the fog is so thick that rays penetrate only a fraction of the average thickness, the approach of stepping into the volume until some limit value is reached is much quicker. And there's no limit how much you can 'fake' or precompute an effect. The more parameters you fix, the quicker it gets. The final limit is a pre-rendered image, which can be shown instantaneously:) Best regards, Vesa
Re: Material, seamless_circular_noise-cloudy
On Sun, 19 Feb 2006 11:32:03 +0100, studio [EMAIL PROTECTED] wrote: Jyrki did some experiments some time ago - see RS web gallery/Animations/Experimental. Vesa Yes (hi Vesa) : As I make mention in the Planetary Tutorial ... http://rswiki.matthias-kappenberg.de/index.php/Planetary_Project_Tutorial Jyrki did get the ball rolling with his volumetric cloud animation tests using Mark Heuymans shader (someone please correct me if I have my facts mixed up) . I did some more work in this area and the 'Mark's Marble' renderings were a kind of result (below) . http://www.studiodynamics.net/images/marks_marble_shad.jpg Really should do a tutorial about that whole era of RS procedural experiments , since they appear to be a bit forgotten already . studio Garry: Yeah, more proof that the mailing list is very volatile. Old stuff, June 2005. The main trick was to use a steep curve to map noise turbidity to itself, to get areas of dense fog and relative clarity instead of a blurry noise. I'm using volumetrics routinely. It's slow (maybe problematic for animations) but despite that: really useful! For example, I use a gradient fog material, getting a blue-white sky gradient and volumetric shadows at the same time. Will be included in the -hopefully upcoming- Wiki tut. -Mark H
Re: Material, seamless_circular_noise-cloudy
On Sun, 19 Feb 2006 23:57:14 +0100, Vesa Meskanen [EMAIL PROTECTED] wrote: Hello, You say _brute-force_. Are there other methods that can be used? I know the game industry is looking into approaches made available with DX10. Then again, not sure if it deals with selfshadowing or just simple volumetrics, haven't really looked into it. I guess if you have any knowledge - in advance - of the nature of the fog, you can save majority of the computations. For example, if the fog is so thick that rays penetrate only a fraction of the average thickness, the approach of stepping into the volume until some limit value is reached is much quicker. And there's no limit how much you can 'fake' or precompute an effect. The more parameters you fix, the quicker it gets. The final limit is a pre-rendered image, which can be shown instantaneously:) Best regards, Vesa Hi Vesa, list, Would that be possible in V5 VSL already? 'Smart volume sampling', I like the sound of it ;) Thanks for going into this happy clouds, Mark Heuymans