PDA

View Full Version : 3D Rendering History - Pt.2


DanDiaz
10-22-2003, 01:24 AM
Hi all,
Dmitry Shklyar has followed up his first article on 3D Rendering History with '3D Rendering History - Pt.2: To Photorealism and Beyond'. Find out about the history of global illumination and the search for photorealism!

Read the article (http://www.cgnetworks.com/story_custom.php?story_id=1724&page=)

http://www.cgnetworks.com/newsimages/2003_10/2003_10_22_render_lg.jpg (http://www.cgnetworks.com/story_custom.php?story_id=1724&page=)

Enjoy!

Dan

peanut
10-22-2003, 03:59 AM
Interesting article

Its in my favorites !

any link for part 1 of this interview ?


Regards

linus3d
10-22-2003, 04:12 AM
Hrm... has photon mapping been implemented in any commercial 3D package? If so, which ones?

jum'bok
10-22-2003, 04:15 AM
yet another great article :applause:

Geco
10-22-2003, 04:36 AM
Originally posted by linus3d
Hrm... has photon mapping been implemented in any commercial 3D package? If so, which ones? yes, messiah:studio 1.5
messiah:forum (http://www.cgtalk.com/forumdisplay.php?s=&forumid=92)
messiah:homepage (http://www.projectmessiah.com/)

dingo
10-22-2003, 04:56 AM
Originally posted by linus3d
Hrm... has photon mapping been implemented in any commercial 3D package? If so, which ones?

Sure, Animation Master.

Here is a link to a tutorial:

http://www.ypoart.com/tutorials/Photon-Mapping.htm

and to the company site

http://www.hash.com

Array
10-22-2003, 05:00 AM
Originally posted by linus3d
Hrm... has photon mapping been implemented in any commercial 3D package? If so, which ones?

mentalray (and therefore maya, softimage|xsi, and 3d studio max utilize photon mapping)

brazil, finalrender, and vray also use photon mapping to compute GI.

Array
10-22-2003, 05:07 AM
Originally posted by jum'bok
yet another great article :applause:

Thank you. :beer:

erakesh
10-22-2003, 05:07 AM
Nice Article. Wish more of Subsurface scattering was mentioned. I was looking a lot for it.

Array
10-22-2003, 05:11 AM
here is a link to my first article if anyone missed it:

http://www.cgnetworks.com/story_custom.php?story_id=1647&page=

JA-forreal
10-22-2003, 05:41 AM
Good article. But I can't help but thinking about the future of 3d rendering. I have been using a collection of open source rendering solutions for most of my work. And most of these rendering apps have spawned from the individual coders take on an algorithmic principle or two. I love the way that this process is going. Our options for 3d expression through great renders are expanding all of the time with some amazingly surprising discoveries along the way. I have even been influenced to learn a little code to get in on the development side of my own 3d design tools. Artist who want to mess around with science.... Go figure. Much fun to be had. The more we artist learn about computer science the better.

phoenix
10-22-2003, 07:04 AM
very interesting and informative......:thumbsup:

kiaran
10-22-2003, 07:43 AM
First off, thanks for the great article!

I too was looking forward to more discussion on the topic of sub surface scattering. I went on to take a look at Henrik Jensen's website. I would recommend anybody interested in rendering to check it out.

This has left we wondering when a true translucent shader is going to be implimented in retail rendering engines like mental ray. How nice it would be if we got a SSS shader in Maya 6.0, complete with a full BSSRDF implimentation.

Great stuff, very interesting.

Kiaran

oxyg3n
10-22-2003, 08:26 AM
As always, thanks for the article, it was an interesting read. I am looking forward to a new one now. What could it possibly be on??

fxgogo
10-22-2003, 08:45 AM
Great article, but I feel a whole bunch of history has been left out since part 1.

ThirdEye
10-22-2003, 11:39 AM
Originally posted by Array
mentalray (and therefore maya, softimage|xsi, and 3d studio max utilize photon mapping)

brazil, finalrender, and vray also use photon mapping to compute GI.

Same for C4D (part of its GI code is the same of finalRender), dunno about Lightwave

Wonderer
10-22-2003, 01:30 PM
Great article! I only hope I'll still alive when real time GI be avaiable.:buttrock:

neovisionz
10-22-2003, 02:28 PM
Really great both docs!!

:thumbsup:

Neo

Kel Solaar
10-22-2003, 04:15 PM
Great article! very interesting!

Neison
10-22-2003, 05:41 PM
I look forward to reading the article, however:

Can we please have a link to Part 1? I'd like to read them in consecutive order.

Thanks! :thumbsup:

edit: nevermind/found it :blush:

linus3d
10-22-2003, 06:10 PM
Strange how most 3D packages incorporate photon mapping, except for LightWave (which is known for its great renders.)

ThirdEye
10-22-2003, 08:39 PM
Originally posted by linus3d
Strange how most 3D packages incorporate photon mapping, except for LightWave (which is known for its great renders.)

Lw's been the 1st one to integrate global illumination, its algos are quite old now i think (but the quality is still really high)

Annuostivix
10-23-2003, 04:49 AM
very cool, I always like a good read

Meshroom
10-23-2003, 12:17 PM
Nice article as usual. Part 1 can be found HERE (http://www.cgnetworks.com/story_custom.php?story_id=1647&page=)
Take care ppl :wavey:

renderwoman
10-23-2003, 03:11 PM
i love it!:D:love:

PATER
10-24-2003, 04:18 PM
There is another technique worth mentioning:
Irradiance Mapping

Vray does not use Photons by default, it's default GI technique is Irradiance Mapping. Irradiance Mapping is View-Dependant but you can render multiple views to add detail to the Map and it is VERY FAST. Vray CAN use Photon mapping to Speed up the Rendering Process even further. Taking it's speed and Features into account i'd say Irradiance Mapping is state of the art, not Photon Mapping. (The Irradiance Mapping technique is definately newer. There are rumors it'll be included in the next mental ray)

Otherwise, good Article

mdabrov
10-24-2003, 08:24 PM
greg ward author of radiance and hdr
http://radsite.lbl.gov/radiance/contributors/GregWard.html

paul debevec mother of image based rendering
http://www.debevec.org/

and people behind fractals like musgrave and worley
http://www.kenmusgrave.com/
http://www.worley.com

craiggulow
10-25-2003, 06:36 PM
Does anyone know if the Appel paper is online? A few years ago I searched for it, but was unable to locate it. Although irrelevant to anything I have anything to do with now, when I was with Meme-X and we were working on The Incredible Comicshop we wanted to examine Appels work. Now I just curious to read it.

TIA

Array
10-25-2003, 08:17 PM
Originally posted by craiggulow
Does anyone know if the Appel paper is online? A few years ago I searched for it, but was unable to locate it. Although irrelevant to anything I have anything to do with now, when I was with Meme-X and we were working on The Incredible Comicshop we wanted to examine Appels work. Now I just curious to read it.

TIA

When writing this article, I searched quite a bit for it. I couldnt even come up with Appel's first name though :thumbsdow I just knew the name because it was mentioned in so many other related papers.

craiggulow
10-28-2003, 04:53 AM
When writing this article, I searched quite a bit for it. I couldnt even come up with Appel's first name though I just knew the name because it was mentioned in so many other related papers.

His first name is Arthur.

clokkevi
10-31-2003, 09:59 PM
Ray tracing works by shooting a ray from the viewer's eye into the scene, where it either leaves that scene or hits another object. If a ray meets an object, it then sends a secondary ray in the direction of the surface normal,
This is a typo, right?


..until it either hits another surface, in which case the object is in shadow, or until it hits a light.
Hmm.. what are the chances of any ray hitting a infinitely small point-light? There must be something wrong here, or?

chautruong
04-01-2004, 02:52 AM
hi every body.
I'm retouching an old issue:) , I have an question: what is the first meaning of term "global illumination" (global: atmosphere of our planet etc...?)and who is the author? I'm translating this term in my language for an article and i'd like to hearing about that.

James Jacobs
04-01-2004, 08:34 AM
Originally posted by ThirdEye
Lw's been the 1st one to integrate global illumination, its algos are quite old now i think (but the quality is still really high)

Dispite what it might say in its UI,
Lightwave does in fact use an implementation
of Photon Mapping, and not Radiosity.

I think that Newtek might have chosen to refer to their GI as
Radiosity, becuase at the time that they added the feature, Photon Mapping was an uncommon term that not many understood.

Sorta the same reason why they choose to call their "subdivision surfaces" "metaNURBS".


-jj

pklunn
12-11-2005, 07:47 AM
Hrm... has photon mapping been implemented in any commercial 3D package? If so, which ones?

The answer to this is yes....photon mapping is part of the Bunkspeed UDRIVE rendering solution.

Jozvex
12-11-2005, 12:12 PM
There is another technique worth mentioning:
Irradiance Mapping

Vray does not use Photons by default, it's default GI technique is Irradiance Mapping. Irradiance Mapping is View-Dependant but you can render multiple views to add detail to the Map and it is VERY FAST. Vray CAN use Photon mapping to Speed up the Rendering Process even further. Taking it's speed and Features into account i'd say Irradiance Mapping is state of the art, not Photon Mapping. (The Irradiance Mapping technique is definately newer. There are rumors it'll be included in the next mental ray)

Otherwise, good Article

Hmm, well Mental Ray has had Final Gathering for possibly longer than VRay's been around and it's an irradience mapping type of technique. VRay's is most definitely more developed but now that MR's FG can do multiple bounces it somewhat closer in ability. Also, irradience mapping (on its own) doesn't give you a physically based result whereas photon mapping does. Mixed together (or mixed with some brute force path tracing) you get a balance between speed and accuracy (or you try to anyway).

playmesumch00ns
12-12-2005, 08:06 AM
I always thought VRay's lightmap thing was more of an irradiance-caching path tracer.

This thread's a bit of a zombie innit?

Jozvex
12-12-2005, 08:29 AM
Yeah I think it's got just about every method there is and you can mix and match them for the first bounce and secondary bounces

adyy
12-12-2005, 11:00 AM
Ray tracing works by shooting a ray from the viewer's eye into the scene, where it either leaves that scene or hits another object. If a ray meets an object, it then sends a secondary ray in the direction of the surface normal, . . . ..until it either hits another surface, in which case the object is in shadow, or until it hits a light.

not quite!

If a ray meets an object, the following things could happen:
- for shadow calculation : it is shooted a ray from that point to each the light in the scene. If that ray (or rays) do not intersect anything that point is fully iluminated, by the light (or by the lights).
- for reflection calculation : it is shooted a ray in the direction of the surface normal till it meets something or it heats a distance limitation factor. If it hits something it stores the color of the hited object. If that object is transparent, is stored the color and the transparency factor and another (third) ray is shooted along . . . If something like refraction is activated/available the third ray is deviated by the refraction index.
- for ambient occlusion : is is shooted a number of rays in a hemispheric patern, and depending of the hit rate of all those rays, it is calculated a value that is added as black color to the point.

the explanations above are very simplistic. take them only as a "briefing".

playmesumch00ns
12-12-2005, 11:36 AM
That's not really correct: if you're calculating reflection, you need to shoot one or more rays in some direction in the hemipshere around the surface normal. If you're doing specular reflection this will be in the mirror direction, if you're doing diffuse reflection, this will be a random direction in the hemisphere

Jozvex
12-12-2005, 11:45 AM
- for reflection calculation : it is shooted a ray in the direction of the surface normal till it meets something or it heats a distance limitation factor.

That's still not right! If the reflection ray were sent in the direction of the surface normal, looking at a flat mirror you would see the same reflection from every angle! I believe that might be called an orthagonal reflection and it's useful for texture baking purposes, but not for getting proper view dependent reflection. The reflection ray is sent out at the same (but opposite) angle that the incident ray came in at.

Oh dear that wasn't a very good explanation but it's late and I'm tired!

:surprised

adyy
12-12-2005, 01:01 PM
Jozvex: yeah! right! i tied to explain and write down in hurry and I made a mistake. =D
playmesumch00ns : correctly! but my explanation was for a very simplistic reflection calculation.

CGTalk Moderation
12-12-2005, 01:01 PM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.