General Discussion Thread


Hey Jeremy,

Been a while! That scene is amazing looking forward to some artist work with it!


Here I’ve merged all the general discussion about LC


1) When people post in the old threads there is less interested and feedback.

Not certain about that; would be interested in the thoughts of others. Seems to me that a single threads isn’t going to encourage more feedback.

2) Gathering all the past challenges in one thread can help in increasing the feedback also from other users that are actually work on a different challenge.
So in this way there are only 4 thread unlocked and for people should be easier to take part to the forum.

Posts added to a thread move it to the top (recent) of the forum; otherwise, I’m not sure the difference in this point to the first point … can you provide more clarification.

My thoughts on the Lighting Challenge forum:

  1. Should clearly identify each challenge in it’s own thread - i.e. standardized thread titles. That looks to be in-place
  2. A master thread listing all the challenges should exist which also discusses guidelines. That looks to be in-place.
  3. Each lighting challenge thread should be available for members to read through, post their progress, final, and questions. The context can be specific to the challenge as another forum exists on CGTalk to discuss lighting and shaders … although in-thread questions specific to this should be fine. That looks to be in-place
  4. Sticky-threads should likely be: Master Thread Listing, Discussion Thread, and maybe the current active lighting challenge.

At the moment, it seems members only post feedback in the current active lighting challenge; not many post in the “feedback” or old lighting challenge threads.

Hopefully this post will stimulate further discussion on the future of the CGTalk Lighting Challenges forum.



Looks like the thread pertaining to portfolios has been deleted so I’ll post this information here for others looking to embed a video in their CG Society Portfolio. You have to use the following weblink format in order to embed a video:{VIDEOCODEHERE}&

{VIDEOCODEHERE} is the code found after the last backslash when you click Share YouTube video.




Just a brief note to say that some of the scenes shared in this section of CGTalk are used elsewhere by render engine makers and on rendering papers without crediting authors, which is not in the spirit of this community or the people who shared their scenes.


Thanks for the message, can you send me a pm where you’ve found the scene used ?


Links checked. LC#47 links now are on. Pls if you find any broken link let me know


All links are not working at all.


Thanks for letting me know. Checking right now


I would like to create a “repository” of ideas for lighting challenges. Please focus both on the technical and artistic aspect. Just to start the thread

  1. Realtime renders vs “traditional”
    I’m sure this one will become obsolete (hopefully). Do you think non-real time renderers are superior in some specialties ? Do you think RT renders can complete in all the fields with them or there is some niche that is still hard to colonize?
  2. Hard challenges. No matter how fast pc will become is there some problem too complex to work out or where only few render engine can succeed ? (mostrous level of displacement, non-visible light rendering, thousands of mirrors and glasses in the same scene etc. ?)
  3. Photorealistic scene still difficult to render . How can we create a Turing test for rendering ? An image with such a complex behaviour of light that would be impossible to reproduce without a believable approximation ?
    Post your ideas and answer here


I believe nowadays the difficulty of rendering a Photorealistic scene is much easier than how it would have been 5 or 10 years back.
realtime renders soon will catch up with offline renders.
so I think we need to focus more on what makes a Photorealistic image believable. rather than how lights are bouncing around and how the light approximation is done.
for example, most of the modern render engine is capable of rendering whatever you throw at them .lights are bouncing, scattering through the surface etc!
but still, we can end up with the bad looking shot I don’t think any rendering advancement can fix that.

There is a secret sauce in making a believable image but I don’t think that’s the render engine capability they (Pixar, ILM, etc) have compared to the others.

The photorealistic scenes are not difficult to render(render time, how light behave and stuff) but what difficult is to achieve the believability in those images.

so yeah basically I think we have to learn how to produce a believable image and let the AI do the rendering and denoise :slight_smile:


Hello, we have not seen a new challenge for a while. Looking forward for some comments.


We are organizing new challenges. Hopefully we will back soon. I need also to clean the forum a bit :wink:


Hello to all,
so I’m gathering a few ideas for the next challenge !. From now on the pace of the challenge will be more “scheduled” with a challenge every 3 months, a better final gallery and hopefully other news :wink:
So the themes that comes up to my mind are

  • A “benchmark” challenge So a difficult realistic small model to render in order to compare renderers, for this we need ideas to create the “ultimate” benchmark so a mixed scene that will involve everything difficult to render (if there is still something :wink:
    -A realtime/still The idea is to have a scene and to render it with realtime engines vs “classic” non realtime renderers (an environment or also a small benchmark scene)
    -A mood challenge… so something like a sci/fi corridor that need to be rendered with different mood (think of same corridor lighted as in Star Wars/Alien)
    -A composite challenge.

— Your ideas…

Also… since I want to revamp some scene… and to create a redux challenge (like the one for the bottles) what’s your favorite challenge since now ?



I would be up for either the scifie corridor or the real-time still. Both are great ideas.


IMHO, reviewing the purpose of this forum/sub-forum should probably be the first step before the next challenge.

If the goal is to simply improve one’s lighting and composition (aka cinematic) skills, then it may be helpful to obtain scenes that are “ready-to-render” - i.e. add lighting only.

If the goal is to develop one’s skills as a Lighting TD, then the current challenges meet this requirement. Most of the assets in all of the existing challenges (1 - 44) require a significant amount of cleanup and texturing before they can be adequately lit.

BTW, this discussion should probably move to:




10 years ago things were very different and now the quality has improved a lot, even if small challenges still seems to still work best . I’m thinking about still life without involving too much textures were it’s easier to include procedurals shaders and also when you don’t need to work too much on textures uv unwrap and other stuff.
Generally we had donations for the models so it was an obj that people started working on and share with others
I’m in the effort of cleaning the old scenes to create a better repository and also cleaning and creating galleries for the last challenges we need voluntaries and modellers. can someone help ? Send me a PM :slight_smile:

Also the next challenge will be HUGE with pro models,textures like it was never seen. I promise. In the meantime let’s do spring cleaning :slight_smile:


For what it’s worth, I really like the challenges which are rougher - requiring shading and texturing and maybe even some additional modeling … they allow everyone to go their own direction at a more rudimentary level … and personally they help me learn a lot more.

I am wanting to go back and work through the challenges until I’ve done them all, these are great :slight_smile:

I know it’s hard to pull off in multiple formats, but an animated shot for lighting would be a lot of fun …


Any timeline on when the next challenge will start? Thanks!



This week I should receive the new model. So stay tuned :wink: