Leaderboard

There are some flaws to Lumen and Nanite based on some videos I have seen. The first is that the Lumen engine has a hard time consistently converging all GI light rays in real time. There is this shifting blotchiness which you would notice should the scene contain smooth surfaces. I would imagine to get the performance you need, the error threshold is set higher than you would use in a standard render engine. That is why all demo scenes use rocks and rough surfaces where the subtle shifting of brightness and color is not as noticeable. If smooth surfaces are used, they will probably not be a brightly lit. Another problem is with thin objects, especially leaves. If the camera is up close, all looks good but at a distance the leaves just disappear. That one I cannot explain nor could the demo operator in the video I was watching. Nevertheless, Unreal 5 still remains hugely impressive. It is just not the be-all and end-all yet of real time, hyper-realistic rendering. But they are close. I also have to admire that all demos are run on a PS5. I mean, that alone is impressive. Now, I hear two things when it comes to real time 3D being used in Stagecraft for The Mandalorian. I hear it is based on Unreal Engine but it also uses ILM's Helios render engine. Pretty certain they are not running the 70 foot diameter of LED "volumes" of Stagecraft using PS5 so the partnership between ILM and Unreal probably has developed the be-all and end-all of real time rendering. It probably requires a render farm of 1000 nVidia RTX-A8000's to make it all work - especially if they want to use for more than TV and make large format movies to be shown in IMAX theaters. If anyone has the specs on Stagecraft 2.0, please share. Interesting fact: The partnership between ILM and Unreal probably was very easy to create given that Unreal's CTO is Kim Libreri, a past ILM VFX supervisor who left in 2015 for Epic. I met Kim during Siggraph 2009 and we actually had a nice chat about fluid simulations. Quiet guy with a searing intelligence. Dave

Igor - the first thing to say is thank you for your efforts in thinking how best to set this up and your proposals : ) However - for me the whole thing is too complex and too rigid. Personally I'd just have one forum. However I would make it essential that the original poster tags the post to make it searchable. Maybe you could set up tags for 'Home DCC' - eg: I'm doing this in 'C4D' and 'Target DCC(s) eg; I'm interested in how to do this in 'Blender' or 'Houdini' or 'Any'. And tags of course for the 'Topic area' I think having all the subforums just makes it too complex, too hard work for moderators and too hard to discover for occasional users. It also makes it difficult when a topic area falls outside of easily identifiable categories. For example, one area I would like to discuss is C4D 'Takes' - ie, does Blender or Houdini have something eqivalent to that functionality? Where would that go? It wouldn't fit anywhere and would therefore be easily lost. As Dast suggests - I'd also drop any idea of a specific number of DCC's - there's lots of related tools and techniques that might come into play. For example a modelling request might best be solved with something like Zbrush, or MOI3d or something else that is not necessarily a full blown DCC. So, in summary: I'd keep it simple. Very simple.

@Igor I know the subforum titles are only a work-in-progress, but I would suggest not to include any number or names of DCC's in the title itself. As this title might soon become obsolete as soon as a new DCC is added to the list. So while "The beauty of 4 ..." might sound nice and to the point NOW, it might be already obsolete by tomorrow. Just a suggestion. Also, I am a bit afraid that the need to move requests to the appropriate sub-section might soon become a bottle-neck. Since you already mention that you don't have enough hours in a day, I feel that the need to move topics manually is a big no-no. Can you provide a tag for each of the DCC to be discussed? This would require the user who creates a request to select the appropriate DCC-tag. This could help manage the different topics. And it would also avoid the need to include the DCC in the title (as your example illustrated). On the other hand, when does one decide to move the topic? When does one tag the request as being "solved/done"? Who does this tagging? I have more than once seen people ask question in the forum, with members providing one or more answers ... but no further actions or response from the original poster. There needs to be some discipline from the requestor to tag the request, but there is no guarantee. Which will soon lead to a cluttered subforum. And I don't think moderators have so much free time at their disposal to clean up after everyone lacking discipline, right?