hvanderwegen

Vanilla Cycles still doesn't offer lightgroups or light linking. This $49 version of Cycles called K-Cycles now allows for these, simplifying the lighting workflow considerably. What is K-Cycles Ultra Lighting Interactive real-time viewport and final lighting workflow. Ultimate Lightgroups features with ease of use. Solo preview lightgroup mode. Remainder lightgroup mode. Complete accurate lighting setup. Real-time light mixing for viewport and final render Light linking in real-time with many options. Tone mapping per lightgroup. Full and fast denoising for all the lightgroups. https://blendermarket.com/products/k-cycles https://kcycles.gumroad.com/l/ByeMh E-Cycles also supports lightgroups and light linking, but is far too expensive in my opinion. Until Cycles introduces these, this is sort-of essential for anyone doing complex lighting setups such as in architectural rendering.

Welll... (Warning, tongue in cheek retort 😉 ) Here's one Cycles(X) only studio that's used it since Cycles was first introduced: https://studio.blender.org/welcome/ All joking aside, Cycles was the main render engine for (the now sadly) defunct Tangent Animation, who produced Next Gen and Maya and the Three (Cycles is the main render engine). BarnstormVFX also uses Blender and Cycles for their work (Man in the High Castle, Good Wife, etcetera). There are studios in India, Indonesia, and Japan that also use Blender and Cycles. So it's not as if ANY studio has not used Cycles as their main render engine.

This is too cool not to share. And real-time in the viewport too!

Interesting. I get 3.67 seconds on my 3080ti 12GB with the same settings (I assume you selected Optix as a denoiser). Seems your overclocked 3080 is as fast as the newer 3080ti. Anyway, I agree rendering on GPU is great fun. And the Cycles render preview viewport is just plain amazing to work in.

The Cottage generator is great fun. As are the other node generators from Blenderesse: https://blenderesse.gumroad.com/ The castle is also rather impressive.

Three weeks ago I bought myself a Raspberry PI 400. It's great fun, and reminds me of the old days of 8bit and 16bit computing: an all-in-one keyboard computer, a custom branded mouse and even a paper manual - connect to a screen, and the included micro SD comes pre-installed with Raspberry PI OS - boots when the power is connected. A real throwback to the 80s and 90s. And it runs Blender (2.79) 😉 It cost the same as Apple's mouse for the Studio... Why do I mention this? Well, if Apple would release a similar M1/M2 "Mac Mini in a keyboard" THAT would be the Mac that I'd get in a sec. Interestingly enough, I've been told that Apple registered patents in that direction. Who knows? I'd love a Mac like this 🙂

I hope you will test it with Blender 3.1 - it has much improved rendering performance for Metal M1, so I am very curious how it runs and compares to your CUDA machines for rendering the same scenes.

Aaah... The GPU accelerated subd modifier is a game changer. One of the outstanding performance regressions from the old 2.79 series. What a difference. And finally vertex level control over subd weights creasing as well! Interestingly enough 3.1 loads up in 1 second on my system, where 3.0 took much longer due to some kind of initialization going on in the background. Might be a plugin getting the way, though. Toggling between object and edit mode is faster too. Hitting the TAB key no longer asks for a mode change confirmation! Nice - hit TAB and switch to Edit mode and vice versa. I like how the small usability showstoppers are taken care of one by one each release. 🙂

...And's it time for yet another Blender release! This time the focus is on improving performance: sub-d performance, nodes performance, export performance, image editor performance, ... But also many new features, including the new Point Cloud object that can be rendered directly with Cycles to create sand, water splashes, particles, and different types of motion graphics. Also for Apple M1 users Cycles now has a Metal GPU backend, improving rendering performance by 30-50%. https://www.blender.org/download/releases/3-1/

FBX might still work though. My experience is that even with paid-for objects the original creator rarely bothers testing the fbx /obj files for proper import into other applications. Definitely seems to be the case here. In my opinion every 3d model seller ought to include a working fully textured blend file that allows anyone to convert it via Blender for import in their target DCC. I have seen an uptake in this already the past few years.

https://3delicious.net/index.php?a=detail&id=101631 It includes textures. Loads fine in Blender, import in C4D fails to resolve the paths. But if the paths are removed, it loads fine in C4D. It's a Max exported file.

Ah, that is unfortunate. 😔 Here is an example that could be fixed. If the paths are removed or edited, C4D will load those. # 3ds Max Wavefront OBJ Exporter v0.94b - (c)2007 guruware # File Created: 14.12.2021 12:11:28 newmtl wood Ns 70.0000 Ni 1.5000 d 1.0000 Tr 1.0000 Tf 1.0000 1.0000 1.0000 illum 2 Ka 0.0000 0.0000 0.0000 Kd 0.9137 0.8157 0.6471 Ks 0.6300 0.6300 0.6300 Ke 0.0000 0.0000 0.0000 map_Ka D:\New folder 2\wood.jpg map_Kd D:\New folder 2\wood.jpg newmtl fabric Ns 10.0000 Ni 1.5000 d 1.0000 Tr 1.0000 Tf 1.0000 1.0000 1.0000 illum 2 Ka 0.0000 0.0000 0.0000 Kd 0.8941 0.7843 0.8784 Ks 0.0000 0.0000 0.0000 Ke 0.0000 0.0000 0.0000 map_Ka D:\New folder 2\Text.jpg map_Kd D:\New folder 2\Text.jpg

Well, Blender is free and doesn't need to be installed - just download, extract, and run. Then import the model (file-->import-->wavefront/obj or fbx), check if the textures are correct, export-->Collada. Done! But yes, obj files are notoriously unreliable in terms of texture import. Another option is to open the MTL file in a text editor (Notepad++), and search and remove all texture paths. Place all textures in the model's root folder, and import in C4D. That also works.

This might be a trick: open the obj file in Blender (which seems to load most obj files' textures just fine), save as Collada, and open in Cinema4d. This works for me. May not work with all obj files, though.

It becomes a pipeline issue when it consistently occurs 10, 20, or more times every work day. The OP may have access to libraries of hundreds or even thousands of such files. It then becomes a colossal waste of time, not to mention a frustrating block in one's workflow. I know of at least one viz team at a company doing furniture renders, and they stick with LightWave because all their libraries that they built over the past two decades are LW format. They'd like to make the switch to another DCC (seeing that LW is no longer developed and plugin developers have left such as Octane), but that would mean having to convert all those libraries, which is a huge task.

Correct, I am well familiar with those sites. The OP mentioned specifically how MTL files (which implies OBJ files) are not correctly opened in C4d, so I downloaded a bog-standard free OBJ file with MTL and textures. Of course native Max files load perfectly well in Max 😉 My own experience with OBJ/MTL files is that it is a fragile format and should be avoided at all costs for interchange between DCCs, if possible. Collada or FBX are much more reliable, and even then there are almost always smaller (and at times bigger) issues. Luckily more and more vendors offer native converted and manually corrected files nowadays for all major DCCs. Blend files seem to become a standard part, which is great since the user can then convert to any exchange file format they need. As for Max: I have had to use it a few times in the past two years again for work. Its GUI feels terribly outdated, and crashed on me multiple times so far, at a rate that I now expect to have it crash at least one time during a work session. No such issues with Blender, C4d, or Houdini* (*which isn't that stable with its new work-in-progress GPU renderer, though...)

Never really thought too much about this, but the OP has a point. For example, take this model: https://3delicious.net/index.php?a=detail&id=101631 I open it in Blender with the default OBJ import settings: Worked without a hitch. Import, select the file, done. May need to scale things to match the scene, but no issues with textures. I open the object in C4D with inverted transparency (otherwise it becomes transparent) and it loads with black mats only (I did try all the material settings) because the texture paths are incorrect: Am I missing something? A setting? It can be fixed by using the Project Asset Inspector, but I can imagine it is rather cumbersome to fix this for every single OBJ file that is imported in C4D. I assume that's what @CullenJB is having issues with. I also tested Max 2022. Dragging and dropping the object file in the viewport results in: Issues with transparency again. This file seems to cause problems with the default obj import settings for Max as well. So, the grass is indeed not always greener on the other side... It's probably a setting in the import - but couldn't immediately discover this option. I opened the object in LightWave as well, and I had to fix the transparency channel again. That said, it may just be this particular file. I tested this file too: https://3delicious.net/index.php?a=detail&id=100035 Same results, though. So... a bit ironic that the open source alternative seems to be the only DCC that just works "out of the box" when importing these OBJ files with textures...

This is a very particular case. For comparison, I converted this scene to Blender, and added the Line Art modifer to create similar lines - and as expected, Blender also plays the animation quite slow (slower than C4D actually). View-port navigation also slows down in B for the view camera that must recalculate the lines from all angles, but the perspective camera is smooth (no need to recalculate lines, but the lines look wrong from a too different angle). So it is just the nature of the 'beast': calculating and processing this type of line effect in real time takes its toll on the CPU rather than the GPU. Only with a dedicated edge line art GPU shader (like in games with edge shader effects such as Borderlands) is it possible to calculate these type of edge effects in real-time. Btw, if these lines are 'baked' per animation frame in Blender the view-port works butter-smooth, as does the animation (at least in Blender, don't know if this is possible in C4D, to be honest...). The lines are then converted to Grease Pencil strokes, and can be adjusted per frame, if required. At the expense of not being able to change the camera's view point too much, of course... Perhaps someone here knows how to bake the lines in Cinema4D.

@Papa LennyMy system is more or less identical to yours (3900x and 3080ti), and I tested the scene for you. When I play the animation in R25 the viewport it plays at between 6.3 up to 7.7 frames per second. The viewport is rather sluggish to orbit - around 9-14fps depending on the viewpoint. Does that answer your question? Not sure how it compares to R21 - I did not test.

Would a simple transparency effect not be clearer in bringing the message across?

https://code.blender.org/2022/02/sculpt-development-update/

Yes, I agree for the most part. Irony is often lost and completely misunderstood by mainstream audiences. Hard to get right. Most audiences can't handle it very well either or just lack the ability to understand good funny irony.

Okay, so this movie isn't taking itself seriously at all, which I actually kinda liked - at least the part at the end with the "fake fake-looking CG". I did smile at that. The rest, meh. It's a streaming movie - not "theater worthy", it seems.