Mash

It will work fine, I run a Ryzen 5950X 16 core machine with a geforce 3090 all day perfectly fine. Ram - Incredibly unlikely, even single channel mode wont slow it down this much CPU - Could be, but gpu rendering is obviously heavy on the gpu, the cpu just doesnt come into the equation much GPU - My guess is its overheating. The fact it works but is just slow points to this C4D settings - Could be searching slow drives for texture when you hit render.

ztl is a zbrush file. If you have a maxon one sub then you can install zbrush and export as something else.

The first thing I would do is simply run another gpu benchmark so you can narrow down the problem to one specific app/render engine, or the entire machine in general. eg go grab the octanebench download and check your render speeds.

The only way Ive found to do it is to disable image sampling and antialiasing. The real solution is to composite the background in afterwards, but that depends what you need.

Here:

Loading textures over the network?

I'm a 10/10 on the confused scale right now. Why would you not want to earn money from your plugin?

just a tip, your entire product page doesn't mention "NFT" even once. Anyone searching for a plugin to do this is going to get zero hits on google.

No proper answer from me, but I will say, the label is penetrating through the glass when the subdivision has happened, this won't be helping you.

Well then upload a scene and let us pull it apart. Include your PS document for the displacement.

With my 13 pro max using '3d scanner app', I found I got much better results with the front facing 'true depth' camera. The rear lidar shots would seem to work well whilst capturing them, but then the final processed shots had textures in all sorts of wrong places. The truedepth camera scans all came out pretty well, its just a shame you have to do it blind with the phone facing away from you. The rear lidar shots worked well for capturing larger room scale shots though.

For something like this 16-bit would be enough, that jumps you from 256 height levels to 65,000 height levels. 32bit with 4 billion height levels tends to only be needed for landscapes.

You really have a whole host of FPS-related problems in that video 0:09 Here you have 2 problems. The most obvious one is the constant flickering from one frame to the next. This is caused by a combination of the camera motion and a lack of motion blur. Motion blur would largely fix this, but then of course youll have a much softer animation. The real answer here is that this simply isn't a shot you should do on low fps projects (24, 25 30fps) The second problem here is that you have a 24/25 fps mismatch. Once per second the camera skips an entire frame, this usually means you have rendered at 24fps, but your editing app is trying to stretch this to 25 fps. 0:22 Here you have a different fps mismatch. You can see the image jumping 4-5 times per second, this tells us that there is a 24/25 to 30 fps conversion problem.

Maxon have ignored requests to make public forums for over 2 decades, I'll be amazed if they create anything at this stage. Though the irony here is that maybe you've potentially just shown David that he can charge a subscription for forums, and so will have one released by next week 😉

If you're aiming for realism, then yes, use real world units, or at least something somewhat approximately close to them. DOF is based on the real physical size of the lens, its aperture, the camera sensor size and the focal length. If you use values 10x higher then you'll find you have to use silly values 10x higher/lower elsewhere to compensate. Or when you grab a readymade material with SSS like milk, it won't work well because it is set up to allow light to travel X distance through the substance. If you find c4d's cameras and materials are too large at default scale, then press ctrl + D and change the default project scale from 1cm to 1mm (or smaller) This will bring down the default size for primitives, materials, modelling tools etc.

But if you ignore them, you cant lower their score. Dilemma 😉

Im not a redshift user really, but in octane I would be looking for one area in particular. This looks very much like an ambient occlusion inaccuracy. Most render engines have some sort of setting to select how much detail small things should be rendered to. Octane has Ray Epsilon standard has Ray threshold Physical has shading Error Threshold

Without your texture we will get a different result, but if I hit render, then I have no black lines

This should still work for you:

Older versions of c4d didn't use this current online licence system, nobody is going to go back and update old software to add the licence system to it. Is there any reason you want an old c4d version to learn with?

In a nutshell, no, you're screwed. In the grand scheme of things these are actually pretty good meshes for a CAD import.

If you can spare that single model and let us look we can get you an answer very quickly, otherwise we're just stabbing in the dark. The first stab though, don't use those UV offset, scale and tiling settings to change a material. Instead use the texture mode (top middle toolbar, 5th icon) This will allow you to move, scale and rotate the texture using the normal 3D tools. This is much more likely to work. Second, what is the first black texture projection set to on the left side of the tag list? Octane has this vague set of rules when using multiple textures with multiple projections where sometimes one projection mode will override the others.

I've been here a while. I sort of fell out of reading forums after cgtalk managed to turn itself from a great place to chat, into an ocean of garbage. I went back a few months ago to see if it had improved, but it was still just unconstrained trolling and faux-outrage. Im perfectly happy to read about the bad aspects of c4d, but watching the same people say the same thing relentlessly in every single thread every single day, just pure misery.

Serious point though, don't stick 8 gpus in one machine if theyre using redshift (the built in render engine) It trips over its own shoelaces if you try to go much past 4 gpus. You'll be better off with 2 machines with 4 gpus.