As I wrote, the last few days I have been dabbling with the Reflectance stuff in Cinema 4D R16. I pulled one of my old files from my archives and tried to create some reasonably “real” materials.
As you see in the first picture, in my work i did not necessarily aim for ultra-realism. Since those machines inside consist of various blank metals, often the same kind of material for different components, you either end up with everything being indiscernible reflections only or a grey blob, depending on the individual material settings. Incidentally the same problem makes it so tough doing photographs or filming this stuff. Therefore finding a good balance has always been the challenge. Combined with other factors like not having suitable HDR environment maps at hand most of the time ultimately one had do employ a lot of tricks. Despite all that with Cinema 4D things still looked flat and not necessarily convincing. That, however, is now changed considerably.
If you look at the second image, things simply look more metal-y to begin with. Yepp, even with only a short time spent on this and my own limited texturing skills you can actually feel that it’s much more like how materials behave in terms of bouncing around light in the real world. The great thing is that this is all within the same system, so specular highlights show up in reflections as they would on natural materials and for the most part this more than anything else gives this feeling of realism. I can’t give you specific advice about the various models since I haven’t experimented nearly enough myself, but you’ll get the hang of it.
As a downside of course all this is a lot more computationally intensive than your old specular, even more so since you will need to use the Physical renderer to avoid antialiasing and sampling issues in certain situations (the materials work with the Standard renderer as well, though). This also reveals another flaw – if you want to call it that – For now this is an isolated solution to a specific problem and does not go hand in hand with other things. I for instance would just love if in order to make it truly physical we could control Diffusion and Transparency/ Absorption in a similar fashion. Inevitably you will have to fiddle with these parameters and your old existing materials will not translate too well. Most of the time you will want to reduce the diffusion or use much darker, more desaturated colors since the reflections and speculars will boost brightness notably.
It’s also obvious that this kinda plugged-on solution introduces a lot of redundancy with the user needing to make sure bump maps and so on are always synced, which without the program providing instances for textures and materials can be a major pain. Anyone who has ever created a reasonably complex material knows this. These problems can multiply exponentially since the number of layers is almost unlimited. So in summary, it is still not trivial to create good materials and as I said in my other posts, this is perhaps a bit too much and overdone. In fairness, though, some of this may resolve itself if and when they get around to also bring the other shading stuff to that level.
There are a few annoyances I noted as well. Let me just summarize them shortly:
- The panel can be towering high, especially with the anisotropic and fabric shading and the extra texture slots twirled open. This is slightly stupid, because it makes little sense to e.g. change the anisotropic pattern far at the bottom and having to scroll up again to see the preview. They should have put more thought into this.
- There are drawing glitches, which probably have to do with the panel being so long. When you numerically edit values for instance, some of the sliders/ scrubbers above and below will not refresh properly and break up. Only scrolling vertically will enforce a redraw.
- The Layer Fresnel settings seem to have a bug. When choosing Conductor materials and a preset, the IOR and Absorption readouts are greyed out and don’t update. Only when you select Custom do they become available. Now this may be by design, but do I know what these values are for any of the metals? It’s inconsistent with the Dielectric materials and by all means should function the same, be it just that the user may need a baseline reference for tweaking his own values.