Russia
Joined: Aug 13, 2022
Post Count: 233
Status:
Offline
Re: How to make a mirror ball?
I sometimes feel that reflections in mirrors are not quite natural and are often strangely miscalculated. But in the case of mirror balls, I have no confidence that there is a lie with reflections. Everything is very difficult and therefore convincingly real))). I used to ask a question on this forum about endless reflections in the "mirror corridor", the program has restrictions on this in two cycles, then a black square appears instead of the next reflection. This is done to save processor effort (?) when calculating. In my example with balls, there seem to be no black squares, probably no endless reflections (or the processor is not tired yet))). I want to try to work with translucent textures in my free time - I need to understand the possibilities of light transmission of mirror surface parameters. This is necessary for the backlight to work inside the form. And for example, for realism with the focus of the rays on the surface of the table. It's still black under the vases, if it's a mirror setting, it's wrong. I wanted to get a semi-transparent mirror, while it turns out to be deaf.
France
Joined: Oct 20, 2017
Post Count: 192
Status:
Offline
Re: How to make a mirror ball?
Bonjour.
Pour répondre à la question de Keet concernant les fichiers MTL utilisés pour créer mon miroir sphérique, j'ai d'abord utilisé le MTL du miroir de base (un simple rectangle).
newmtl sweethome3d_window_mirror_1 illum 2 Ka 0.2 0.2 0.2 Kd 1.0 1.0 1.0 Ks 0.5 0.5 0.5 Ns 96.07843 Ni 1.0 "
Je n'ai pas repris la texture d'origine du miroir.
Et pour info le début du fichier .OBJ correspondant.
"# # Generated by Sweet Home 3D - dim. f?vr. 11 23:25:20 CET 2024 # http://www.sweethome3d.com/ # mtllib materials.mtl g sweethome3d_window_mirror_1_1_1_1_1 usemtl sweethome3d_window_mirror_1 v 834.0 100.0 316.00018 v 848.4201 100.72226 313.13174 ........."
Russia
Joined: Aug 13, 2022
Post Count: 233
Status:
Offline
Re: How to make a mirror ball?
I assume that for a mirror sphere, it does not matter what value the Ns and Ni parameters are, they only work for a transparent sphere with parameter d. The mirror wins everything. There must be some other factors that can change the rendering, not the parameters in MTL.
Netherlands
Joined: Apr 8, 2022
Post Count: 790
Status:
Offline
Re: How to make a mirror ball?
Thank you @captaincook. Very clear information and understandable for users not yet familiar with the intricacies of obj/mtl files. (Merci @captaincook. Des informations très claires et compréhensibles pour les utilisateurs qui ne sont pas encore familiarisés avec les subtilités des fichiers obj/mtl.)
Here's a MirrorShpere for other users to download.
@Andrei Maybe the Ni and Ns don't make a difference. At least the Ni shouldn't without transparency. Did you try different values that didn't appear to make a difference? I think the most important part is the sweethome3d_window_mirror prefix in the OBJ file. That might overrule some values in the MTL file for the photo renderer.
Russia
Joined: Aug 13, 2022
Post Count: 233
Status:
Offline
Re: How to make a mirror ball?
Yes, I made several attempts with the mirrored surface of the vase and stopped. The result was the same every time. I will try other experiments, report success or loss)). My research based on the transparent sphere also cannot be nominated for the Nobel prize, the work from "d" and from "Ni" is visible, even the maximum Ns did not stand out in any way, let alone the average values. I tried the boundary values of "d" - every other time a vase appeared on the render, then disappeared completely, depending on the lightness of the color or texture. Dark shades at the value boundary (0.05 or 0.9995) behave unstable on the render, light shades are more stable. That's all I've understood so far)). We should try with translucent textures, I suspect there are prospects there.
Russia
Joined: Aug 13, 2022
Post Count: 233
Status:
Offline
Re: How to make a mirror ball?
even for a transparent vase, the glass material with the parameter "d" 0.1-0.2 looks more spectacular and elegant, not as is customary for glass 0.4-0.5. With a small number, a more clarified glass is obtained, it is more expensive than usual on sale)). And with average values, it is impossible to obtain glass of very light shades, but glass of dark shades (graphite or bronze) is obtained. Therefore, you can change the "d" for the tasks of the model.
