Chiron (carpaint WIP)

Forums Gallery Chiron (carpaint WIP)

This topic contains 32 replies, has 5 voices, and was last updated by Jeremy Hill 4 months, 2 weeks ago.

Viewing 15 posts - 1 through 15 (of 33 total)
  • Author
    Posts
  • #2331
    Jeremy Hill
    Jeremy Hill

    Doing some research for a carpaint material node, this is the first test.

    #2336
    Jeremy Hill
    Jeremy Hill

    Second test.

    #2338
    Jeremy Hill
    Jeremy Hill

    Side view (uff, bad uv mapping on b-pillar panel).

    #2340
    Jeremy Hill
    Jeremy Hill

    Comparison of metallic vs. solid color.

    Metallic:

    Solid color:

    #2341
    Philip
    Philip

    Nice! 🙂

    Philip

    #2342
    Jeremy Hill
    Jeremy Hill

    Great! Just working on this on the side as I do other things (namely .. adding obj/fbx/etc import for the gui 🙂 )

    Here are the same materials, but with different base colors. First metallic white with solid-color blue, then the opposite combination.

    Not too pleased with the metallic white, so there is yet work to do, but there is also an issue Albert found with the material layer simulation that may improve this case, as well.

    In the end, the goal is to have a nice carpaint material with just a few easy parameters, which covers these sorts of cases (metallic, solid, etc), with built-in orange peel, and such.

    #2343
    Philip
    Philip

    <Not too pleased with the metallic white, so there is yet work to do>

    Perhaps not perfect, but a good start.

    <In the end, the goal is to have a nice carpaint material with just a few easy parameters, which covers these sorts of cases (metallic, solid, etc), with built-in orange peel, and such.>

    Sounds really good – and orange peel too…! Great!
    I really like the direction you’re going with the material system and the whole philosophy of Bella.
    Looking forward to the Rhino plug-in 😉

    Philip

    #2345
    Eric
    Eric

    ferrari

    Do a pale two tone, but you will need a good screen.
    I’m not so happy with this white metallic (maxwell render), but last week I got a good screen. A cheap 27″ hanspree quantum dot, was new but with a slight defect was £162 delivered…. so I will try again the difficult white metallic soon.

    Is making material presets useful right now? I think the material system itself needs some thought, and the naming conventions are very offputting at the moment. I even think ‘layer’ should be renamed ‘ThinLayer’ or ‘ThinFilm’ layer is too common a word (even without PS) it becomes difficult to communicate in forums..

    low-sl

    low

    Low-sl.. only have pentium 920 quad core… big machine died..

    #2346
    Jeremy Hill
    Jeremy Hill

    Very nice materials, as usual. 🙂 Regarding the question of presets, I’m open to any suggestions, especially if they can be expressed using formulas, which can then be encapsulated in a node that implements those formulas.

    On the question of terminology, that was always going to be a bit thorny, but it had to be decided at some point, and is unlikely to be changed at this point, sorry to say. I don’t intend to change your mind, but I will give you our rationale, which is summed up in this diagram:

    So, the core concept consists of two main components, the substrate and the layer, with the logic for how to build a material following from that — you may have a dielectric substrate, or a conductor substrate, with an optional layer overlaid. And the sheet material is a layer with no substrate, used on its own as a material.

    That is what we are simulating, physically, and we did not find other terms that would communicate it (in our opinion) more effectively. But whatever the case, this is something that is infeasible to be changed at this point.

    #2347
    Eric
    Eric

    What would half polished stone look like in that diagram?, just substrate + multiple semi-opaque layers? Or would you use a fine noise map in the roughness slot with 1 layer.

    I have read the docs but I am confused.

    #2349
    Avatar
    Thomas An.

    I don’t understand this effort. How are you determining if something looks real without a reference photo taken in a controlled way and try to recreate it ?

    Asking people for realism “opinions” is entirely useless; once we are told something is a render we manufacture criticism out of thin air.

    #2350
    Jeremy Hill
    Jeremy Hill

    @eric: we currently support only one layer in the model, so you could use any combination of roughness-mapping in the conductor, and the layer, as well as opacity-mapping in the layer. Beyond this, you can also use the blend & stack materials, to combine multiple materials.

    #2351
    Jeremy Hill
    Jeremy Hill

    @thomas, I am just exploring at this point, I can proceed to compare with references once I have got a better idea of the parameters I’ll need, and how they are related to one another, at which point I’ll probably have to revise, and so forth.

    #2352
    Eric
    Eric

    So I could blend indentical substrates+layer with different roughness for their respective layers. What is the broad term for substrate/dielectric+layer ?

    Another way would be to offer dual mapping slots; overall map and high frequency map which always blend at 50% as a nod to simplicity.

    I expect there is a compromise here between rendering speed, and ease of use. Maybe an idea is to offer a fuzziness value to interpret any given roughness across a broader range.

    #2353
    Jeremy Hill
    Jeremy Hill

    In discussion, I guess we have basically used “layered dielectric” and “layered conductor” to refer to those.

    Regarding dual maps, that is something that will generally be handled using a mix/multiply type of node (which I have yet to write). On the other hand, for something like the carpaint material I am working on here, if it made sense to expose two separate inputs (as you say, perhaps for ease of use), that could be done, as well.

Viewing 15 posts - 1 through 15 (of 33 total)

You must be logged in to reply to this topic.