All posts related to V2
By bograt
#365710
Lumion sucks... There are an endless number of materials out there so for realistic results it is crucial to make each from scratch, or adapt others to suit. The wizard is to assist by giving you a base from which to build your material.
If your only specific enough as to say 'fabric' I see no reason why any of the materials labelled 'fabric' in the library should not suit your purpose perfectly
User avatar
By choo-chee
#365716
hatts wrote:
Today, software is measured not only by essence but also by libraries
Yeah if only Maxwell had some sort of material library
Some sort of centralized, standardized community of user-created materials

http://resources.maxwellrender.com/sear ... v2=0&tipo=
That's almost there but having 1000's of materials of which 99% are common-wizard based materials...
Still I believe maybe a better default library with more examples of real-life materials - not 20 materials from 3 "families" but 100 or 150 that cover
most of the stuff out there.
User avatar
By Half Life
#365728
I suppose I should be clear as to why I think this is a somewhat pointless thing to do (for now).

Fabric is usually made of thread -- and SSS(and I mean thinSSS, because you wouldn't want to use real SSS for this as things stand now) does not allow for the spaces between the threads (meaning you can't get fully transparent areas just using thinSSS controls). Now you could create a layer clip map for the spaces between the threads but then you run afoul of the next limit in the simulation... which is to say that threads are not infinitely flat (like the polygons our thinSSS is applied to), each thread has it's own mass. Furthermore, threads will have their own SSS/anisotropy/roughness qualities which contribute to the overall effect of the cloth... and then of top of that you will have thread "fuzz", due to factors like wear and fiber type. Then you have the issue of how they are woven together -- which also has a huge impact on the final effect.

One of the frustrating things about Maxwell is: it is so good as calculating the light that flaws in creating materials and/or building geometry easily reveals the "fakeness" of them. This is really brought home clearly when you realize that essentially what we've got for creating fabrics is just "shader" level controls (R2/roughness/anisotropy/angle being the primary tools)...

So my stance is: it really is pointless to create a library of "ready-made" fabric materials until the controls offered in MXED are refined enough to make such a library more worthwhile than simple shaders.

All that said, even having a ready-made material is somewhat limited in usefulness, because at the end of the day you will want to customize it significantly... take for instance the example of a couch, best practices would dictate that you would create a custom texture set specifically for that geometry/UV layout to account for things like wear/stitching/etc.

So for now my advice would be to stick to easier to simulate surfaces in Maxwell (if you find it difficult to fake the look you want) -- meaning (p)leather, plastic, metal... and even wood (although I have some issues with how that is currently simulated as well).

I'm excited for the future possibilities of simulation in Maxwell, but also painfully aware of the limitations of what is currently offered in MXED... which unfortunately means that advanced trickery/fakery is the only real viable solution to many of these simulation problems. This runs counter to the very idea of what Maxwell is meant to be, but there it is.

Now, if you are asking how to fake the look of what you want, Mihai already answered that... good textures are the order of the day... just don't look too hard at the final results.

Best,
Jason.
Sketchup 2024 Released

I would like to add my voice to this annual reques[…]