- This topic has 6 replies, 4 voices, and was last updated 3 years, 1 month ago by Jeremy Hill.
August 22, 2019 at 9:39 pm #1702yoann
i Did some first test with bella and the output result look very good !
I only play with the renderer for an hour but i must admit that if you dont have the proper technical knowledge,
a normal user will be completly lost with the actual materials.
Something less technical and more user friendly, like the way renderman handle material is a good exemple. 🙂
yoannAugust 22, 2019 at 9:54 pm #1703Jeremy Hill
Thanks very much for the feedback! 🙂
In time we will be creating various specialized material nodes, similar to how it is done with the existing water and glass nodes (which are used internally to create dielectric materials).
So, these will be higher-level materials, exposing simplified sets of parameters, to make it easier/more efficient to create and work with plastics and so forth. The trick is defining them correctly, to give enough control, but not to clutter things with rarely-used parameters.August 22, 2019 at 11:01 pm #1706Oscar Cano
Welcome aboard Yoann, and thank you very much for your interest in Bella!
Bella’s material system has been designed to be highly realistic and accurate while at the same time be affordable and intuitive to use. We have been very focused in their raw usage, but making them easier to use and understand is, a Jeremy said, a direction in which we are moving. Not only is a thing in the UI side, but through the use of a good documentation, sample scenes, video tutorials, and so on.
For the moment I think the IPR is helpful in the sense that it allows you to touch and see interactively, but definitely we need, and we will, to take more actions.August 24, 2019 at 4:28 pm #1716yoann
Thanks for the answer !
Another quick question, do you plan to implement OSL ?
btw i think adding some kind of roadmap to the website would be a good idea.August 24, 2019 at 10:20 pm #1718Albert Martinez
It is possible that we can provide free code eventually. It’s planned to provide an SDK to be connected with external tools and our GUI/CLI application is already making use of it. We are not planning to make OSL for entire Bella source code at the moment.
Absolutely, some roadmap will be added, as soon as we are completing other priority areas like adding better documentation, sharing free scenes to show Bella features, etc.
We also want to know your Bella feedback because that will help us to prioritize some of our roadmap elements, as we want to be aware of the needs of our customers, for example prioritizing VFX features over architecture or the opposite.August 26, 2019 at 3:43 pm #1728yoann
I didn’t have the time to dig through bella so i can’t give good feedback but
i really like the layered system approach over the classical udershader is a really nice idea
and if it’s flexible enough to adapt to different requirement and give a
bit of artistic freedom, it could become one of the strongest parts of your renderer.
In my opinion, what we need the most is the ability to handle complex material interaction,
skin is a good example, it’s really hard to handle complex skin shading (like blood on the skin, ink for tattoo etc)…
Also, be able to layer several different bsdf could be really nice too.
Other features like efficient motion blur, volume is also very important.
I’ll try to give more interesting feedback as soon as possible. And sorry for my english :D!
One nice addition to the gallery could be a render of the moana island!
https://www.technology.disneyanimation.com/islandsceneAugust 26, 2019 at 11:07 pm #1738Jeremy Hill
Thanks for the feedback, and I think your english seems great. 🙂
Regarding OSL, I think my colleague misunderstood what you were referring to. I would say, first, that this is not something to look for in the near term, as we have many other core features yet to implement (as you mention, motion blur, volumetrics, etc). And second, though we are open to the idea, it ultimately depends on a pretty careful evaluation, to see if/how it can be integrated with Bella’s internal architecture.
Regarding skin, we think it will probably be best to implement something like a custom skin material node (or a layer-like node), rather than try to handle it generically, since there exist a few nice cheats/hacks specifically created for rendering skin.
As for multiple bsdfs, it should already be possible using the blend & stack materials, so please feel free to let us know if/how those are unable to help you achieve whatever it is you are attempting to do.
- You must be logged in to reply to this topic.