Nvidia released Mental Ray as a plugin for 3DS MAX - maybe petition for Blender?

Just saw this: http://escape-technology.com/en/news-blog-events/847-nvidia-releases-mental-ray-for-3ds-max-2018

I wonder if it would be a big deal for them to make plugin for Blender. Maybe we can file some kind of open petition for Nvidia to connect Blender with their Mental Ray ?

I assume you’ve never used Mental Ray, correct?
I used 3dsmax for about 15 years, and for most of the time the main render engine i used was mental ray. and i hated it! (pretty much everyone i know hates it)

Is there a specific reason for wanting to have Mental ray in Blender?

I was just wondering exactly the same thing. I learned to use Mental Ray with Maya and used it for some time. I don’t have a lot of emotions related to it as hate or something, but there is absolutely nothing special about it as far as I can say. Relatively a lot of learning and no advantages against other renders. I would definitely go for Octane, Maxwell, Cycles, Luxrender or even Thea instead.

Damn, I remember when Mental Ray integration was considered a good thing. When XSI came with Mental Ray by default and built their whole software around it people were impressed (including me). How times and technology has changed.

The reason Blender doesn’t have any capable renderer is it’s infectious GPL licence. There’s no way nVidia, or any other sane company is going to make their renderer opensource just to integrate it into blender. There are some attempts with custom blender builds and exporters, but it harms the renderer workflow to the point there’s not much point using it if you want to be efficient and productive.

As for Mental Ray, sure, it’s not a very good renderer, but neither is Cycles. At the very least, mental ray can render an interior scene in a resonable time and has usable shadowcatcher solution. Cycles doesn’t.

I’ve worked with Mental Ray, Vray, Renderman, Arnold, Cycles and Octane. Mental Ray is an alright machine but in order for it to be truly good you need to really understand it. It’s a really steep learning curve (even comparing it with Renderman - which is also known for a steep curve).

Cycles gets great results with far less effort and Renderman is great if you want to learn about a widely used engine in the animation industry. Both of these are available in Blender and are free. Mental Ray has its uses but it’s probably not worth the trouble for 95% of Blender users.

Blender has a few external professional renderers including Renderman, Octane and Vray. Biggest problem with integrating with Blender is not some much the GPL license. Is that their is a need to develop a C++ API for external developers to allow easy connection to all the internal Blender C++ data structures and functionality. The way it is now their are two option. One use python scripting plugin which is not a powerful, direct or flexible. Second modified Blender code directly and create a communication layer to the external renderer engine e.g.(like Octane does). This option is more tedious and needs to updated often as Blender core code changes.

You could start a thunderclap

You can’t do that without making your code GPL-compatible (i.e. free and open source). All the commercial renderers ultimately go through the file system or the network.

The way it is now their are two option. One use python scripting plugin which is not a powerful, direct or flexible.

For external renderers, it’s not all that bad - Luxrender is pretty well integrated, for example. The GPL restriction still applies.

Vray for Blender has poor workflow compared to for example Vray for Max. And developers themselves say it’s pretty much due to the Blender’s licence throwing sticks under their feet.

All in all, Blender’s biggest enemy is Blender.

There is a saying where I am from. I am struggling to translate it a bit, please forgive me. It goes something like this: For a bad dancer even his balls get in the way.

You could try switching to Bforartists, but they can’t do anything about the GPL either.

While there are some known workarounds, the only real solutions is to either create a team to create a brand new FOSS program (from scratch) with a permissive license or never use Open-source creative software again due to the GPL’s prevalence there (even if the apps end up getting things that’s the equivalent of finding a cure for cancer with a 100 percent success rate).

Cycles is getting better with time. The denoiser cuts render time 10x.

The GPL is there to make Blender an open creative program. I rather have it open like it is and it’s also nice that a capable application like blender is available for free for anyone.

:slight_smile: Yes, under same license as Cycles (APL 2)

For a company it’s simple: “Does it pay off? No.”
Anyone else able can check if kick start will run the endeavor or choose to just do it…

They just implemented micro displacement while Mental ray had it back then in 2004, that’s like 13 years ago + dev time. And you guys are planning to change the GPL license to a more permissive one? or re-code it? that’s a big lol to me.

I see a lot of potential in Blender, that’s why I’m still around here. Most of the problem resides in the user base, the majority of users thinks Blender has nothing wrong and there’s no need to add features or enhance them. On the other hand developers thinks the way it has to be done is, they code it first, they add logic after that and then they add some design/Better UX-UI. Which should the the other way around.

On this, Cycles has a bright future ahead of it when you count the sheer number of new features and improvements going into Blender 2.79’s Cycles and how many of them were not written by Sergey (who has been the main developer since Brecht started working for Solid Angle). Sure, the shadow catcher could use a way to allow its usage with no need for additional compositing, but all the other stuff is already at a good state.

Sure, it’s still not the best engine to use for indoor scenes, but I haven’t seen a lot of development news out of the Luxrender community lately (from reading their forums at least, even though it is very capable with Luxcore) and Yafaray is still incomplete when it comes to material capabilities (no SSS yet for instance).

Appleseed could be an alternative, it’s getting better ways faster that other FOSS render engines.

Keep in mind we also have radeon pro render, which is also developing fast. PBR, Denoiser, CPU+GPU, Out-of-Core rendering, Material portability between all major dcc apps, etc.

why use mental ray at all?

It’s likely that Mental Ray still gets a fair bit of professional use but it’s getting a bit old now.

Is it still considered to be competitive these days with the industry looking increasingly towards path tracers?

Not any more - http://images.nvidia.com/content/newsletters/email/pdf/Mental-Ray-FAQ-201.pdf

If Nvidia is dropping support and development for Mental Ray, it is probably no longer relevant.