Sticky: Please Let Us Know What Vantage Developers Are Not Going to Support

[jujubee]

It was brought to my attention that some of my material and function requests would "just turn Vantage into another VRay GPU and we have that already."
So if I'm interpreting those words correctly, Vantage is just a stripped down cross-platform GPU renderer (perhaps attempting to tackle Omniverse eventually) and that's the only real benefit you get out of this currently (aside from speed because it's light) over the built-in VRay GPU.

So if I can make a helpful suggestion/request:

Please add a Sticky at the top of the Wishlist noting what the developers of Vantage are NOT going to or planning to implement.

This way, we're wasting time requesting these features and also know what limitations to expect moving forward.
I've been converting many libraries to work with Vantage only to just find out that you aren't planning to support a bunch of VRay material types because you're saying those features are redundant.

Thank you!

[s.balabanov]

Hello jujubee,

Since its inception, Vantage is designed to be a real-time ray tracing engine with speed being key. Vantage's primary goal is to enable users to experience their scenes interactively in real-time - that means with the fastest rendering speed and the minimum setup effort possible. That does not mean we do not want to support V-Ray features - in fact we are gradually expanding the supported features set. The recently added normal maps and material override are great examples of this. However, we are careful not to reduce rendering speed. In order to do so, we spend additional time to figure out how to add rendering features with as little performance impact as possible. This means that we cannot put a fixed estimate on when a feature will be added. If we have such a roadmap, it will be frequently edited both in terms of time estimates and development order.

The comment about Vantage becoming another V-Ray GPU refers to the way V-Ray GPU was developed. I'm sure you remember it was initially named "V-Ray RT", where RT originally stood for real-time. However, as numerous new features were added, they slowed it down and it evolved into a faster production render engine instead of a real-time one. This is exactly what we want to avoid with Vantage. We know Vantage is great for interactive presentations and reviews with customers, since it provides a fast and convenient way of getting your project to real-time. Vantage is also great at rendering lighting-accurate previsualizations for exploration, spatial awareness, lighting/material variations, etc. with just a few seconds per frame. We are also happy to see that Vantage is sometimes used for production rendering, but this is not a priority for real-time rendering. When quality is a priority you can always rely on V-Ray GPU which is a feature complete and robust engine used in many production pipelines. That is one of the reasons we provide the means to transfer Vantage cameras back to 3ds Max.

As for feature support, we have published a quite extensive list of what is and what is not supported. Were you able to find it before committing to your material conversion process? If not, we should find a way to communicate it better.
https://docs.chaos.com/display/LAV/3ds+Max

You mention that your materials require conversion and specifically VRayBlendMtl. Could you share, what are the features that you need to be supported? What is your intended workflow with Vantage? Where do you position it in your pipeline? Answers to these questions will help us understand the issues you experience and provide possible solutions/workarounds.

Many thanks,
Simeon

[Phillip Miller]

Great explanation Simeon.
Another fact that is worth understanding is that adding any new functionality to Vantage (or most GPU-compute renderers) will slow it down a bit - whether you use that functionality for your scene or not - so we need to be very careful about what we add as we will need to claw back performance with each addition.

Kind regards,

- Phil

[q976431]

Thank you very much for your questions.
I think if there are some unique functions that cannot be replaced by other ways. The development team should temporarily consider sacrificing some performance, adding it, and then optimizing it in later versions, such as IFL sequence file. At present, I can't find any alternative method to complete the production of LED screen.
Performance degradation can upgrade hardware
However, if some key functions are not supported, it is difficult to complete some specific animations

[jujubee]

I do appreciate your explanations.
As for supported or unsupported features, I suppose if I worked in this every day and wasn't fresh again to VRay after such a long break, perhaps I would remember or know what is supported and not supported. I'm learning to be very careful and check that list knowing its limitations.
Even if you were to add a list or an indicator in some sort of log file which would tell users new or old that a part of a material isn't supported, it would be pretty handy. I know it's probably a big ask but I figured it would be worth asking.

When quality is a priority you can always rely on V-Ray GPU which is a feature complete and robust engine used in many production pipelines.

I can understand some items not being high priority, while others (bump mapping and vrayblendmtl) have been pretty standard and not having access to these or having to find a workaround - I'm not the only person, company or developer that will be feeling this. On the other hand, general normal mapping for example had worked, the regular blend material works, and a few other items. Not sure from a code perspective how much more performance would be impacted if the regular ones already work (even if poorly supported.)

VRay GPU IMO still isn't very good. It was buggy back when it was VRay RT in Beta and beyond and still crashes on me with alarming frequency 10+ years later (or however long it has been.)
If I wanted regular GPU I would have stuck with Octane or F-Storm, both of which are faster and crash less. I'm not trying to be difficult, but just plain honest. I switched back because of Vantage. I was actually exploring Omniverse prior to this but this is much faster (albeit different from a shared workspace perspective.)

I do like the speeds and see the promise. In proper hands it is more than a simple draft/WIP tool and I think you fully realize this or at least the potential. I mean seriously, do you always plan to keep this free/beta and a draft tool forever? I don't believe that's true.
If it gets most of the features of regular VRay/GPU, then I could anticipate it eventually replacing it at some point when RTX cards or greater become the norm.

[jujubee]

The development team should temporarily consider sacrificing some performance, adding it, and then optimizing it in later versions, such as IFL sequence file.

The real answer to this question/suggestion is "how much" of a performance impact do each of these added features create?
I can make assumptions and it's very easy to ask for things, but as you probably know and in Vlado's own words to me back in 2006 at a SIGGRAPH convention, "people expect there's a Make Pretty Picture button..."