V-Ray Suggestion: VFB - Real Time, Camera, Lens and Film Attributes

[stevejjd]

Grant warwick - a heavy weight weighing in!

In VR, you dont want camera, film and lens effects because they distract people. VR is trying to immerse people in a virtual word and wants to trick you into thinking its real.

Great film making is the opposite. Quentin Tarantino not only admits that he is crafting a film, he goes on to embellish the limitations and textures of film. He is not trying to trick you into a virtual, clean world, instead he is telling a story through the limits of a medium. He wants you to know your watching a film and to enjoy that qualities that brings. He not only chooses cameras and films but has them custom made for their attributes. He loves to screen his film through a projector in 70mm Because the receptors in the human eye are randomly placed and the response to light is also non linear.

The character of a lens and the response of film has a natural character more pleasing than linear, digital, uniform pixels. Film is made with a random grain which adds a softness and yet is sharper and more clear than digital. I'm surprised they ere are no TVs with randomized, emulsion-like, pixels.

I don't think anyone would want Photoshop in the buffer but if Chaos group can focus on making real-time lens and film, and as Grant said, standardized LUTs it would be so amazing.

Thanks for the support everyone. Please feel free to add your thoughts....dont let this thread slip. Let Vlado know what you think. This is so important.

I agree with Grant, Vray has the speed and this is just a bit of eye candy. So worth it and so achievable. I think this request is real bang for buck.

[instinct]

I don't think anyone would want Photoshop in the buffer but if Chaos group can focus on making real-time lens and film, and as Grant said, standardized LUTs it would be so amazing.

The problem with that is, that everyone is going to ask for 3 different things, resulting in photoshop in the VFB when all of that is implemented.
For example: For various reasons we go all linear from 3D and do everything in post. So personally i'd prefer if the time was spent on other areas of vray.

And that's the dilemma Vlado is referring to. There is a nearly infinite amount of workflows and requirements out there and you can (and should) only satisfy so many of them. And something that may seem indispensible to you does not matter at all to many other people.

Cheers.
Thorsten

[joconnell]

And that's the dilemma Vlado is referring to. There is a nearly infinite amount of workflows and requirements out there and you can (and should) only satisfy so many of them. And something that may seem indispensible to you does not matter at all to many other people.

Screw you Thorsten, the egg spline has revolutionised my workflow.

[instinct]

Screw you Thorsten, the egg spline has revolutionised my workflow.

I'll give you that one. The egg spline is a noteable exception!

[grantwarwick]

Screw you Thorsten, the egg spline has revolutionised my workflow.

Haha

Seriously though, I think everyone pretty damn happy with V-Ray atm.
I'm surprised that Vlado doesn't remember the buzz when he implemented Glare in the first place. It was better than Octanes and people loved it!

[stevejjd]

The problem with that is, that everyone is going to ask for 3 different things, resulting in photoshop in the VFB when all of that is implemented.
For example: For various reasons we go all linear from 3D and do everything in post. So personally i'd prefer if the time was spent on other areas of vray.

And that's the dilemma Vlado is referring to. There is a nearly infinite amount of workflows and requirements out there and you can (and should) only satisfy so many of them. And something that may seem indispensible to you does not matter at all to many other people.

Cheers.
Thorsten

Ah no, no, no with all due respect, this thread is very specific about what is being asked for and agreed upon by the community. Its limited to the 3 items proposed and no more. We are all trying to discuss those proposed elements and not trying to add other ones. And if someone has 3 more suggestions they are free to ask on there own thread. But they wont because there is no need for a complex of post effect tools beyond what I asked for. There is no potential flood gate!. You simply cannot argue that all people are or will ask for more post effect tools as they have not or if they did it does not belong in this thread I would ask them to remove it.

The issues are the ones I clearly, painstakingly outlined....please do not muddy the exact issue. Those are false fears and I already stated with painful clarity why theses incredibly specific tools be added.

....its not a "slipperly slope idea" where if its implemented will open the flood gate to more photoshop tools.

Photoshop tools are NOT BEING ASKED FOR we are aasking for VERY SPECIFIC TOOOLS that give SPECIFIC results to the render because they are the MAIN ATTRIBUTES OF A CAMERA and we are shooting images with a VIRTUAL CAMERA. That is why the proposed tools are HIGHLIGHTED and NO OTHERS. The proposed tools are DISTINCT from all other post effects in Photoshop or nuke.

and again I say with great respect

Vray is already very fast - I think its time that Chaos group really put some time into studying film, photography and cameras. Study movies like Blade Runner and get an appreciation for the beauty of seeing through a lens and the response of film. They make vrayscans of materials...its literally the same principal but for camera/ film effects. Chaos group should Study samples of lenses and film-stocks and approximate them. This has nothing to do with photoshop and everything to do with appreciating what makes a beautiful image and whats most important to have when rendering interactively!

After this issue is addressed, of course Chaos group can continue optimization. We can achieve a fast image but its always dull without the suggested ideas. We will have fast renders in the buffer that look final and filmic.

If you read what I painstakingly wrote there is no lose, only win.

[instinct]

I disagree. First of all, i DID read your posts and i am aware that YOU only requested these three things. I did not write clearly enough probably. I did not mean "if you start implementing it, people will request all sorts of things" i meant "there are hundreds of threads already requesting said features and everyone 'only wants these 2 or 3 things'" and it is as Vlado noted a tough decision to make on what to spend developer time.

"The main attributes of a camera" is also not set in stone and people will list different attributes there. You are making the assumption that your definition has to count for all, and again i disagree with that. (e.g. i would say that film stock profiles and grain are an attribute of film, not the camera. You get the idea)

And in regards to "Photoshop tools are NOT BEING ASKED FOR we are aasking for VERY SPECIFIC TOOOLS" let me quote Grant:

I like to look at the VFB as an app in itself these days. A mini compositing software.

With all due respect, you are asking for developer time to be spend features that i don't need and i am objecting that. I am not trying to muddy your requests or intentions, i simply disagree with it.

Cheers,
Thorsten

[kagemaru]

The problem with that is, that everyone is going to ask for 3 different things, resulting in photoshop in the VFB when all of that is implemented.
For example: For various reasons we go all linear from 3D and do everything in post. So personally i'd prefer if the time was spent on other areas of vray

I agree with this on so many levels.
I understand i'm possibly within a big minority as a vray user here but i will never understand why i would want a baked in post process effects on my renders.
Post effects that i could change anyway afterwards in thousands of variations in a matter of seconds.
But, as specified before, i think i'm not the standard vray user seeing people going nuts for realtime bloom/glare in the vfb, so i don't think there's anything wrong having more options then before. From my vray use i'd like though to have developers' time spent more on other issues, but that's me.

[stevejjd]

Okay I respect that.

I have a conviction that this is what is needed and of course I'm going to fight for it.

The VFB is like mini compositing software now that it has the color controls and if the requested features are implemented it would still be limited to a mini compositing software.

Mini compositing software is not dedicated software like Nuke. There is an important distinction.

We are asking for exact and distinct features and no more. Yes, Mini is a good name as it infers limits. Limited to what is necessary.

It would be up to chaos group to create a standard by going out and studying film and camera and lens. That's the whole pont. Even Grant requested that they formalize and standardize LUTS.

What I am asking for is on the same level as vray scans.

[joconnell]

Even Grant requested that they formalize and standardize LUTS.

The film and colour industry has been trying to do that for years with no success unfortunately.

What I am asking for is on the same level as vray scans.

That'd involve having a calibrated environment where you know the light levels hitting it and the material properties of every surface in it and then shooting it with every possible combination of camera, lens and film stock, then comparing it to a ground truth "pure" image and measuring the difference that the lens and stock is adding, let alone whatever process you used to scan the film negative itself. That's a lot of possibilities

[stevejjd]

I dont think Grant is asking for Chaos group to standardize LUTs across the film and VFX industry but to have a standard of LUT that works with Vray.

It doesn't have to be a literal camera lens and film stock matching - just study it and approximate it, fake it even. It doesn't have to be hard.

That is why Vlado asked if its for matching of for achieving an aesthetics. And its is for aesthetics. So I guess on that level I made a mistake in saying its on the same level as vray scans. What I mean is that they should just take a look at lenses and film to see why it looks that way and just do an approximation of what its like. It doesn't have to be hard. Lots of software like Davinci resolve fake that and it looks great.

[stevejjd]

I agree with this on so many levels.
I understand i'm possibly within a big minority as a vray user here but i will never understand why i would want a baked in post process effects on my renders.
Post effects that i could change anyway afterwards in thousands of variations in a matter of seconds.
But, as specified before, i think i'm not the standard vray user seeing people going nuts for realtime bloom/glare in the vfb, so i don't think there's anything wrong having more options then before. From my vray use i'd like though to have developers' time spent more on other issues, but that's me.

As I said there would be an option to bake in. This whole idea is so you don't have to look at raw renders while making things as it look so harsh an unnatural. People who are happy with raw renders and waiting till the end for post work have enjoyed that workflow since forever. People who want to see their work as if through a real camera live with this as a frustration.

[jasonhuang1115]

If I understand the discussions so far correctly, we are gonna get the real-time bloom/glare in 3.5 and for the LUT, wouldn't we already can make a LUT that is generated through externally? i.e. if you like a certain film response curve, you can export it from said Magic Bullet (plus tons of stock film LUT available online) and load that into VFB. Did I miss something?

[stevejjd]

If I understand the discussions so far correctly, we are gonna get the real-time bloom/glare in 3.5 and for the LUT, wouldn't we already can make a LUT that is generated through externally? i.e. if you like a certain film response curve, you can export it from said Magic Bullet (plus tons of stock film LUT available online) and load that into VFB. Did I miss something?

We are just discussing the possibilities of a product suggestion. Just a few additions to the frame buffer.

Simply put, there are V-Ray users who would like to have real-time bloom, glare lens effects and some film type effects like film response grain and Standardized LUT's that work with V-Ray. Those things will help remove the raw feeling of renders. Its a bit like how in substance painter you can turn on lense and other effects to work with and it takes away the raw feeling. Saving a frame and testing it in Nuke then changing the scene and testing in Nuke can be annoying.

Lots of people really enjoy enjoy these effects in other render engines so V-Ray users are asking if they can enjoy that too. And we are discussing the Pro and Con of chaos group implementing the tools.

Some people here are worried and voiced their concern that it's just too hard and would take development time away from optimizing V-Ray to be faster, more powerful. A totally legitimate concern. The Artists who request these features are saying that V-Ray is already fast and should take a moment to address this because they feel its important to have as its adding image quality or beauty to an already fast engine.

It been clarified that the desired camera, lens, film, lut profiles are just for aesthetics only and not there to perfectly match actually existing film camera technology. It purely to have renders look really nice in the buffer instead of waiting till post work.

its also been noted that no one is asking for really super complicated tools, like Photoshop in the buffer. This idea I appose. so the proposal is limited to camera, lens, film, lut controls because those are the ones that add naturalism to the raw render.

I admit that I should go back to my original post and edit to make it more clear. Writing clearly is surprisingly hard. I'm really glad I met opposition because it allowed me to clarify my idea and understand the negative points. I'm still convinced it can be done without harm to those who wont use it.

I wanted to stimulate a discussion and even a debate over this. Debating this is great because it will force people who want this to really think about what exacly they want, why they want it and if its truly worth it. Also, just because I want it and believe in the idea doesn't mean its good for chaos group and I want to challenge them to think about it and to express their worries and to find out if we all think its good.

What would really take this debate or discussion to the next level is if the artists who want the tools could work together to produce and image of the frame buffer controls they want with diagrams and explanations. I get the feeling it way more simple then people think. Also if people who want this don't speak up then this idea is just a fart. Vlado should be skeptical and its up to people who want this to explain and refine the idea. If this idea is not worth fighting for then its a waist of time and just a big fluff on the forums.

But some ideas break through like Luc AKA BigGuns with the Aisurface shader

[jasonhuang1115]

Steve, any video to show what you are referring to in other renderers in this regard? Not the real-time bloom/glare but real-time camera, lens, and film attributes...I am trying to wrap my head around how it's significantly different from loading desired LUTs manually either during or after rendering is done.

Love the idea of having some crucial tools around for artists to showcase renders without going through an actual comp process, but just trying to visualize how big of a difference your proposal would make.