Request- Film Response Curves (Comparing Octane to V-Ray)

[peteristrate]

For me, the problem is not necessarily the embed of film response curves.
These are just film filters and, as Vlado said, i don't find any relevance in using them now that the era of video-cameras is almost gone.
If you really want to apply such an "instagram" filter, you are free to do it in your compositing software... and if you want to have the consistency between your filter in compositing software and the VFB, then i think the LUT profiles are quite ok (unless there are any bugs, that need fixing).

Now, my actual issue, as Grant also mentioned and tested it, and as i have noticed it long time ago when i mentioned this to you, Vlado, about half a year or an year ago...(but i didn't know exactly what was the problem or where to look into for doing any tests - Grant got it right though), is that the photographic exposure (specifically, the curves) are not working as realistically as they do in iRay or Corona (and now Octane - i haven't tested Octane personally).
Vray comes with a flat linear curve that is hard and quite impossible adjusting to match photorealism exactly.

All in all, the image in Vray comes with more vibrant colors, and has stronger highlights and darks...(i render as 32bit .exr and apply the curves in comp) that i find quite impossible to balance in comp(using curves). So, as part of my compositing routine, on top of everything, i always have to apply a desaturation filter and 2 curve filters to brighten the blacks a bit and clamp the highlights a bit as well... in order to try to match iRay or Corona... and even then, i'm still not happy with the result. iRay and Corona seem to have a much more and balanced lighting and colors in the raw render.
Maybe it's just me that i can't get it right, but Grant noticed that too in his IronMan test:
Vray image has:
- stronger highlights
- darker darks
- more vibrant / saturated colors (even in the HDRI behind)
- falloff / curve / balance between bright points and darks is steep, not that mild: Look at the highlight on the golden part of the IronMan's legs: Octane's transition from highlights to the less reflective area next to that highlight is much more natural, while Vray's is much more contrasty.

[kosso_olli]

Carefully read Grants post again. He said that he used the Motiva plugin for the renders of Vray, which makes it look a little different. To really compare the images, we would need the raw images from both renderers.

[jenya.andersson]

Have you tired softclip? Take your render into nuke apply softclip node. With logarithmic compress. Should mimic film soft roll-off in highlights.

As in the film response curve, have you tired to make some? its pretty simple. Photograph Macbeth chart with film stock, use equalizer in nuke gizmo. To extract macbeth chart info. Export LUT out with your film stock.

Personally i think its best to keep film response curve with compositing side of things or DI (in nuke/ Baselight). As 90% time you would be working with a neutral graduated plate in 3D.

But a softclip would be nice to have in Vray.
What be cool is a normalize preview button in the vray frame buffer. To preview that you have information in highlights and blacks is not clipping.

[peteristrate]

What be cool is a normalize preview button in the vray frame buffer. To preview that you have information in highlights and blacks is not clipping.

That would be a cool feature to implement as a button in the VFB.

[grantwarwick]

The raw renders are very close.
It comes down to shader differences. But that is near impossible to match as the EXACT same colors in Octane don't match V-Ray whereas greyscale values do! Very weird!
Either way, it took me hundreds of attempts to get gold and other materials looking photoreal in octane. Now I'm actually insanely close in V-Ray! It just sucks I can't adjust gamma AND apply film response curves from motiva.
Getting there though!

(Edit: I'd love to show all the different examples but basically it looks like it's impossible to render objects with pure 255 colors in Octane, something often overlooked in V-Ray where you can get things rendering with pure solid colors. For example, in V-Ray you can get a pure red sphere completely red whereas octane introduces a small amount of green and blue into the mix resulting in a more realistic saturation. I dno, it's going to require some more studying and hard evidence but I'm confident I've discovered some new tricks for getting V-Ray looking more realistic and they are totally not based on physical settings, I've gone down the more artistic path but I could never do that before octane because I didn't have realtime feedback which is the primary reason you get away with it

[3LP]

Did you had a chance to see/compare the visual and render time difference that you would have between Vray RT GPU and octane?

Would be great to compare both renderers that use the same hardware/technology.

Although I'm always partisan for a better render output than a faster one.

I can understand that octane is faster than Vray for some situations, but against Vray RT GPU.?
I'd love to see a fair fight
Unless it's not possible to get what you want to achieve in RT GPU because of missing features.

Stan

[grantwarwick]

Did you had a chance to see/compare the visual and render time difference that you would have between Vray RT GPU and octane?

Would be great to compare both renderers that use the same hardware/technology.

Although I'm always partisan for a better render output than a faster one.

I can understand that octane is faster than Vray for some situations, but against Vray RT GPU.?
I'd love to see a fair fight
Unless it's not possible to get what you want to achieve in RT GPU because of missing features.

Stan

V-Ray is faster by a considerable amount so far but that is with adaptive mode. Octanes power comes from it's initial passes in the first 20-30 seconds allowing for realtime feedback.
Because V-Ray is so optimizable I can comfortable half the end render time it seems!

[^Lele^]

Grant, if you set a red colour, and the renderer doesn't output PRECISELY the same colour (times your lighting), the renderer is wrong.
For if it wasn't, then a renderer may decide your pure red would be mighty cooler if it looked a tad more toadish green. And then your spheres as cubes, and your straight lines as spirals (perception is, after all, personal, right? Perceptual rendering is something which scares the socks off me.).

All this talk of "photorealism" going amiss in VRay (with a SLEW of BRDFs, provenly accurate radiometric equations, highly customisable render bias) while being present in Octane (ONE specular BRDF. Two tracing modes which fall to bits when a few GI bounces are used, no matter the tweaking one inputs in the controls. ) which can't even render a pure color, set my blood boiling.
MATHS is king in the rendering world, and evaluating what looks photoreal and what doesn't should be done -at least in public- with exact references and complete disclosure of the setups used, or avoided entirely, for without objective data backing the claims up, the exercise falls into the realm of rumor (or personal preference. just as worthy, for public consumption.).

Besides this, given the camera response curves are mono-tonal (ie, they simply represent a brightness curve which ain't linear, but kinky due to the patchy response of chemical film across the ranges), there is no way applying those to a vray render would produce a hue skew like the one you described for Octane.

I am on record OPPOSING the introduction of OCIO and more advanced LuT controls in the VFB, precisely because it would add one more confusing layer for most users, while those which would really work with an OCIO workflow would know they don't need to see the LWF renders skewed in the VFB by a custom LuT (rather, one would do that in post, lest one ends up chasing fringe values in non-linear color spaces for the rest of time... If any of you ever lit under Log space, you'd know.), so i guess to some it's no news, but i have yet to see one single case of VFB-baked color controls which can't be replicated in post and as such, to ME, they are a low-to-no-priority avenue.

We render unclamped (don't we?), so nothing is precluded to Post as far as image and colour treatment goes.

[3LP]

It isn't; the curves are full floats. A faster algorithm is used when the image is processed for display so that it is interactive, but when writing the final image to disk, full floating-point calculations are used.

Best regards,


Vlado

Hi Vlado, sorry to coma back so late on this one.
I feel like what you're saying isn't working for me.
simple scene here done in 3 sec and I can "break" the curve in simple easy steps.

No Exposure used with/without curves :


With exposure used :

[3LP]

(New message as forum limit max upload files to 5)

Saved file from the VFB (where your said the full float is supposed to be computed ):


As you can see, simple scene where both results should be exactly the same.
I just overlid the scene and compensate with the exposure in the VFB but the visual results in the VFB are just not usable and the exported results are not right either.
For me, if I can't see it in the VFB, then it's not worth using as I will not play by saving all my renders/elements to view them outside the VFB every-time I make a test render.

Maybe a check-box in the settings to enable the quick/bypass cheat the make the VFB faster would be a good workaround?
For those who don't botter waiting a bit longer but who does want to have the "true" CC inside the VFB, it would be very welcome.

Unless I'm mistaken and I'm doing something wrong in my example, the curves are just not usable in production for me.

Thanks
Stan

[3LP]

Grant, if you set a red colour, and the renderer doesn't output PRECISELY the same colour (times your lighting), the renderer is wrong.
For if it wasn't, then a renderer may decide your pure red would be mighty cooler if it looked a tad more toadish green. And then your spheres as cubes, and your straight lines as spirals (perception is, after all, personal, right? Perceptual rendering is something which scares the socks off me.).

All this talk of "photorealism" going amiss in VRay (with a SLEW of BRDFs, provenly accurate radiometric equations, highly customisable render bias) while being present in Octane (ONE specular BRDF. Two tracing modes which fall to bits when a few GI bounces are used, no matter the tweaking one inputs in the controls. ) which can't even render a pure color, set my blood boiling.
MATHS is king in the rendering world, and evaluating what looks photoreal and what doesn't should be done -at least in public- with exact references and complete disclosure of the setups used, or avoided entirely, for without objective data backing the claims up, the exercise falls into the realm of rumor (or personal preference. just as worthy, for public consumption.).

Besides this, given the camera response curves are mono-tonal (ie, they simply represent a brightness curve which ain't linear, but kinky due to the patchy response of chemical film across the ranges), there is no way applying those to a vray render would produce a hue skew like the one you described for Octane.

I am on record OPPOSING the introduction of OCIO and more advanced LuT controls in the VFB, precisely because it would add one more confusing layer for most users, while those which would really work with an OCIO workflow would know they don't need to see the LWF renders skewed in the VFB by a custom LuT (rather, one would do that in post, lest one ends up chasing fringe values in non-linear color spaces for the rest of time... If any of you ever lit under Log space, you'd know.), so i guess to some it's no news, but i have yet to see one single case of VFB-baked color controls which can't be replicated in post and as such, to ME, they are a low-to-no-priority avenue.

We render unclamped (don't we?), so nothing is precluded to Post as far as image and colour treatment goes.

Hi Lele,

I understand what you're saying, and I can understand why a renderer that takes decisions for you that are not "mathematically" correct is scary. And I would to, as you really don't want to end up in a place where you fumble around for hours because it doesn't react the way it should.

Although, I do like have my everyday work be done faster. And if this means be able to do more inside the VFB and be able to avoid having to go back and forward between VFB and post-soft to see a end result, I'd would be able to have that possibility.
Not saying I will use it on every project, you will not end up using VrayFur on every single mesh you create in max either, but you like be able to quickly add one if you need one...
For me, same goes for the post treatment in the VFB.
Even if this doesn't have any mathematical meaning or is outdated, "film response curves" are just a starting point for grading, and they are useful, even if in the end you just consider them as a bunch of presets of color grading (and leave all the historical meaning behind)

Be able to to load those curves in the VFB would make me win few steps in my everyday process and thus make me work faster and have either more time to make my work better or just be able to go home sooner to enjoy more family time. Not saying this curves will change my life and make me win 2h a day, just that every single small or large improvement in the workflow (active shade RT, progressive sample, render elements, CC in VFB, etc) added up and make a big difference in the end.

In the same thought, it been few years I'm asking to get Vray saving straight to *.dng as this will let us use Adobe camera raw to grade quickly or shots. I know it's not something that could be part of a huge production pipeline like Nuke, but for a lot of people that use either AFX or PS or lightroom to grade their shots, it will be a huge game changer. Just another small stone you add to the edifice that would make a big difference in the end.

As for the fact that a full red object doesn't render out pure red in octane versus Vray, like I said on top, I understand your concern, but if it does "look" and "feel" more realistic to have a bit of green/blue introduced (maybe because a real lens would do so, don't know?) well it's something that I could consider work with. Why? simply because my renders would look better straight out of the box without more work from my end. Not saying I'd like the renderer to make my work but any help and push in the back is appreciated.

Stan

[^Lele^]

All in all, the image in Vray comes with more vibrant colors, and has stronger highlights and darks...(i render as 32bit .exr and apply the curves in comp) that i find quite impossible to balance in comp(using curves). So, as part of my compositing routine, on top of everything, i always have to apply a desaturation filter and 2 curve filters to brighten the blacks a bit and clamp the highlights a bit as well... in order to try to match iRay or Corona... and even then, i'm still not happy with the result. iRay and Corona seem to have a much more and balanced lighting and colors in the raw render.

No. I specifically tested iRay (0x1 version. Am on the NVidia beta, will soon switch to it, with identical results.), Arnold, Renderman (19 and 20) and Corona.
Done proper, ALL of them have the SAME, bar nothing, rendered result from an identical scene, shaders, lights, and sampling-wise.
Speed/cleanliness may differ, as not all the methods are identical (Not even the Brute-force ones.), and most surely BRDFs vary (even with identical models and parameters, implementations may vary slightly), but most definitely not the light transport equation when used in LINEAR space.

Maybe it's just me that i can't get it right, but Grant noticed that too in his IronMan test:
Vray image has:
- stronger highlights
- darker darks
- more vibrant / saturated colors (even in the HDRI behind)
- falloff / curve / balance between bright points and darks is steep, not that mild: Look at the highlight on the golden part of the IronMan's legs: Octane's transition from highlights to the less reflective area next to that highlight is much more natural, while Vray's is much more contrasty.

Point by point:
*) "Stronger highlights": than what, exactly? Do you have any information on the BRDF model you're comparing against VRay? In Vray alone you can choose FOUR different BRDFs, all of which will behave differently from one another under the same settings (color, gloss and ior). So which one of those are you claiming is stronger than which other?

*) "Darker Darks": As in, it's supposedly extinguishing your lighting before it should? It's well possible, if your textures/shaders are pitch black, if you have cutoff thresholds which are too high, samping which is too low, if you're not working under LWF, and the list goes on and on. Show me a render, and two scenes, and i will accept this. For it's clear to me that amongst the current limits of VRay, a leaky light transport equation isn't one.

*) "more vibrant / saturated colors (even in the HDRI behind)": again, this is tantamount to claiming you input a color, and VRay renders you another. Show me (along with your color mapping settings, for both renderers.) and i'll believe you against years of measured results.

*) "falloff / curve / balance between bright points and darks is steep" : See point 1: you are comparing two different BRDF models, what else do you expect if not for them to be vastly different? And who's to say the Octane render looks prettier? The author? You? Because i personally strongly disagree, as the way hilights are caught at grazing angles is clearly vastly better for VRay, to the point it shows highlights in parts where octane falls flat.
But this would be falling back into the personal and particular, and the whole point is to avoid it.

Share scene setups, and informations on the BRDFs models Octane uses, and perhaps we may be able to collectively test and get a clear picture of things.
For now, these are results that can't be peer-reviewed, and further show more than one sign of arbitrary choices in the benchmarking setup, which would need verification images alone can't provide.

[^Lele^]

As for the fact that a full red object doesn't render out pure red in octane versus Vray, like I said on top, I understand your concern, but if it does "look" and "feel" more realistic to have a bit of green/blue introduced (maybe because a real lens would do so, don't know?) well it's something that I could consider work with. Why? simply because my renders would look better straight out of the box without more work from my end. Not saying I'd like the renderer to make my work but any help and push in the back is appreciated.

Stan

I strongly disagree: artistic flair should be PURELY into the hands of the artist.
Entire markets REQUIRE utter material and lighting accuracy (Automotive, product design, high-end archviz), and that specific sector is one where research is aplenty, with very good reasons.
A renderer (well, ANY tool.) that makes choices a user hasn't initiated, to skew the linear, Normal result, is doing it wrong, there are no two words to describe it, as far as i am concerned.
Anyone's free to use what anyone wishes.
Just don't show me a broken tool, and tell me to break the one which works to match.

If there's something you can't do in Post, but can do in the VFB, you likely have to dig a bit deeper into the modern post workflows, for they have been pushing out movies which look allright, i'd say, for the past 10 years, and those tools are today entirely free for personal, and small commercial use (if i hear "Adobe" and "post" mentioned in the same sentence, i switch off, sorry.).

[3LP]

I understand your point
I also strongly agree that the artistic flair should be in the artists hands, but you could also have to possibility to switch that on and off, leaving both possibility to the artist. Like all post options in the VFB, you can switch it on, and off.

You come from a pipeline where the tools that are used from start to end needs to work hand in hand (like me as I worked 6 years in VFX) to produce what we all see everyday in show and movies, and no point about saying it's not incredible, it is, and I know there are tool to achieve those results.
They just imply the fact of going from one tool to another to another, to another, to be able to see the end result.
I was referring to try and skip a few of those steps to be able to see faster what the end result would be, inside your 3d software
I never said you'd need to bake that in your render, I would still go linear and make those adjustments down the line in my post-soft, but at least I'd have been able to skip a few step for the WIP and previews.

Why do you think :
* lens effect was introduced there in the first place?
* the CC tools are being introduced as well?
* people are asking to have even more tools in the VFB like those response curves?
* people are asking (and other software company delivers) to be able to interact more with your workflow inside the frame buffer? (like picking the dof point in the VFB like octane, or having even more post effect done straight in the frame buffer like octane or unreal or unity)

I know they can all be done in post soft like nuke or other, it's just more convenient to have them straight in the VFB because you win time and steps.
Why not add some other presets to switch quickly between looks? Those would be "film response curve" named now, but could have another name tomorrow or just a bunch of home made or bought presets.

As for the fact of having a red not rendering pure red in your vfb, I though it could have something to do with the lens (or CCD?)

I might be wrong, but take a stage with a white floor and a big 1m pure red ball.
Shoot it with a Red scarlet, and a Red Epic Dragon, and a Arri alexa, and a Sony CineAlta and with cooke or Arri / zeiss lenses.
Overlay all those footage on top of each other, are you really getting pixel perfect 1:1 match for all those (let's say) 8 different shots?

Everything is not pure straight maths. because real life is not pure straight. Otherwise people would not be adding dirt everywhere and make lens distortions and chromatic aberration and film grain and whatever other impurity that would end up that your pixel in the end, is not the pixel you would have with a pure mathematical render.

Again, I understand you want to have the absolute power and be able to achieve this in post, but some things are just not feasible in post or require a lot of workarounds. Example, it's only with deep rendering (who has been "reasonably" recently introduced) that it has been possible to make correct/true DOF in post. Before that it was just not feasible straight out of the box in one pass. You needed to deconstruct your render in several passes to be able to have a correct dof.
Well nowadays, even if deep rendering is possible, it's still easier to enable dof in your cam, and done.
This would also be with settings like blades, anisotropy and center bias etc. Again, all that is coming from "real life" camera, and it's only recently also that the Vray cam lost his Physical name as before it was called "VrayPhysicalCam".

So if there was a way to get a more "realistic" render out of vray easily, why not let you enable that feature, like having a bit of blue or green in you supposed to be 100% pure red? By the way, GGX wasn't also introduced because it was better suited and with a more "real looking" end-results? Even if you where able to achieve the same result with 5 layered vray blend materials?

Now with your background I can also understand that you don't like to read "Adobe" and "post" in the same sentence And I get that, but that doesn't mean that some other people would like to use those technique, that's personal preference for each person out there, and just neglect their wishes because yours are different, doesn't mean they are unworthy of consideration.

Stan (who doesn't like to play devil's advocate )

[peterguthrie]

my 2 pence:

more tools and features in the frame buffer please. 99% of the renders I do never make it past the frame buffer, so the more creative stuff we can do there the better. Just offering this up as an example of how different people use vray in different ways