Bxdf fixes cook torrance #930

devshgraphicsprogramming · 2025-09-16T14:03:10Z

Description

Continues #899 , #916 and #919

Testing

TODO list:

…eturn quants

include/nbl/builtin/hlsl/bxdf/common.hlsl

include/nbl/builtin/hlsl/math/functions.hlsl

include/nbl/builtin/hlsl/cpp_compat/vector.hlsl

include/nbl/builtin/hlsl/bxdf/fresnel.hlsl

include/nbl/builtin/hlsl/bxdf/cook_torrance_base.hlsl

…cache with defaults

devshgraphicsprogramming · 2025-10-20T09:52:57Z

include/nbl/builtin/hlsl/bxdf/common.hlsl

+    bool isValid(const scalar_type eta) NBL_CONST_MEMBER_FUNC
    {
+        fresnel::OrientedEtas<monochrome_type> orientedEta = fresnel::OrientedEtas<monochrome_type>::create(scalar_type(1.0), hlsl::promote<monochrome_type>(eta));


I don't get what you're doing here

why not take OrientedEta<monochrome_type> ?

devshgraphicsprogramming · 2025-10-20T10:01:08Z

include/nbl/builtin/hlsl/bxdf/common.hlsl

+    bool isValid(const scalar_type eta) NBL_CONST_MEMBER_FUNC
    {
-        return iso_cache.isValid(orientedEta);
+        return iso_cache.isValid(eta);


I don't get whats going on here

its super misleading to call orientedEta into eta now, it makes one think that it doesn't depend on NdotV

include/nbl/builtin/hlsl/bxdf/common.hlsl

include/nbl/builtin/hlsl/bxdf/fresnel.hlsl

devshgraphicsprogramming · 2025-10-20T13:57:58Z

include/nbl/builtin/hlsl/bxdf/base/cook_torrance_base.hlsl

-        assert(notTIR);
-        
+        const bool valid = impl::checkValid<fresnel_type, IsBSDF>::template __call<sample_type, Interaction, MicrofacetCache>(_f, _sample, interaction, cache);
+        assert(valid);


add a comment qhy

devshgraphicsprogramming · 2025-10-20T14:03:29Z

include/nbl/builtin/hlsl/bxdf/base/cook_torrance_base.hlsl

    }

    template<typename C=bool_constant<!IsBSDF> >
    enable_if_t<C::value && !IsBSDF, sample_type> generate(NBL_CONST_REF_ARG(anisotropic_interaction_type) interaction, const vector2_type u, NBL_REF_ARG(anisocache_type) cache)


NBL_FUNC_REQuirES is better syntax than the enable_if

btw why do you still have two generate ? where you could have a const conditional_t<IsBSDF,vector3_type,vector2_type> u

Continues: #930 (comment)

devshgraphicsprogramming · 2025-10-20T14:26:10Z

include/nbl/builtin/hlsl/bxdf/base/cook_torrance_base.hlsl

+        const scalar_type viewShortenFactor = hlsl::mix(scalar_type(1.0), rcpEta.value[0], transmitted);
+        const scalar_type NdotL = localH.z * (VdotH * viewShortenFactor + LdotH) - NdotV * viewShortenFactor;


double check (assert) that this optimized NdotL actually computes to same as actual NdotL computed the slow way

include/nbl/builtin/hlsl/bxdf/base/cook_torrance_base.hlsl

devshgraphicsprogramming · 2025-10-20T14:34:25Z

include/nbl/builtin/hlsl/bxdf/base/cook_torrance_base.hlsl

+        const vector3_type H = hlsl::mul(interaction.getFromTangentSpace(), localH);
        Refract<scalar_type> r = Refract<scalar_type>::create(V.getDirection(), H);
        const scalar_type LdotH = hlsl::mix(VdotH, r.getNdotT(rcpEta.value2[0]), transmitted);


you want to compute your own getNdotT (LdotH ) so you don't need the worldspace H and V until you already clear the isTransmissionPath check

That also lets you compute NdotL without using any world-space inputs

the reason why we made the whole reflect_refract_wrapper was to skip computing NdotL using worldspace dependant variables.

devshgraphicsprogramming · 2025-10-20T14:36:44Z

include/nbl/builtin/hlsl/bxdf/base/cook_torrance_base.hlsl

+    static bool __call(NBL_CONST_REF_ARG(F) fresnel, NBL_CONST_REF_ARG(Sample) _sample, NBL_CONST_REF_ARG(Interaction) interaction, NBL_CONST_REF_ARG(MicrofacetCache) cache)
    {
-        return true;
+        return _sample.getNdotL() > numeric_limits<scalar_type>::min && interaction.getNdotV() > numeric_limits<scalar_type>::min;
    }
 };

 template<class F>
-struct check_TIR_helper<F, true>
+struct checkValid<F, true>
 {
    using vector_type = typename F::vector_type;    // expect monochrome

-    template<class MicrofacetCache>
-    static bool __call(NBL_CONST_REF_ARG(F) fresnel, NBL_CONST_REF_ARG(MicrofacetCache) cache)
+    template<class Sample, class Interaction, class MicrofacetCache>
+    static bool __call(NBL_CONST_REF_ARG(F) fresnel, NBL_CONST_REF_ARG(Sample) _sample, NBL_CONST_REF_ARG(Interaction) interaction, NBL_CONST_REF_ARG(MicrofacetCache) cache)


fresnel rename to orientedFresnel !

devshgraphicsprogramming · 2025-10-20T14:40:34Z

include/nbl/builtin/hlsl/bxdf/base/cook_torrance_base.hlsl

+        else
+            return hlsl::promote<spectral_type>(0.0);


you cannot have this else if you want the compiler to throw away calls to ndf.template D

you need to add an if (DG>max_value) return 0.0; AFTER the overwrite_DG

another way is to add that sidechannel NBL_REF(bool) on the D and Dcorrelated functions which tells us if PDF is inf for this sample

devshgraphicsprogramming · 2025-10-20T14:47:52Z

include/nbl/builtin/hlsl/bxdf/base/cook_torrance_base.hlsl

+    template<typename C=bool_constant<!HasOverwrite> >
+    static enable_if_t<C::value && !HasOverwrite, void> __call(NBL_REF_ARG(scalar_type) DG, N ndf, NBL_CONST_REF_ARG(g2g1_query_type) query, NBL_CONST_REF_ARG(quant_query_type) quant_query, NBL_CONST_REF_ARG(LS) _sample, NBL_CONST_REF_ARG(Interaction) interaction, NBL_CONST_REF_ARG(MicrofacetCache) cache)
+    {
+    }
+    template<typename C=bool_constant<HasOverwrite> >
+    static enable_if_t<C::value && HasOverwrite, void> __call(NBL_REF_ARG(scalar_type) DG, N ndf, NBL_CONST_REF_ARG(g2g1_query_type) query, NBL_CONST_REF_ARG(quant_query_type) quant_query, NBL_CONST_REF_ARG(LS) _sample, NBL_CONST_REF_ARG(Interaction) interaction, NBL_CONST_REF_ARG(MicrofacetCache) cache)
+    {
+        quant_type dg = ndf.template Dcorrelated<LS, Interaction, MicrofacetCache>(query, quant_query, _sample, interaction, cache);
+        DG = dg.projectedLightMeasure;
+    }


cool, but if you're gonna do it this way, you could even stick the methods in SCookTorrance which would save you a lot of typedef aliases

which kinda gives me an idea that all the impl functors could be done and hidden this way

devshgraphicsprogramming · 2025-10-20T15:03:07Z

include/nbl/builtin/hlsl/bxdf/ndf/ggx.hlsl

+        if (dg < bit_cast<scalar_type>(numeric_limits<scalar_type>::infinity))
+            dg *= correlated_wo_numerator<LS, Interaction>(query, _sample, interaction);
+        else
+            dg = scalar_type(0.0);


you can already return 0,0 for both and call it a day

…ng invalid

devshgraphicsprogramming · 2025-10-25T09:32:37Z

include/nbl/builtin/hlsl/bxdf/base/cook_torrance_base.hlsl

-        const scalar_type NdotL = localH.z * (VdotH * viewShortenFactor + LdotH) - NdotV * viewShortenFactor;
-        // VNDF sampling guarantees that `VdotH` has same sign as `NdotV`
-        // and `transmitted` controls the sign of `LdotH` relative to `VdotH` by construction (reflect -> same sign, or refract -> opposite sign)
+        const vector3_type H = hlsl::normalize(hlsl::mul(interaction.getFromTangentSpace(), localH));


you should assert the length of H is already near 1 and the matrix is orthonormal

devshgraphicsprogramming · 2025-10-25T09:33:02Z

include/nbl/builtin/hlsl/bxdf/base/cook_torrance_base.hlsl

+            bxdf::ReflectRefract<scalar_type> rr;
+            rr.refract = r;
+            L = V.reflectRefract(rr, transmitted, rcpEta.value[0]);
+            L.direction = hlsl::normalize(L.direction);


result of refraction should be normalized already when inputs are normalized

keptsecret added 23 commits September 4, 2025 10:46

changes to dual measure quant, old stuff commented out for now

5218545

create query structs in beckmann ndf, removed query concepts, funcs r…

f7525af

…eturn quants

did the same for ggx ndf, some fixes to beckmann

c50db68

Merge branch 'master' into bxdf_fixes_cook_torrance

c49bedb

fixes to ndf concept

319c954

moved eval, quotient/pdf into cook torrance base

4eeacf1

moved generate H into ndfs, generate impl in cook torrance base

e7bc784

use new cook torrance base in microfacet bxdfs

72226bb

Merge branch 'master' into bxdf_fixes_cook_torrance

f34b348

numerous typo bug fixes

ad13044

fixes to ggx ndf

d8d2116

fixed conductor fresnel naming ior as eta

238b08e

minor fixes to beckmann ndf, removed obsolete comments

8bc707f

added luminosity contribution hint to cook torrance bsdfs

9fb28cf

added aniso overloads to isotropic bxdf methods

4cb8a0a

checks for invalid generate sample

392dc31

removed obsolete commented out stuff

02de86e

added cartesian-polar conversions from unit tests

4a7f532

Merge branch 'fix_rotation_mat' into bxdf_fixes_cook_torrance

80c4f67

Merge branch 'master' into bxdf_fixes_cook_torrance

6233bd1

vector hashes

ce5fbac

Merge branch 'fix_rotation_mat' into bxdf_fixes_cook_torrance

47e814b

Merge branch 'master' into bxdf_fixes_cook_torrance

c983975

devshgraphicsprogramming commented Sep 16, 2025

View reviewed changes

include/nbl/builtin/hlsl/bxdf/common.hlsl Outdated Show resolved Hide resolved

devshgraphicsprogramming commented Sep 16, 2025

View reviewed changes

include/nbl/builtin/hlsl/math/functions.hlsl Outdated Show resolved Hide resolved

devshgraphicsprogramming commented Sep 16, 2025

View reviewed changes

include/nbl/builtin/hlsl/cpp_compat/vector.hlsl Outdated Show resolved Hide resolved

keptsecret added 2 commits September 17, 2025 15:26

fix fresnel usage in bxdf pdf

0f2ee0b

fix F calc in generate

b5f02e6

devshgraphicsprogramming commented Sep 17, 2025

View reviewed changes

include/nbl/builtin/hlsl/bxdf/fresnel.hlsl Outdated Show resolved Hide resolved

devshgraphicsprogramming commented Sep 17, 2025

View reviewed changes

include/nbl/builtin/hlsl/bxdf/cook_torrance_base.hlsl Outdated Show resolved Hide resolved

keptsecret added 10 commits October 9, 2025 16:55

removed passthrough funcs, template directly on required interaction/…

444b656

…cache with defaults

precompute more stuff in quant query

971e140

various fixes to ggx ndf, cook torrance base

6a0e28d

change ndf type alias macro usage

f93772e

overwrite DG concept, added Dcorrelated to ggx ndf with optimizations

3519866

fix pdf call

2a7cf4b

minor fixes to cook torrance generate

79aea6d

eval transmitted should do 1-F

5229d25

fixes to generateH, VdotL calculation

a08fdfc

minor fixes to ndf

a77d59b

devshgraphicsprogramming commented Oct 20, 2025

View reviewed changes

include/nbl/builtin/hlsl/bxdf/common.hlsl Outdated Show resolved Hide resolved

devshgraphicsprogramming commented Oct 20, 2025

View reviewed changes

include/nbl/builtin/hlsl/bxdf/fresnel.hlsl Show resolved Hide resolved

devshgraphicsprogramming commented Oct 20, 2025

View reviewed changes

include/nbl/builtin/hlsl/bxdf/base/cook_torrance_base.hlsl Outdated Show resolved Hide resolved

devshgraphicsprogramming commented Oct 20, 2025

View reviewed changes

fixes to VdotL calculation, contigency for beckmann vndf sampling bei…

309e216

…ng invalid

devshgraphicsprogramming commented Oct 25, 2025

View reviewed changes

keptsecret added 4 commits October 27, 2025 11:04

use beta function in G2 correlated as well

c172b19

fixes ggx G2 using beta

ee864a6

pdf should also depend on surface interaction

b01ebf7

account for non-symmetry in transmission, added PathOrigin

36c1f4b

		const scalar_type viewShortenFactor = hlsl::mix(scalar_type(1.0), rcpEta.value[0], transmitted);
		const scalar_type NdotL = localH.z * (VdotH * viewShortenFactor + LdotH) - NdotV * viewShortenFactor;

Uh oh!

Bxdf fixes cook torrance #930

Are you sure you want to change the base?

Bxdf fixes cook torrance #930

Conversation

devshgraphicsprogramming commented Sep 16, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Description

Testing

TODO list:

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

devshgraphicsprogramming Oct 20, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

devshgraphicsprogramming commented Sep 16, 2025 •

edited

Loading

devshgraphicsprogramming Oct 20, 2025 •

edited

Loading