Eevee: AO: Removes samples and denoise options.

This comes with a lot of code simplification that leads to a small performance improvement.

12 files changed, 70 insertions, 208 deletions

diff --git a/source/blender/draw/engines/eevee/eevee_effects.c b/source/blender/draw/engines/eevee/eevee_effects.c
index 5322f84679b..c09e180b36d 100644
--- a/source/blender/draw/engines/eevee/eevee_effects.c
+++ b/source/blender/draw/engines/eevee/eevee_effects.c
@@ -456,6 +456,9 @@ void EEVEE_draw_effects(EEVEE_Data *vedata)
 			if (stl->g_data->gtao_horizons_debug) DRW_transform_to_display(stl->g_data->gtao_horizons_debug);
 			break;
 		case 7:
+			if (txl->gtao_horizons) DRW_transform_to_display(txl->gtao_horizons);
+			break;
+		case 8:
 			if (txl->sss_data) DRW_transform_to_display(txl->sss_data);
 			break;
 		default:
diff --git a/source/blender/draw/engines/eevee/eevee_engine.c b/source/blender/draw/engines/eevee/eevee_engine.c
index 678501528c8..b59f2cd8c0b 100644
--- a/source/blender/draw/engines/eevee/eevee_engine.c
+++ b/source/blender/draw/engines/eevee/eevee_engine.c
@@ -381,12 +381,10 @@ static void eevee_view_layer_settings_create(RenderEngine *UNUSED(engine), IDPro
 
 	BKE_collection_engine_property_add_bool(props, "gtao_enable", false);
 	BKE_collection_engine_property_add_bool(props, "gtao_use_bent_normals", true);
-	BKE_collection_engine_property_add_bool(props, "gtao_denoise", true);
 	BKE_collection_engine_property_add_bool(props, "gtao_bounce", true);
 	BKE_collection_engine_property_add_float(props, "gtao_distance", 0.2f);
 	BKE_collection_engine_property_add_float(props, "gtao_factor", 1.0f);
 	BKE_collection_engine_property_add_float(props, "gtao_quality", 0.25f);
-	BKE_collection_engine_property_add_int(props, "gtao_samples", 2);
 
 	BKE_collection_engine_property_add_bool(props, "dof_enable", false);
 	BKE_collection_engine_property_add_float(props, "bokeh_max_size", 100.0f);
diff --git a/source/blender/draw/engines/eevee/eevee_lightprobes.c b/source/blender/draw/engines/eevee/eevee_lightprobes.c
index 7edf9013239..dd49ae0481e 100644
--- a/source/blender/draw/engines/eevee/eevee_lightprobes.c
+++ b/source/blender/draw/engines/eevee/eevee_lightprobes.c
@@ -1113,7 +1113,7 @@ static void render_scene_to_probe(
 	EEVEE_StorageList *stl = vedata->stl;
 	EEVEE_LightProbesInfo *pinfo = sldata->probes;
 
-	float winmat[4][4], wininv[4][4], posmat[4][4], tmp_ao_dist, tmp_ao_samples, tmp_ao_settings;
+	float winmat[4][4], wininv[4][4], posmat[4][4], tmp_ao_dist, tmp_ao_settings;
 
 	unit_m4(posmat);
 
@@ -1127,7 +1127,6 @@ static void render_scene_to_probe(
 
 	/* Disable AO until we find a way to hide really bad discontinuities between cubefaces. */
 	tmp_ao_dist = stl->effects->ao_dist;
-	tmp_ao_samples = stl->effects->ao_samples;
 	tmp_ao_settings = stl->effects->ao_settings;
 	stl->effects->ao_settings = 0.0f; /* Disable AO */
 
@@ -1214,7 +1213,6 @@ static void render_scene_to_probe(
 	stl->g_data->minzbuffer = tmp_minz;
 	txl->maxzbuffer = tmp_maxz;
 	stl->effects->ao_dist = tmp_ao_dist;
-	stl->effects->ao_samples = tmp_ao_samples;
 	stl->effects->ao_settings = tmp_ao_settings;
 }
 
diff --git a/source/blender/draw/engines/eevee/eevee_materials.c b/source/blender/draw/engines/eevee/eevee_materials.c
index 31cda5936ad..f4bef9afed1 100644
--- a/source/blender/draw/engines/eevee/eevee_materials.c
+++ b/source/blender/draw/engines/eevee/eevee_materials.c
@@ -395,11 +395,10 @@ static void add_standard_uniforms(
 
 	if (vedata->stl->effects->use_ao) {
 		DRW_shgroup_uniform_buffer(shgrp, "horizonBuffer", &vedata->txl->gtao_horizons);
-		DRW_shgroup_uniform_ivec2(shgrp, "aoHorizonTexSize", (int *)vedata->stl->effects->ao_texsize, 1);
 	}
 	else {
-		/* Use shadow_pool as fallback to avoid sampling problem on certain platform, see: T52593 */
-		DRW_shgroup_uniform_buffer(shgrp, "horizonBuffer", &sldata->shadow_pool);
+		/* Use maxzbuffer as fallback to avoid sampling problem on certain platform, see: T52593 */
+		DRW_shgroup_uniform_buffer(shgrp, "horizonBuffer", &vedata->txl->maxzbuffer);
 	}
 
 	if (vedata->stl->effects->use_volumetrics && use_alpha_blend) {
diff --git a/source/blender/draw/engines/eevee/eevee_occlusion.c b/source/blender/draw/engines/eevee/eevee_occlusion.c
index 02b52f75e8a..0ae5fb91db3 100644
--- a/source/blender/draw/engines/eevee/eevee_occlusion.c
+++ b/source/blender/draw/engines/eevee/eevee_occlusion.c
@@ -72,7 +72,9 @@ int EEVEE_occlusion_init(EEVEE_ViewLayerData *UNUSED(sldata), EEVEE_Data *vedata
 
 	const DRWContextState *draw_ctx = DRW_context_state_get();
 	ViewLayer *view_layer = draw_ctx->view_layer;
-	IDProperty *props = BKE_view_layer_engine_evaluated_get(view_layer, COLLECTION_MODE_NONE, RE_engine_id_BLENDER_EEVEE);
+	IDProperty *props = BKE_view_layer_engine_evaluated_get(view_layer,
+	                                                        COLLECTION_MODE_NONE,
+	                                                        RE_engine_id_BLENDER_EEVEE);
 
 	if (BKE_collection_engine_property_value_get_bool(props, "gtao_enable")) {
 		const float *viewport_size = DRW_viewport_size_get();
@@ -85,8 +87,6 @@ int EEVEE_occlusion_init(EEVEE_ViewLayerData *UNUSED(sldata), EEVEE_Data *vedata
 		effects->ao_dist = BKE_collection_engine_property_value_get_float(props, "gtao_distance");
 		effects->ao_factor = BKE_collection_engine_property_value_get_float(props, "gtao_factor");
 		effects->ao_quality = 1.0f - BKE_collection_engine_property_value_get_float(props, "gtao_quality");
-		effects->ao_samples = BKE_collection_engine_property_value_get_int(props, "gtao_samples");
-		effects->ao_samples_inv = 1.0f / effects->ao_samples;
 
 		effects->ao_settings = 1.0; /* USE_AO */
 		if (BKE_collection_engine_property_value_get_bool(props, "gtao_use_bent_normals")) {
@@ -101,29 +101,10 @@ int EEVEE_occlusion_init(EEVEE_ViewLayerData *UNUSED(sldata), EEVEE_Data *vedata
 		effects->ao_texsize[0] = ((int)viewport_size[0]);
 		effects->ao_texsize[1] = ((int)viewport_size[1]);
 
-		/* Round up to multiple of 2 */
-		if ((effects->ao_texsize[0] & 0x1) != 0) {
-			effects->ao_texsize[0] += 1;
-		}
-		if ((effects->ao_texsize[1] & 0x1) != 0) {
-			effects->ao_texsize[1] += 1;
-		}
-
-		CLAMP(effects->ao_samples, 1, 32);
-
-		if (effects->hori_tex_layers != effects->ao_samples) {
-			DRW_TEXTURE_FREE_SAFE(txl->gtao_horizons);
-		}
-
-		if (txl->gtao_horizons == NULL) {
-			effects->hori_tex_layers = effects->ao_samples;
-			txl->gtao_horizons = DRW_texture_create_2D_array((int)viewport_size[0], (int)viewport_size[1], effects->hori_tex_layers, DRW_TEX_RG_8, 0, NULL);
-		}
-
-		DRWFboTexture tex = {&txl->gtao_horizons, DRW_TEX_RG_8, 0};
+		DRWFboTexture tex = {&txl->gtao_horizons, DRW_TEX_RGBA_8, 0};
 
 		DRW_framebuffer_init(&fbl->gtao_fb, &draw_engine_eevee_type,
-		                    effects->ao_texsize[0], effects->ao_texsize[1],
+		                    (int)viewport_size[0], (int)viewport_size[1],
 		                    &tex, 1);
 
 		if (G.debug_value == 6) {
@@ -174,8 +155,6 @@ void EEVEE_occlusion_cache_init(EEVEE_ViewLayerData *UNUSED(sldata), EEVEE_Data
 		DRW_shgroup_uniform_vec4(grp, "viewvecs[0]", (float *)stl->g_data->viewvecs, 2);
 		DRW_shgroup_uniform_vec2(grp, "mipRatio[0]", (float *)stl->g_data->mip_ratio, 10);
 		DRW_shgroup_uniform_vec4(grp, "aoParameters[0]", &stl->effects->ao_dist, 2);
-		DRW_shgroup_uniform_float(grp, "sampleNbr", &stl->effects->ao_sample_nbr, 1);
-		DRW_shgroup_uniform_ivec2(grp, "aoHorizonTexSize", (int *)stl->effects->ao_texsize, 1);
 		DRW_shgroup_uniform_texture(grp, "utilTex", EEVEE_materials_get_util_tex());
 		DRW_shgroup_call_add(grp, quad, NULL);
 
@@ -187,8 +166,6 @@ void EEVEE_occlusion_cache_init(EEVEE_ViewLayerData *UNUSED(sldata), EEVEE_Data
 		DRW_shgroup_uniform_vec4(grp, "viewvecs[0]", (float *)stl->g_data->viewvecs, 2);
 		DRW_shgroup_uniform_vec2(grp, "mipRatio[0]", (float *)stl->g_data->mip_ratio, 10);
 		DRW_shgroup_uniform_vec4(grp, "aoParameters[0]", &stl->effects->ao_dist, 2);
-		DRW_shgroup_uniform_float(grp, "sampleNbr", &stl->effects->ao_sample_nbr, 1);
-		DRW_shgroup_uniform_ivec2(grp, "aoHorizonTexSize", (int *)stl->effects->ao_texsize, 1);
 		DRW_shgroup_uniform_texture(grp, "utilTex", EEVEE_materials_get_util_tex());
 		DRW_shgroup_call_add(grp, quad, NULL);
 
@@ -202,7 +179,6 @@ void EEVEE_occlusion_cache_init(EEVEE_ViewLayerData *UNUSED(sldata), EEVEE_Data
 			DRW_shgroup_uniform_vec4(grp, "viewvecs[0]", (float *)stl->g_data->viewvecs, 2);
 			DRW_shgroup_uniform_vec2(grp, "mipRatio[0]", (float *)stl->g_data->mip_ratio, 10);
 			DRW_shgroup_uniform_vec4(grp, "aoParameters[0]", &stl->effects->ao_dist, 2);
-			DRW_shgroup_uniform_ivec2(grp, "aoHorizonTexSize", (int *)stl->effects->ao_texsize, 1);
 			DRW_shgroup_uniform_texture(grp, "utilTex", EEVEE_materials_get_util_tex());
 			DRW_shgroup_call_add(grp, quad, NULL);
 		}
@@ -213,7 +189,6 @@ void EEVEE_occlusion_compute(
         EEVEE_ViewLayerData *UNUSED(sldata), EEVEE_Data *vedata, struct GPUTexture *depth_src, int layer)
 {
 	EEVEE_PassList *psl = vedata->psl;
-	EEVEE_TextureList *txl = vedata->txl;
 	EEVEE_FramebufferList *fbl = vedata->fbl;
 	EEVEE_StorageList *stl = vedata->stl;
 	EEVEE_EffectsInfo *effects = stl->effects;
@@ -223,20 +198,13 @@ void EEVEE_occlusion_compute(
 		effects->ao_src_depth = depth_src;
 		effects->ao_depth_layer = layer;
 
-		for (effects->ao_sample_nbr = 0.0;
-		     effects->ao_sample_nbr < effects->ao_samples;
-		     ++effects->ao_sample_nbr)
-		{
-			DRW_framebuffer_texture_detach(txl->gtao_horizons);
-			DRW_framebuffer_texture_layer_attach(fbl->gtao_fb, txl->gtao_horizons, 0, (int)effects->ao_sample_nbr, 0);
-			DRW_framebuffer_bind(fbl->gtao_fb);
-
-			if (layer >= 0) {
-				DRW_draw_pass(psl->ao_horizon_search_layer);
-			}
-			else {
-				DRW_draw_pass(psl->ao_horizon_search);
-			}
+		DRW_framebuffer_bind(fbl->gtao_fb);
+
+		if (layer >= 0) {
+			DRW_draw_pass(psl->ao_horizon_search_layer);
+		}
+		else {
+			DRW_draw_pass(psl->ao_horizon_search);
 		}
 
 		/* Restore */
diff --git a/source/blender/draw/engines/eevee/eevee_private.h b/source/blender/draw/engines/eevee/eevee_private.h
index 85878a6acc3..3c3b495f95c 100644
--- a/source/blender/draw/engines/eevee/eevee_private.h
+++ b/source/blender/draw/engines/eevee/eevee_private.h
@@ -521,7 +521,7 @@ typedef struct EEVEE_EffectsInfo {
 
 	/* Ambient Occlusion */
 	bool use_ao, use_bent_normals;
-	float ao_dist, ao_samples, ao_factor, ao_samples_inv;
+	float ao_dist, pad1, ao_factor, pad2;
 	float ao_offset, ao_bounce_fac, ao_quality, ao_settings;
 	float ao_sample_nbr;
 	int ao_texsize[2], hori_tex_layers;
diff --git a/source/blender/draw/engines/eevee/eevee_screen_raytrace.c b/source/blender/draw/engines/eevee/eevee_screen_raytrace.c
index 62f75800e14..f4b8ab807f6 100644
--- a/source/blender/draw/engines/eevee/eevee_screen_raytrace.c
+++ b/source/blender/draw/engines/eevee/eevee_screen_raytrace.c
@@ -260,11 +260,10 @@ void EEVEE_screen_raytrace_cache_init(EEVEE_ViewLayerData *sldata, EEVEE_Data *v
 		DRW_shgroup_uniform_vec4(grp, "aoParameters[0]", &effects->ao_dist, 2);
 		if (effects->use_ao) {
 			DRW_shgroup_uniform_buffer(grp, "horizonBuffer", &vedata->txl->gtao_horizons);
-			DRW_shgroup_uniform_ivec2(grp, "aoHorizonTexSize", (int *)vedata->stl->effects->ao_texsize, 1);
 		}
 		else {
-			/* Use shadow_pool as fallback to avoid sampling problem on certain platform, see: T52593 */
-			DRW_shgroup_uniform_buffer(grp, "horizonBuffer", &sldata->shadow_pool);
+			/* Use ssr_specrough_input as fallback to avoid sampling problem on certain platform, see: T52593 */
+			DRW_shgroup_uniform_buffer(grp, "horizonBuffer", &txl->ssr_specrough_input);
 		}
 
 		DRW_shgroup_call_add(grp, quad, NULL);
diff --git a/source/blender/draw/engines/eevee/shaders/ambient_occlusion_lib.glsl b/source/blender/draw/engines/eevee/shaders/ambient_occlusion_lib.glsl
index 5ec72833379..302b7596cc3 100644
--- a/source/blender/draw/engines/eevee/shaders/ambient_occlusion_lib.glsl
+++ b/source/blender/draw/engines/eevee/shaders/ambient_occlusion_lib.glsl
@@ -30,87 +30,29 @@
 #ifndef UTIL_TEX
 #define UTIL_TEX
 uniform sampler2DArray utilTex;
+#define texelfetch_noise_tex(coord) texelFetch(utilTex, ivec3(ivec2(coord) % LUT_SIZE, 2.0), 0)
 #endif /* UTIL_TEX */
 
 uniform vec4 aoParameters[2];
-uniform sampler2DArray horizonBuffer;
-
-/* Cannot use textureSize(horizonBuffer) when rendering to it */
-uniform ivec2 aoHorizonTexSize;
+uniform sampler2D horizonBuffer;
 
 #define aoDistance   aoParameters[0].x
-#define aoSamples    aoParameters[0].y
+#define aoSamples    aoParameters[0].y /* UNUSED */
 #define aoFactor     aoParameters[0].z
-#define aoInvSamples aoParameters[0].w
+#define aoInvSamples aoParameters[0].w /* UNUSED */
 
 #define aoOffset     aoParameters[1].x /* UNUSED */
 #define aoBounceFac  aoParameters[1].y
 #define aoQuality    aoParameters[1].z
 #define aoSettings   aoParameters[1].w
 
+/* aoSettings flags */
 #define USE_AO            1
 #define USE_BENT_NORMAL   2
 #define USE_DENOISE       4
 
-vec2 pack_horizons(vec2 v) { return v * 0.5 + 0.5; }
-vec2 unpack_horizons(vec2 v) { return v * 2.0 - 1.0; }
-
-/* Returns the texel coordinate in horizonBuffer
- * for a given fullscreen coord */
-ivec2 get_hr_co(ivec2 fs_co)
-{
-	bvec2 quarter = notEqual(fs_co & ivec2(1), ivec2(0));
-
-	ivec2 hr_co = fs_co / 2;
-	hr_co += ivec2(quarter) * (aoHorizonTexSize / 2);
-
-	return hr_co;
-}
-
-/* Returns the texel coordinate in fullscreen (depthBuffer)
- * for a given horizonBuffer coord */
-ivec2 get_fs_co(ivec2 hr_co)
-{
-	hr_co *= 2;
-	bvec2 quarter = greaterThanEqual(hr_co, aoHorizonTexSize);
-
-	hr_co -= ivec2(quarter) * (aoHorizonTexSize - 1);
-
-	return hr_co;
-}
-
-/* Returns the phi angle in horizonBuffer
- * for a given horizonBuffer coord */
-float get_phi(ivec2 hr_co, ivec2 fs_co, float sample)
-{
-	bvec2 quarter = greaterThanEqual(hr_co, aoHorizonTexSize / 2);
-	ivec2 tex_co = ((int(aoSettings) & USE_DENOISE) != 0) ? hr_co - ivec2(quarter) * (aoHorizonTexSize / 2) : fs_co;
-	float blue_noise = texture(utilTex, vec3((vec2(tex_co) + 0.5) / LUT_SIZE, 2.0)).r;
-
-	float phi = sample * aoInvSamples;
-
-	if ((int(aoSettings) & USE_DENOISE) != 0) {
-		/* Interleaved jitter for spatial 2x2 denoising */
-		phi += 0.25 * aoInvSamples * (float(quarter.x) + 2.0 * float(quarter.y));
-		blue_noise *= 0.25;
-	}
-	/* Blue noise is scaled to cover the rest of the range. */
-	phi += aoInvSamples * blue_noise;
-	phi *= M_PI;
-
-	return phi;
-}
-
-/* Returns direction jittered offset for a given fullscreen coord */
-float get_offset(ivec2 fs_co, float sample)
-{
-	float offset = sample * aoInvSamples;
-
-	/* Interleaved jitter for spatial 2x2 denoising */
-	offset += 0.25 * dot(vec2(1.0), vec2(fs_co & 1));
-	offset += texture(utilTex, vec3((vec2(fs_co / 2) + 0.5 + 16.0) / LUT_SIZE, 2.0)).r;
-	return offset;
-}
+vec4 pack_horizons(vec4 v) { return v * 0.5 + 0.5; }
+vec4 unpack_horizons(vec4 v) { return v * 2.0 - 1.0; }
 
 /* Returns maximum screen distance an AO ray can travel for a given view depth */
 vec2 get_max_dir(float view_depth)
@@ -120,6 +62,13 @@ vec2 get_max_dir(float view_depth)
 	return vec2(ProjectionMatrix[0][0], ProjectionMatrix[1][1]) * max_dist;
 }
 
+vec2 get_ao_dir(float jitter)
+{
+	/* Only half a turn because we integrate in slices. */
+	jitter *= M_PI;
+	return vec2(cos(jitter), sin(jitter));
+}
+
 void get_max_horizon_grouped(vec4 co1, vec4 co2, vec3 x, float lod, inout float h)
 {
 	co1 *= mipRatio[int(lod + 1.0)].xyxy; /* +1 because we are using half res top level */
@@ -161,10 +110,8 @@ void get_max_horizon_grouped(vec4 co1, vec4 co2, vec3 x, float lod, inout float
 	h = mix(h, max(h, s_h.w), blend.w);
 }
 
-vec2 search_horizon_sweep(float phi, vec3 pos, vec2 uvs, float jitter, vec2 max_dir)
+vec2 search_horizon_sweep(vec2 t_phi, vec3 pos, vec2 uvs, float jitter, vec2 max_dir)
 {
-	vec2 t_phi = vec2(cos(phi), sin(phi)); /* Screen space direction */
-
 	max_dir *= max_v2(abs(t_phi));
 
 	/* Convert to pixel space. */
@@ -214,11 +161,8 @@ vec2 search_horizon_sweep(float phi, vec3 pos, vec2 uvs, float jitter, vec2 max_
 	return h;
 }
 
-void integrate_slice(vec3 normal, float phi, vec2 horizons, inout float visibility, inout vec3 bent_normal)
+void integrate_slice(vec3 normal, vec2 t_phi, vec2 horizons, inout float visibility, inout vec3 bent_normal)
 {
-	/* TODO OPTI Could be precomputed. */
-	vec2 t_phi = vec2(cos(phi), sin(phi)); /* Screen space direction */
-
 	/* Projecting Normal to Plane P defined by t_phi and omega_o */
 	vec3 np = vec3(t_phi.y, -t_phi.x, 0.0); /* Normal vector to Integration plane */
 	vec3 t = vec3(-t_phi, 0.0);
@@ -251,71 +195,43 @@ void integrate_slice(vec3 normal, float phi, vec2 horizons, inout float visibili
 	bent_normal += vec3(sin(b_angle) * -t_phi, cos(b_angle) * 0.5);
 }
 
-void denoise_ao(vec3 normal, float frag_depth, inout float visibility, inout vec3 bent_normal)
+void gtao_deferred(vec3 normal, vec3 position, vec4 noise, float frag_depth, out float visibility, out vec3 bent_normal)
 {
-	vec2 d_sign = vec2(ivec2(gl_FragCoord.xy) & 1) - 0.5;
-
-	if ((int(aoSettings) & USE_DENOISE) == 0) {
-		d_sign *= 0.0;
-	}
-
-	/* 2x2 Bilateral Filter using derivatives. */
-	vec2 n_step = step(-0.2, -abs(vec2(length(dFdx(normal)), length(dFdy(normal)))));
-	vec2 z_step = step(-0.1, -abs(vec2(dFdx(frag_depth), dFdy(frag_depth))));
-
-	visibility -= dFdx(visibility) * d_sign.x * z_step.x * n_step.x;
-	visibility -= dFdy(visibility) * d_sign.y * z_step.y * n_step.y;
+	/* Fetch early, hide latency! */
+	vec4 horizons = texelFetch(horizonBuffer, ivec2(gl_FragCoord.xy), 0);
 
-	bent_normal -= dFdx(bent_normal) * d_sign.x * z_step.x * n_step.x;
-	bent_normal -= dFdy(bent_normal) * d_sign.y * z_step.y * n_step.y;
-}
-
-void gtao_deferred(vec3 normal, vec3 position, float frag_depth, out float visibility, out vec3 bent_normal)
-{
+	vec4 dirs;
+	dirs.xy = get_ao_dir(noise.x * 0.5);
+	dirs.zw = get_ao_dir(noise.x * 0.5 + 0.5);
 	vec2 uvs = get_uvs_from_view(position);
 
-	vec4 texel_size = vec4(-1.0, -1.0, 1.0, 1.0) / vec2(textureSize(depthBuffer, 0)).xyxy;
-
-	ivec2 fs_co = ivec2(gl_FragCoord.xy);
-	ivec2 hr_co = get_hr_co(fs_co);
-
 	bent_normal = vec3(0.0);
 	visibility = 0.0;
 
-	for (float i = 0.0; i < MAX_PHI_STEP; i++) {
-		if (i >= aoSamples) break;
+	horizons = unpack_horizons(horizons);
 
-		vec2 horiz = unpack_horizons(texelFetch(horizonBuffer, ivec3(hr_co, int(i)), 0).rg);
-		float phi = get_phi(hr_co, fs_co, i);
+	integrate_slice(normal, dirs.xy, horizons.xy, visibility, bent_normal);
+	integrate_slice(normal, dirs.zw, horizons.zw, visibility, bent_normal);
 
-		integrate_slice(normal, phi, horiz.xy, visibility, bent_normal);
-	}
+	visibility *= 0.5; /* We integrated 2 slices. */
 
-	visibility *= aoInvSamples;
 	bent_normal = normalize(bent_normal);
 }
 
-void gtao(vec3 normal, vec3 position, vec2 noise, out float visibility, out vec3 bent_normal)
+void gtao(vec3 normal, vec3 position, vec4 noise, out float visibility, out vec3 bent_normal)
 {
 	vec2 uvs = get_uvs_from_view(position);
-
-	float homcco = ProjectionMatrix[2][3] * position.z + ProjectionMatrix[3][3];
-	float max_dist = aoDistance / homcco; /* Search distance */
-	vec2 max_dir = max_dist * vec2(ProjectionMatrix[0][0], ProjectionMatrix[1][1]);
+	vec2 max_dir = get_max_dir(position.z);
+	vec2 dir = get_ao_dir(noise.x);
 
 	bent_normal = vec3(0.0);
 	visibility = 0.0;
 
-	for (float i = 0.0; i < MAX_PHI_STEP; i++) {
-		if (i >= aoSamples) break;
-
-		float phi = M_PI * (i + noise.x) * aoInvSamples;
-		vec2 horizons = search_horizon_sweep(phi, position, uvs, noise.g, max_dir);
-
-		integrate_slice(normal, phi, horizons, visibility, bent_normal);
-	}
+	/* Only trace in 2 directions. May lead to a darker result but since it's mostly for
+	 * alpha blended objects that will have overdraw, we limit the performance impact. */
+	vec2 horizons = search_horizon_sweep(dir, position, uvs, noise.y, max_dir);
+	integrate_slice(normal, dir, horizons, visibility, bent_normal);
 
-	visibility *= aoInvSamples;
 	bent_normal = normalize(bent_normal);
 }
 
@@ -337,7 +253,7 @@ float gtao_multibounce(float visibility, vec3 albedo)
 }
 
 /* Use the right occlusion  */
-float occlusion_compute(vec3 N, vec3 vpos, float user_occlusion, vec2 randuv, out vec3 bent_normal)
+float occlusion_compute(vec3 N, vec3 vpos, float user_occlusion, vec4 rand, out vec3 bent_normal)
 {
 #ifndef USE_REFRACTION
 	if ((int(aoSettings) & USE_AO) > 0) {
@@ -345,11 +261,10 @@ float occlusion_compute(vec3 N, vec3 vpos, float user_occlusion, vec2 randuv, ou
 		vec3 vnor = mat3(ViewMatrix) * N;
 
 #ifdef ENABLE_DEFERED_AO
-		gtao_deferred(vnor, vpos, gl_FragCoord.z, visibility, bent_normal);
+		gtao_deferred(vnor, vpos, rand, gl_FragCoord.z, visibility, bent_normal);
 #else
-		gtao(vnor, vpos, randuv, visibility, bent_normal);
+		gtao(vnor, vpos, rand, visibility, bent_normal);
 #endif
-		denoise_ao(vnor, gl_FragCoord.z, visibility, bent_normal);
 
 		/* Prevent some problems down the road. */
 		visibility = max(1e-3, visibility);
diff --git a/source/blender/draw/engines/eevee/shaders/effect_gtao_frag.glsl b/source/blender/draw/engines/eevee/shaders/effect_gtao_frag.glsl
index 730abcc89cf..32edf709d88 100644
--- a/source/blender/draw/engines/eevee/shaders/effect_gtao_frag.glsl
+++ b/source/blender/draw/engines/eevee/shaders/effect_gtao_frag.glsl
@@ -24,15 +24,14 @@ void main()
 	vec3 bent_normal;
 	float visibility;
 
-	gtao_deferred(normal, viewPosition, depth, visibility, bent_normal);
+	vec4 noise = texelfetch_noise_tex(gl_FragCoord.xy);
 
-	denoise_ao(normal, depth, visibility, bent_normal);
+	gtao_deferred(normal, viewPosition, noise, depth, visibility, bent_normal);
 
 	FragColor = vec4(visibility);
 }
 
 #else
-uniform float sampleNbr;
 
 #ifdef LAYERED_DEPTH
 uniform sampler2DArray depthBufferLayered;
@@ -48,10 +47,7 @@ uniform int layer;
 
 void main()
 {
-	ivec2 hr_co = ivec2(gl_FragCoord.xy);
-	ivec2 fs_co = get_fs_co(hr_co);
-
-	vec2 uvs = saturate((vec2(fs_co) + 0.5) / vec2(textureSize(gtao_depthBuffer, 0).xy));
+	vec2 uvs = saturate(gl_FragCoord.xy / vec2(textureSize(gtao_depthBuffer, 0).xy));
 	float depth = gtao_textureLod(gtao_depthBuffer, uvs, 0.0).r;
 
 	if (depth == 1.0) {
@@ -64,14 +60,17 @@ void main()
 	depth = saturate(depth - 3e-6); /* Tweaked for 24bit depth buffer. */
 
 	vec3 viewPosition = get_view_space_from_depth(uvs, depth);
-
-	float phi = get_phi(hr_co, fs_co, sampleNbr);
-	float offset = get_offset(fs_co, sampleNbr);
+	vec4 noise = texelfetch_noise_tex(gl_FragCoord.xy);
 	vec2 max_dir = get_max_dir(viewPosition.z);
+	vec4 dirs;
+	dirs.xy = get_ao_dir(noise.x * 0.5);
+	dirs.zw = get_ao_dir(noise.x * 0.5 + 0.5);
 
-	FragColor.xy = search_horizon_sweep(phi, viewPosition, uvs, offset, max_dir);
+	/* Search in 4 directions. */
+	FragColor.xy = search_horizon_sweep(dirs.xy, viewPosition, uvs, noise.y, max_dir);
+	FragColor.zw = search_horizon_sweep(dirs.zw, viewPosition, uvs, noise.y, max_dir);
 
 	/* Resize output for integer texture. */
-	FragColor = pack_horizons(FragColor.xy).xyxy;
+	FragColor = pack_horizons(FragColor);
 }
 #endif
diff --git a/source/blender/draw/engines/eevee/shaders/effect_ssr_frag.glsl b/source/blender/draw/engines/eevee/shaders/effect_ssr_frag.glsl
index bc78d622093..fef55d82e96 100644
--- a/source/blender/draw/engines/eevee/shaders/effect_ssr_frag.glsl
+++ b/source/blender/draw/engines/eevee/shaders/effect_ssr_frag.glsl
@@ -199,7 +199,7 @@ void fallback_cubemap(
 
 	vec4 rand = texelfetch_noise_tex(gl_FragCoord.xy);
 	vec3 bent_normal;
-	float final_ao = occlusion_compute(N, viewPosition, 1.0, rand.rg, bent_normal);
+	float final_ao = occlusion_compute(N, viewPosition, 1.0, rand, bent_normal);
 	final_ao = specular_occlusion(dot(N, V), final_ao, roughness);
 
 	/* Starts at 1 because 0 is world probe */
diff --git a/source/blender/draw/engines/eevee/shaders/lit_surface_frag.glsl b/source/blender/draw/engines/eevee/shaders/lit_surface_frag.glsl
index d425a8575e2..a2917842d43 100644
--- a/source/blender/draw/engines/eevee/shaders/lit_surface_frag.glsl
+++ b/source/blender/draw/engines/eevee/shaders/lit_surface_frag.glsl
@@ -396,7 +396,7 @@ void CLOSURE_NAME(
 	/* ---------------------------- */
 #if defined(CLOSURE_GLOSSY) || defined(CLOSURE_DIFFUSE)
 	vec3 bent_normal;
-	float final_ao = occlusion_compute(N, viewPosition, ao, rand.zw, bent_normal);
+	float final_ao = occlusion_compute(N, viewPosition, ao, rand, bent_normal);
 #endif
 
 
diff --git a/source/blender/makesrna/intern/rna_layer.c b/source/blender/makesrna/intern/rna_layer.c
index 6cdd77dc7ce..ef3d7c0ae88 100644
--- a/source/blender/makesrna/intern/rna_layer.c
+++ b/source/blender/makesrna/intern/rna_layer.c
@@ -350,12 +350,10 @@ RNA_LAYER_ENGINE_CLAY_GET_SET_FLOAT(hair_brightness_randomness)
 /* ViewLayer settings. */
 RNA_LAYER_ENGINE_EEVEE_GET_SET_BOOL(gtao_enable)
 RNA_LAYER_ENGINE_EEVEE_GET_SET_BOOL(gtao_use_bent_normals)
-RNA_LAYER_ENGINE_EEVEE_GET_SET_BOOL(gtao_denoise)
 RNA_LAYER_ENGINE_EEVEE_GET_SET_BOOL(gtao_bounce)
 RNA_LAYER_ENGINE_EEVEE_GET_SET_FLOAT(gtao_factor)
 RNA_LAYER_ENGINE_EEVEE_GET_SET_FLOAT(gtao_quality)
 RNA_LAYER_ENGINE_EEVEE_GET_SET_FLOAT(gtao_distance)
-RNA_LAYER_ENGINE_EEVEE_GET_SET_INT(gtao_samples)
 RNA_LAYER_ENGINE_EEVEE_GET_SET_BOOL(dof_enable)
 RNA_LAYER_ENGINE_EEVEE_GET_SET_FLOAT(bokeh_max_size)
 RNA_LAYER_ENGINE_EEVEE_GET_SET_FLOAT(bokeh_threshold)
@@ -1425,13 +1423,6 @@ static void rna_def_view_layer_engine_settings_eevee(BlenderRNA *brna)
 	RNA_def_property_flag(prop, PROP_CONTEXT_UPDATE);
 	RNA_def_property_update(prop, NC_SCENE | ND_LAYER_CONTENT, "rna_ViewLayerEngineSettings_update");
 
-	prop = RNA_def_property(srna, "gtao_denoise", PROP_BOOLEAN, PROP_NONE);
-	RNA_def_property_boolean_funcs(prop, "rna_LayerEngineSettings_Eevee_gtao_denoise_get",
-	                               "rna_LayerEngineSettings_Eevee_gtao_denoise_set");
-	RNA_def_property_ui_text(prop, "Denoise", "Use denoising to filter the resulting occlusion and bent normal but exhibit 2x2 pixel blocks");
-	RNA_def_property_flag(prop, PROP_CONTEXT_UPDATE);
-	RNA_def_property_update(prop, NC_SCENE | ND_LAYER_CONTENT, "rna_ViewLayerEngineSettings_update");
-
 	prop = RNA_def_property(srna, "gtao_bounce", PROP_BOOLEAN, PROP_NONE);
 	RNA_def_property_boolean_funcs(prop, "rna_LayerEngineSettings_Eevee_gtao_bounce_get",
 	                               "rna_LayerEngineSettings_Eevee_gtao_bounce_set");
@@ -1463,14 +1454,6 @@ static void rna_def_view_layer_engine_settings_eevee(BlenderRNA *brna)
 	RNA_def_property_flag(prop, PROP_CONTEXT_UPDATE);
 	RNA_def_property_update(prop, 0, "rna_LayerCollectionEngineSettings_update");
 
-	prop = RNA_def_property(srna, "gtao_samples", PROP_INT, PROP_UNSIGNED);
-	RNA_def_property_int_funcs(prop, "rna_LayerEngineSettings_Eevee_gtao_samples_get",
-	                           "rna_LayerEngineSettings_Eevee_gtao_samples_set", NULL);
-	RNA_def_property_ui_text(prop, "Samples", "Number of samples to take to compute occlusion");
-	RNA_def_property_range(prop, 2, 32);
-	RNA_def_property_flag(prop, PROP_CONTEXT_UPDATE);
-	RNA_def_property_update(prop, NC_SCENE | ND_LAYER_CONTENT, "rna_ViewLayerEngineSettings_update");
-
 	/* Depth of Field */
 	prop = RNA_def_property(srna, "dof_enable", PROP_BOOLEAN, PROP_NONE);
 	RNA_def_property_boolean_funcs(prop, "rna_LayerEngineSettings_Eevee_dof_enable_get",