Skip to content
Snippets Groups Projects
Commit 9eba0b91 authored by Alexander Rose's avatar Alexander Rose
Browse files

add smaa antialiasing option (new default)

parent 58bc6722
No related branches found
No related tags found
No related merge requests found
Show whitespace changes
Inline Side-by-side
Showing
with 759 additions and 9 deletions
src/mol-canvas3d/passes/postprocessing.ts
+ 26
− 5
View file @ 9eba0b91
......@@ -21,6 +21,7 @@ import quad_vert from '../../mol-gl/shader/quad.vert';
import postprocessing_frag from '../../mol-gl/shader/postprocessing.frag';
import { StereoCamera } from '../camera/stereo';
import { FxaaParams, FxaaPass } from './fxaa';
import { SmaaParams, SmaaPass } from './smaa';
const PostprocessingSchema = {
...QuadSchema,
......@@ -96,10 +97,11 @@ export const PostprocessingParams = {
}),
off: PD.Group({})
}, { cycle: true, description: 'Draw outline around 3D objects' }),
antialiasing: PD.MappedStatic('on', {
on: PD.Group(FxaaParams),
antialiasing: PD.MappedStatic('smaa', {
fxaa: PD.Group(FxaaParams),
smaa: PD.Group(SmaaParams),
off: PD.Group({})
}, { cycle: true, description: 'Smooth pixel edges' }),
}, { options: [['fxaa', 'FXAA'], ['smaa', 'SMAA'], ['off', 'Off']], description: 'Smooth pixel edges' }),
};
export type PostprocessingProps = PD.Values<typeof PostprocessingParams>
......@@ -113,6 +115,7 @@ export class PostprocessingPass {
private readonly tmpTarget: RenderTarget
private readonly renderable: PostprocessingRenderable
private readonly fxaa: FxaaPass
private readonly smaa: SmaaPass
constructor(private webgl: WebGLContext, private drawPass: DrawPass) {
const { colorTarget, depthTexture } = drawPass;
......@@ -123,6 +126,7 @@ export class PostprocessingPass {
this.tmpTarget = webgl.createRenderTarget(width, height, false, 'uint8', 'linear');
this.renderable = getPostprocessingRenderable(webgl, colorTarget.texture, depthTexture);
this.fxaa = new FxaaPass(webgl, this.tmpTarget.texture);
this.smaa = new SmaaPass(webgl, this.tmpTarget.texture);
}
syncSize() {
......@@ -135,6 +139,8 @@ export class PostprocessingPass {
this.tmpTarget.setSize(width, height);
ValueCell.update(this.renderable.values.uTexSize, Vec2.set(this.renderable.values.uTexSize.ref.value, width, height));
this.fxaa.setSize(width, height);
if (this.smaa.supported) this.smaa.setSize(width, height);
}
}
......@@ -204,7 +210,7 @@ export class PostprocessingPass {
}
private _renderFxaa(camera: ICamera, toDrawingBuffer: boolean, props: PostprocessingProps) {
if (props.antialiasing.name !== 'on') return;
if (props.antialiasing.name !== 'fxaa') return;
const input = (props.occlusion.name === 'on' || props.outline.name === 'on')
? this.tmpTarget.texture : this.drawPass.colorTarget.texture;
......@@ -220,13 +226,28 @@ export class PostprocessingPass {
this.fxaa.render();
}
private _renderSmaa(camera: ICamera, toDrawingBuffer: boolean, props: PostprocessingProps) {
if (props.antialiasing.name !== 'smaa') return;
const input = (props.occlusion.name === 'on' || props.outline.name === 'on')
? this.tmpTarget.texture : this.drawPass.colorTarget.texture;
this.smaa.update(input, props.antialiasing.params);
this.smaa.render(camera.viewport, toDrawingBuffer ? undefined : this.target);
}
private _render(camera: ICamera, toDrawingBuffer: boolean, props: PostprocessingProps) {
if (props.occlusion.name === 'on' || props.outline.name === 'on' || props.antialiasing.name === 'off') {
this._renderPostprocessing(camera, toDrawingBuffer, props);
}
if (props.antialiasing.name === 'on') {
if (props.antialiasing.name === 'fxaa') {
this._renderFxaa(camera, toDrawingBuffer, props);
} else if (props.antialiasing.name === 'smaa') {
if (!this.smaa.supported) {
throw new Error('SMAA not supported, missing "HTMLImageElement"');
}
this._renderSmaa(camera, toDrawingBuffer, props);
}
}
......
src/mol-canvas3d/passes/smaa.ts 0 → 100644
+ 245
− 0
View file @ 9eba0b91
/**
* Copyright (c) 2020 mol* contributors, licensed under MIT, See LICENSE file for more info.
*
* @author Alexander Rose <alexander.rose@weirdbyte.de>
*/
import { QuadSchema, QuadValues } from '../../mol-gl/compute/util';
import { ComputeRenderable, createComputeRenderable } from '../../mol-gl/renderable';
import { TextureSpec, UniformSpec, DefineSpec, Values } from '../../mol-gl/renderable/schema';
import { ShaderCode } from '../../mol-gl/shader-code';
import { WebGLContext } from '../../mol-gl/webgl/context';
import { createComputeRenderItem } from '../../mol-gl/webgl/render-item';
import { RenderTarget } from '../../mol-gl/webgl/render-target';
import { createTexture, loadImageTexture, Texture } from '../../mol-gl/webgl/texture';
import { Vec2 } from '../../mol-math/linear-algebra';
import { ValueCell } from '../../mol-util';
import { ParamDefinition as PD } from '../../mol-util/param-definition';
import { blend_vert } from '../../mol-gl/shader/smaa/blend.vert';
import { blend_frag } from '../../mol-gl/shader/smaa/blend.frag';
import { weights_vert } from '../../mol-gl/shader/smaa/weights.vert';
import { weights_frag } from '../../mol-gl/shader/smaa/weights.frag';
import { edges_vert } from '../../mol-gl/shader/smaa/edges.vert';
import { edges_frag } from '../../mol-gl/shader/smaa/edges.frag';
import { Viewport } from '../camera/util';
import { isDebugMode } from '../../mol-util/debug';
export const SmaaParams = {
edgeThreshold:PD.Numeric(0.1, { min: 0.05, max: 0.2, step: 0.01 }),
maxSearchSteps:PD.Numeric(18, { min: 0, max: 98, step: 1 }),
};
export type SmaaProps = PD.Values<typeof SmaaParams>
export class SmaaPass {
private readonly edgesTarget: RenderTarget
private readonly weightsTarget: RenderTarget
private readonly edgesRenderable: EdgesRenderable
private readonly weightsRenderable: WeightsRenderable
private readonly blendRenderable: BlendRenderable
private _supported = false;
get supported() {
return this._supported;
}
constructor(private webgl: WebGLContext, input: Texture) {
if (typeof HTMLImageElement === undefined) {
if (isDebugMode) console.log(`Missing "HTMLImageElement" required for "SMAA"`);
return;
}
const width = input.getWidth();
const height = input.getHeight();
this.edgesTarget = webgl.createRenderTarget(width, height, false, 'uint8', 'linear');
this.weightsTarget = webgl.createRenderTarget(width, height, false, 'uint8', 'linear');
this.edgesRenderable = getEdgesRenderable(webgl, input);
this.weightsRenderable = getWeightsRenderable(webgl, this.edgesTarget.texture);
this.blendRenderable = getBlendRenderable(webgl, input, this.weightsTarget.texture);
this._supported = true;
}
private updateState(viewport: Viewport) {
const { gl, state } = this.webgl;
state.enable(gl.SCISSOR_TEST);
state.disable(gl.BLEND);
state.disable(gl.DEPTH_TEST);
state.depthMask(false);
const { x, y, width, height } = viewport;
gl.viewport(x, y, width, height);
gl.scissor(x, y, width, height);
state.clearColor(0, 0, 0, 1);
gl.clear(gl.COLOR_BUFFER_BIT);
}
setSize(width: number, height: number) {
const w = this.edgesTarget.getWidth();
const h = this.edgesTarget.getHeight();
if (width !== w || height !== h) {
this.edgesTarget.setSize(width, height);
this.weightsTarget.setSize(width, height);
ValueCell.update(this.edgesRenderable.values.uTexSizeInv, Vec2.set(this.edgesRenderable.values.uTexSizeInv.ref.value, 1 / width, 1 / height));
ValueCell.update(this.weightsRenderable.values.uTexSizeInv, Vec2.set(this.weightsRenderable.values.uTexSizeInv.ref.value, 1 / width, 1 / height));
ValueCell.update(this.blendRenderable.values.uTexSizeInv, Vec2.set(this.blendRenderable.values.uTexSizeInv.ref.value, 1 / width, 1 / height));
}
}
update(input: Texture, props: SmaaProps) {
let edgesNeedsUpdate = false;
if (this.edgesRenderable.values.tColor.ref.value !== input) {
ValueCell.update(this.edgesRenderable.values.tColor, input);
edgesNeedsUpdate = true;
}
if (this.edgesRenderable.values.dEdgeThreshold.ref.value !== props.edgeThreshold) {
ValueCell.update(this.edgesRenderable.values.dEdgeThreshold, props.edgeThreshold);
edgesNeedsUpdate = true;
}
if (edgesNeedsUpdate) this.edgesRenderable.update();
if (this.weightsRenderable.values.dMaxSearchSteps.ref.value !== props.maxSearchSteps) {
ValueCell.update(this.weightsRenderable.values.dMaxSearchSteps, props.maxSearchSteps);
this.weightsRenderable.update();
}
if (this.blendRenderable.values.tColor.ref.value !== input) {
ValueCell.update(this.blendRenderable.values.tColor, input);
this.blendRenderable.update();
}
}
render(viewport: Viewport, target: RenderTarget | undefined) {
this.edgesTarget.bind();
this.updateState(viewport);
this.edgesRenderable.render();
this.weightsTarget.bind();
this.updateState(viewport);
this.weightsRenderable.render();
if (target) {
target.bind();
} else {
this.webgl.unbindFramebuffer();
}
this.updateState(viewport);
this.blendRenderable.render();
}
}
//
const EdgesSchema = {
...QuadSchema,
tColor: TextureSpec('texture', 'rgba', 'ubyte', 'linear'),
uTexSizeInv: UniformSpec('v2'),
dEdgeThreshold: DefineSpec('number')
};
const EdgesShaderCode = ShaderCode('smaa-edges', edges_vert, edges_frag);
type EdgesRenderable = ComputeRenderable<Values<typeof EdgesSchema>>
function getEdgesRenderable(ctx: WebGLContext, colorTexture: Texture): EdgesRenderable {
const width = colorTexture.getWidth();
const height = colorTexture.getHeight();
const values: Values<typeof EdgesSchema> = {
...QuadValues,
tColor: ValueCell.create(colorTexture),
uTexSizeInv: ValueCell.create(Vec2.create(1 / width, 1 / height)),
dEdgeThreshold: ValueCell.create(0.1),
};
const schema = { ...EdgesSchema };
const renderItem = createComputeRenderItem(ctx, 'triangles', EdgesShaderCode, schema, values);
return createComputeRenderable(renderItem, values);
}
//
const WeightsSchema = {
...QuadSchema,
tEdges: TextureSpec('texture', 'rgba', 'ubyte', 'linear'),
tArea: TextureSpec('texture', 'rgb', 'ubyte', 'linear'),
tSearch: TextureSpec('texture', 'rgba', 'ubyte', 'nearest'),
uTexSizeInv: UniformSpec('v2'),
dMaxSearchSteps: DefineSpec('number'),
};
const WeightsShaderCode = ShaderCode('smaa-weights', weights_vert, weights_frag);
type WeightsRenderable = ComputeRenderable<Values<typeof WeightsSchema>>
function getWeightsRenderable(ctx: WebGLContext, edgesTexture: Texture): WeightsRenderable {
const width = edgesTexture.getWidth();
const height = edgesTexture.getHeight();
const areaTexture = createTexture(ctx.gl, ctx.extensions, 'image-uint8', 'rgb', 'ubyte', 'linear');
const searchTexture = createTexture(ctx.gl, ctx.extensions, 'image-uint8', 'rgba', 'ubyte', 'nearest');
const values: Values<typeof WeightsSchema> = {
...QuadValues,
tEdges: ValueCell.create(edgesTexture),
tArea: ValueCell.create(areaTexture),
tSearch: ValueCell.create(searchTexture),
uTexSizeInv: ValueCell.create(Vec2.create(1 / width, 1 / height)),
dMaxSearchSteps: ValueCell.create(8),
};
// Note: loading image textures requires `HTMLImageElement` to be available
loadImageTexture(getAreaTexture(), values.tArea, areaTexture);
loadImageTexture(getSearchTexture(), values.tSearch, searchTexture);
const schema = { ...WeightsSchema };
const renderItem = createComputeRenderItem(ctx, 'triangles', WeightsShaderCode, schema, values);
return createComputeRenderable(renderItem, values);
}
//
const BlendSchema = {
...QuadSchema,
tColor: TextureSpec('texture', 'rgba', 'ubyte', 'linear'),
tWeights: TextureSpec('texture', 'rgba', 'ubyte', 'linear'),
uTexSizeInv: UniformSpec('v2'),
};
const BlendShaderCode = ShaderCode('smaa-blend', blend_vert, blend_frag);
type BlendRenderable = ComputeRenderable<Values<typeof BlendSchema>>
function getBlendRenderable(ctx: WebGLContext, colorTexture: Texture, weightsTexture: Texture): BlendRenderable {
const width = colorTexture.getWidth();
const height = colorTexture.getHeight();
const values: Values<typeof BlendSchema> = {
...QuadValues,
tColor: ValueCell.create(colorTexture),
tWeights: ValueCell.create(weightsTexture),
uTexSizeInv: ValueCell.create(Vec2.create(1 / width, 1 / height)),
};
const schema = { ...BlendSchema };
const renderItem = createComputeRenderItem(ctx, 'triangles', BlendShaderCode, schema, values);
return createComputeRenderable(renderItem, values);
}
//
function getAreaTexture() {
return 'data:image/png;base64,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';
}
function getSearchTexture() {
return 'data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAEIAAAAhCAAAAABIXyLAAAAAOElEQVRIx2NgGAWjYBSMglEwEICREYRgFBZBqDCSLA2MGPUIVQETE9iNUAqLR5gIeoQKRgwXjwAAGn4AtaFeYLEAAAAASUVORK5CYII=';
}
\ No newline at end of file
src/mol-gl/shader/smaa/blend.frag.ts 0 → 100644
+ 66
− 0
View file @ 9eba0b91
/**
* Slightly adapted from https://github.com/mrdoob/three.js
* MIT License Copyright (c) 2010-2020 three.js authors
*
* WebGL port of Subpixel Morphological Antialiasing (SMAA) v2.8
* Preset: SMAA 1x Medium (with color edge detection)
* https://github.com/iryoku/smaa/releases/tag/v2.8
*/
export const blend_frag = `
precision highp float;
precision highp int;
precision highp sampler2D;
uniform sampler2D tWeights;
uniform sampler2D tColor;
uniform vec2 uTexSizeInv;
varying vec2 vUv;
varying vec4 vOffset[2];
vec4 SMAANeighborhoodBlendingPS(vec2 texCoord, vec4 offset[2], sampler2D colorTex, sampler2D blendTex) {
// Fetch the blending weights for current pixel:
vec4 a;
a.xz = texture2D(blendTex, texCoord).xz;
a.y = texture2D(blendTex, offset[1].zw).g;
a.w = texture2D(blendTex, offset[1].xy).a;
// Is there any blending weight with a value greater than 0.0?
if (dot(a, vec4(1.0, 1.0, 1.0, 1.0)) < 1e-5) {
return texture2D(colorTex, texCoord, 0.0);
} else {
// Up to 4 lines can be crossing a pixel (one through each edge). We
// favor blending by choosing the line with the maximum weight for each
// direction:
vec2 offset;
offset.x = a.a > a.b ? a.a : -a.b; // left vs. right
offset.y = a.g > a.r ? -a.g : a.r; // top vs. bottom // WebGL port note: Changed signs
// Then we go in the direction that has the maximum weight:
if (abs(offset.x) > abs(offset.y)) { // horizontal vs. vertical
offset.y = 0.0;
} else {
offset.x = 0.0;
}
// Fetch the opposite color and lerp by hand:
vec4 C = texture2D(colorTex, texCoord, 0.0);
texCoord += sign(offset) * uTexSizeInv;
vec4 Cop = texture2D(colorTex, texCoord, 0.0);
float s = abs(offset.x) > abs(offset.y) ? abs(offset.x) : abs(offset.y);
// WebGL port note: Added gamma correction
C.xyz = pow(C.xyz, vec3(2.2));
Cop.xyz = pow(Cop.xyz, vec3(2.2));
vec4 mixed = mix(C, Cop, s);
mixed.xyz = pow(mixed.xyz, vec3(1.0 / 2.2));
return mixed;
}
}
void main() {
gl_FragColor = SMAANeighborhoodBlendingPS(vUv, vOffset, tColor, tWeights);
}
`;
\ No newline at end of file
src/mol-gl/shader/smaa/blend.vert.ts 0 → 100644
+ 32
− 0
View file @ 9eba0b91
/**
* Slightly adapted from https://github.com/mrdoob/three.js
* MIT License Copyright (c) 2010-2020 three.js authors
*
* WebGL port of Subpixel Morphological Antialiasing (SMAA) v2.8
* Preset: SMAA 1x Medium (with color edge detection)
* https://github.com/iryoku/smaa/releases/tag/v2.8
*/
export const blend_vert = `
precision highp float;
attribute vec2 aPosition;
uniform vec2 uQuadScale;
uniform vec2 uTexSizeInv;
varying vec2 vUv;
varying vec4 vOffset[2];
void SMAANeighborhoodBlendingVS(vec2 texCoord) {
vOffset[0] = texCoord.xyxy + uTexSizeInv.xyxy * vec4(-1.0, 0.0, 0.0, 1.0); // WebGL port note: Changed sign in W component
vOffset[1] = texCoord.xyxy + uTexSizeInv.xyxy * vec4(1.0, 0.0, 0.0, -1.0); // WebGL port note: Changed sign in W component
}
void main() {
vUv = (aPosition + 1.0) * 0.5;
SMAANeighborhoodBlendingVS(vUv);
vec2 position = aPosition * uQuadScale - vec2(1.0, 1.0) + uQuadScale;
gl_Position = vec4(position, 0.0, 1.0);
}
`;
\ No newline at end of file
src/mol-gl/shader/smaa/edges.frag.ts 0 → 100644
+ 76
− 0
View file @ 9eba0b91
/**
* Slightly adapted from https://github.com/mrdoob/three.js
* MIT License Copyright (c) 2010-2020 three.js authors
*
* WebGL port of Subpixel Morphological Antialiasing (SMAA) v2.8
* Preset: SMAA 1x Medium (with color edge detection)
* https://github.com/iryoku/smaa/releases/tag/v2.8
*/
export const edges_frag = `
precision highp float;
precision highp int;
precision highp sampler2D;
uniform sampler2D tColor;
uniform vec2 uTexSizeInv;
varying vec2 vUv;
varying vec4 vOffset[3];
vec4 SMAAColorEdgeDetectionPS(vec2 texcoord, vec4 offset[3], sampler2D colorTex) {
vec2 threshold = vec2(dEdgeThreshold, dEdgeThreshold);
// Calculate color deltas:
vec4 delta;
vec3 C = texture2D(colorTex, texcoord).rgb;
vec3 Cleft = texture2D(colorTex, offset[0].xy).rgb;
vec3 t = abs(C - Cleft);
delta.x = max(max(t.r, t.g), t.b);
vec3 Ctop = texture2D(colorTex, offset[0].zw).rgb;
t = abs(C - Ctop);
delta.y = max(max(t.r, t.g), t.b);
// We do the usual threshold:
vec2 edges = step(threshold, delta.xy);
// Then discard if there is no edge:
if (dot(edges, vec2(1.0, 1.0 )) == 0.0)
discard;
// Calculate right and bottom deltas:
vec3 Cright = texture2D(colorTex, offset[1].xy).rgb;
t = abs( C - Cright );
delta.z = max(max(t.r, t.g), t.b);
vec3 Cbottom = texture2D(colorTex, offset[1].zw).rgb;
t = abs(C - Cbottom);
delta.w = max(max(t.r, t.g), t.b);
// Calculate the maximum delta in the direct neighborhood:
float maxDelta = max(max(max(delta.x, delta.y), delta.z), delta.w );
// Calculate left-left and top-top deltas:
vec3 Cleftleft = texture2D(colorTex, offset[2].xy).rgb;
t = abs( C - Cleftleft );
delta.z = max(max(t.r, t.g), t.b);
vec3 Ctoptop = texture2D(colorTex, offset[2].zw).rgb;
t = abs(C - Ctoptop);
delta.w = max(max(t.r, t.g), t.b);
// Calculate the final maximum delta:
maxDelta = max(max(maxDelta, delta.z), delta.w);
// Local contrast adaptation in action:
edges.xy *= step(0.5 * maxDelta, delta.xy);
return vec4(edges, 0.0, 0.0);
}
void main() {
gl_FragColor = SMAAColorEdgeDetectionPS(vUv, vOffset, tColor);
}
`;
\ No newline at end of file
src/mol-gl/shader/smaa/edges.vert.ts 0 → 100644
+ 33
− 0
View file @ 9eba0b91
/**
* Slightly adapted from https://github.com/mrdoob/three.js
* MIT License Copyright (c) 2010-2020 three.js authors
*
* WebGL port of Subpixel Morphological Antialiasing (SMAA) v2.8
* Preset: SMAA 1x Medium (with color edge detection)
* https://github.com/iryoku/smaa/releases/tag/v2.8
*/
export const edges_vert = `
precision highp float;
attribute vec2 aPosition;
uniform vec2 uQuadScale;
uniform vec2 uTexSizeInv;
varying vec2 vUv;
varying vec4 vOffset[ 3 ];
void SMAAEdgeDetectionVS(vec2 texCoord) {
vOffset[0] = texCoord.xyxy + uTexSizeInv.xyxy * vec4(-1.0, 0.0, 0.0, 1.0); // WebGL port note: Changed sign in W component
vOffset[1] = texCoord.xyxy + uTexSizeInv.xyxy * vec4(1.0, 0.0, 0.0, -1.0); // WebGL port note: Changed sign in W component
vOffset[2] = texCoord.xyxy + uTexSizeInv.xyxy * vec4(-2.0, 0.0, 0.0, 2.0); // WebGL port note: Changed sign in W component
}
void main() {
vUv = (aPosition + 1.0) * 0.5;
SMAAEdgeDetectionVS(vUv);
vec2 position = aPosition * uQuadScale - vec2(1.0, 1.0) + uQuadScale;
gl_Position = vec4(position, 0.0, 1.0);
}
`;
\ No newline at end of file
src/mol-gl/shader/smaa/weights.frag.ts 0 → 100644
+ 216
− 0
View file @ 9eba0b91
/**
* Slightly adapted from https://github.com/mrdoob/three.js
* MIT License Copyright (c) 2010-2020 three.js authors
*
* WebGL port of Subpixel Morphological Antialiasing (SMAA) v2.8
* Preset: SMAA 1x Medium (with color edge detection)
* https://github.com/iryoku/smaa/releases/tag/v2.8
*/
export const weights_frag = `
precision highp float;
precision highp int;
precision highp sampler2D;
#define SMAASampleLevelZeroOffset(tex, coord, offset) texture2D(tex, coord + float(offset) * uTexSizeInv, 0.0)
#define SMAA_AREATEX_MAX_DISTANCE 16
#define SMAA_AREATEX_PIXEL_SIZE (1.0 / vec2(160.0, 560.0))
#define SMAA_AREATEX_SUBTEX_SIZE (1.0 / 7.0)
uniform sampler2D tEdges;
uniform sampler2D tArea;
uniform sampler2D tSearch;
uniform vec2 uTexSizeInv;
varying vec2 vUv;
varying vec4 vOffset[3];
varying vec2 vPixCoord;
#if __VERSION__ == 100
vec2 round(vec2 x) {
return sign(x) * floor(abs(x) + 0.5);
}
#endif
float SMAASearchLength(sampler2D searchTex, vec2 e, float bias, float scale) {
// Not required if searchTex accesses are set to point:
// float2 SEARCH_TEX_PIXEL_SIZE = 1.0 / float2(66.0, 33.0);
// e = float2(bias, 0.0) + 0.5 * SEARCH_TEX_PIXEL_SIZE +
// e * float2(scale, 1.0) * float2(64.0, 32.0) * SEARCH_TEX_PIXEL_SIZE;
e.r = bias + e.r * scale;
return 255.0 * texture2D(searchTex, e, 0.0).r;
}
float SMAASearchXLeft(sampler2D edgesTex, sampler2D searchTex, vec2 texCoord, float end) {
/**
* @PSEUDO_GATHER4
* This texCoord has been offset by (-0.25, -0.125) in the vertex shader to
* sample between edge, thus fetching four edges in a row.
* Sampling with different offsets in each direction allows to disambiguate
* which edges are active from the four fetched ones.
*/
vec2 e = vec2(0.0, 1.0);
for (int i = 0; i < dMaxSearchSteps; i++) { // WebGL port note: Changed while to for
e = texture2D( edgesTex, texCoord, 0.0).rg;
texCoord -= vec2(2.0, 0.0) * uTexSizeInv;
if (!(texCoord.x > end && e.g > 0.8281 && e.r == 0.0)) break;
}
// We correct the previous (-0.25, -0.125) offset we applied:
texCoord.x += 0.25 * uTexSizeInv.x;
// The searches are bias by 1, so adjust the coords accordingly:
texCoord.x += uTexSizeInv.x;
// Disambiguate the length added by the last step:
texCoord.x += 2.0 * uTexSizeInv.x; // Undo last step
texCoord.x -= uTexSizeInv.x * SMAASearchLength(searchTex, e, 0.0, 0.5);
return texCoord.x;
}
float SMAASearchXRight(sampler2D edgesTex, sampler2D searchTex, vec2 texCoord, float end) {
vec2 e = vec2( 0.0, 1.0 );
for (int i = 0; i < dMaxSearchSteps; i++) { // WebGL port note: Changed while to for
e = texture2D(edgesTex, texCoord, 0.0).rg;
texCoord += vec2(2.0, 0.0) * uTexSizeInv;
if (!(texCoord.x < end && e.g > 0.8281 && e.r == 0.0)) break;
}
texCoord.x -= 0.25 * uTexSizeInv.x;
texCoord.x -= uTexSizeInv.x;
texCoord.x -= 2.0 * uTexSizeInv.x;
texCoord.x += uTexSizeInv.x * SMAASearchLength( searchTex, e, 0.5, 0.5 );
return texCoord.x;
}
float SMAASearchYUp(sampler2D edgesTex, sampler2D searchTex, vec2 texCoord, float end) {
vec2 e = vec2( 1.0, 0.0 );
for (int i = 0; i < dMaxSearchSteps; i++) { // WebGL port note: Changed while to for
e = texture2D(edgesTex, texCoord, 0.0).rg;
texCoord += vec2(0.0, 2.0) * uTexSizeInv; // WebGL port note: Changed sign
if (!(texCoord.y > end && e.r > 0.8281 && e.g == 0.0)) break;
}
texCoord.y -= 0.25 * uTexSizeInv.y; // WebGL port note: Changed sign
texCoord.y -= uTexSizeInv.y; // WebGL port note: Changed sign
texCoord.y -= 2.0 * uTexSizeInv.y; // WebGL port note: Changed sign
texCoord.y += uTexSizeInv.y * SMAASearchLength(searchTex, e.gr, 0.0, 0.5); // WebGL port note: Changed sign
return texCoord.y;
}
float SMAASearchYDown(sampler2D edgesTex, sampler2D searchTex, vec2 texCoord, float end) {
vec2 e = vec2( 1.0, 0.0 );
for (int i = 0; i < dMaxSearchSteps; i++) { // WebGL port note: Changed while to for
e = texture2D(edgesTex, texCoord, 0.0).rg;
texCoord -= vec2( 0.0, 2.0 ) * uTexSizeInv; // WebGL port note: Changed sign
if (!(texCoord.y < end && e.r > 0.8281 && e.g == 0.0)) break;
}
texCoord.y += 0.25 * uTexSizeInv.y; // WebGL port note: Changed sign
texCoord.y += uTexSizeInv.y; // WebGL port note: Changed sign
texCoord.y += 2.0 * uTexSizeInv.y; // WebGL port note: Changed sign
texCoord.y -= uTexSizeInv.y * SMAASearchLength(searchTex, e.gr, 0.5, 0.5); // WebGL port note: Changed sign
return texCoord.y;
}
vec2 SMAAArea(sampler2D areaTex, vec2 dist, float e1, float e2, float offset) {
// Rounding prevents precision errors of bilinear filtering:
vec2 texCoord = float(SMAA_AREATEX_MAX_DISTANCE) * round(4.0 * vec2(e1, e2)) + dist;
// We do a scale and bias for mapping to texel space:
texCoord = SMAA_AREATEX_PIXEL_SIZE * texCoord + (0.5 * SMAA_AREATEX_PIXEL_SIZE);
// Move to proper place, according to the subpixel offset:
texCoord.y += SMAA_AREATEX_SUBTEX_SIZE * offset;
return texture2D(areaTex, texCoord, 0.0).rg;
}
vec4 SMAABlendingWeightCalculationPS(vec2 texCoord, vec2 pixCoord, vec4 offset[3], sampler2D edgesTex, sampler2D areaTex, sampler2D searchTex, ivec4 subsampleIndices) {
vec4 weights = vec4(0.0, 0.0, 0.0, 0.0);
vec2 e = texture2D(edgesTex, texCoord).rg;
if (e.g > 0.0) { // Edge at north
vec2 d;
// Find the distance to the left:
vec2 coords;
coords.x = SMAASearchXLeft(edgesTex, searchTex, offset[0].xy, offset[2].x );
coords.y = offset[1].y; // offset[1].y = texCoord.y - 0.25 * uTexSizeInv.y (@CROSSING_OFFSET)
d.x = coords.x;
// Now fetch the left crossing edges, two at a time using bilinear
// filtering. Sampling at -0.25 (see @CROSSING_OFFSET) enables to
// discern what value each edge has:
float e1 = texture2D(edgesTex, coords, 0.0).r;
// Find the distance to the right:
coords.x = SMAASearchXRight(edgesTex, searchTex, offset[0].zw, offset[2].y);
d.y = coords.x;
// We want the distances to be in pixel units (doing this here allow to
// better interleave arithmetic and memory accesses):
d = d / uTexSizeInv.x - pixCoord.x;
// SMAAArea below needs a sqrt, as the areas texture is compressed
// quadratically:
vec2 sqrt_d = sqrt(abs(d));
// Fetch the right crossing edges:
coords.y -= 1.0 * uTexSizeInv.y; // WebGL port note: Added
float e2 = SMAASampleLevelZeroOffset(edgesTex, coords, ivec2(1, 0)).r;
// Ok, we know how this pattern looks like, now it is time for getting
// the actual area:
weights.rg = SMAAArea(areaTex, sqrt_d, e1, e2, float(subsampleIndices.y));
}
if (e.r > 0.0) { // Edge at west
vec2 d;
// Find the distance to the top:
vec2 coords;
coords.y = SMAASearchYUp(edgesTex, searchTex, offset[1].xy, offset[2].z );
coords.x = offset[0].x; // offset[1].x = texCoord.x - 0.25 * uTexSizeInv.x;
d.x = coords.y;
// Fetch the top crossing edges:
float e1 = texture2D(edgesTex, coords, 0.0).g;
// Find the distance to the bottom:
coords.y = SMAASearchYDown(edgesTex, searchTex, offset[1].zw, offset[2].w);
d.y = coords.y;
// We want the distances to be in pixel units:
d = d / uTexSizeInv.y - pixCoord.y;
// SMAAArea below needs a sqrt, as the areas texture is compressed
// quadratically:
vec2 sqrt_d = sqrt(abs(d));
// Fetch the bottom crossing edges:
coords.y -= 1.0 * uTexSizeInv.y; // WebGL port note: Added
float e2 = SMAASampleLevelZeroOffset(edgesTex, coords, ivec2(0, 1)).g;
// Get the area for this direction:
weights.ba = SMAAArea(areaTex, sqrt_d, e1, e2, float(subsampleIndices.x));
}
return weights;
}
void main() {
gl_FragColor = SMAABlendingWeightCalculationPS(vUv, vPixCoord, vOffset, tEdges, tArea, tSearch, ivec4(0.0));
}
`;
\ No newline at end of file
src/mol-gl/shader/smaa/weights.vert.ts 0 → 100644
+ 39
− 0
View file @ 9eba0b91
/**
* Slightly adapted from https://github.com/mrdoob/three.js
* MIT License Copyright (c) 2010-2020 three.js authors
*
* WebGL port of Subpixel Morphological Antialiasing (SMAA) v2.8
* Preset: SMAA 1x Medium (with color edge detection)
* https://github.com/iryoku/smaa/releases/tag/v2.8
*/
export const weights_vert = `
precision highp float;
attribute vec2 aPosition;
uniform vec2 uQuadScale;
uniform vec2 uTexSizeInv;
varying vec2 vUv;
varying vec4 vOffset[3];
varying vec2 vPixCoord;
void SMAABlendingWeightCalculationVS(vec2 texCoord) {
vPixCoord = texCoord / uTexSizeInv;
// We will use these offsets for the searches later on (see @PSEUDO_GATHER4):
vOffset[0] = texCoord.xyxy + uTexSizeInv.xyxy * vec4(-0.25, 0.125, 1.25, 0.125); // WebGL port note: Changed sign in Y and W components
vOffset[1] = texCoord.xyxy + uTexSizeInv.xyxy * vec4(-0.125, 0.25, -0.125, -1.25); // WebGL port note: Changed sign in Y and W components
// And these for the searches, they indicate the ends of the loops:
vOffset[2] = vec4(vOffset[0].xz, vOffset[1].yw) + vec4(-2.0, 2.0, -2.0, 2.0) * uTexSizeInv.xxyy * float(dMaxSearchSteps);
}
void main() {
vUv = (aPosition + 1.0) * 0.5;
SMAABlendingWeightCalculationVS(vUv);
vec2 position = aPosition * uQuadScale - vec2(1.0, 1.0) + uQuadScale;
gl_Position = vec4(position, 0.0, 1.0);
}
`;
\ No newline at end of file
src/mol-gl/webgl/texture.ts
+ 26
− 4
View file @ 9eba0b91
......@@ -118,6 +118,10 @@ export function getAttachment(gl: GLRenderingContext, extensions: WebGLExtension
throw new Error('unknown texture attachment');
}
function isImage(x: TextureImage<any> | TextureVolume<any> | HTMLImageElement): x is HTMLImageElement {
return typeof HTMLImageElement !== undefined && (x instanceof HTMLImageElement);
}
function isTexture2d(x: TextureImage<any> | TextureVolume<any>, target: number, gl: GLRenderingContext): x is TextureImage<any> {
return target === gl.TEXTURE_2D;
}
......@@ -138,7 +142,7 @@ export interface Texture {
getDepth: () => number
define: (width: number, height: number, depth?: number) => void
load: (image: TextureImage<any> | TextureVolume<any>) => void
load: (image: TextureImage<any> | TextureVolume<any> | HTMLImageElement) => void
bind: (id: TextureId) => void
unbind: (id: TextureId) => void
/** Use `layer` to attach a z-slice of a 3D texture */
......@@ -193,7 +197,7 @@ export function createTexture(gl: GLRenderingContext, extensions: WebGLExtension
init();
let width = 0, height = 0, depth = 0;
let loadedData: undefined | TextureImage<any> | TextureVolume<any>;
let loadedData: undefined | TextureImage<any> | TextureVolume<any> | HTMLImageElement;
let destroyed = false;
function define(_width: number, _height: number, _depth?: number) {
......@@ -210,13 +214,17 @@ export function createTexture(gl: GLRenderingContext, extensions: WebGLExtension
}
}
function load(data: TextureImage<any> | TextureVolume<any>) {
function load(data: TextureImage<any> | TextureVolume<any> | HTMLImageElement) {
gl.bindTexture(target, texture);
// unpack alignment of 1 since we use textures only for data
gl.pixelStorei(gl.UNPACK_ALIGNMENT, 1);
gl.pixelStorei(gl.UNPACK_COLORSPACE_CONVERSION_WEBGL, gl.NONE);
gl.pixelStorei(gl.UNPACK_PREMULTIPLY_ALPHA_WEBGL, 0);
if (isTexture2d(data, target, gl)) {
if (isImage(data)) {
gl.pixelStorei(gl.UNPACK_FLIP_Y_WEBGL, false);
gl.bindTexture(gl.TEXTURE_2D, texture);
gl.texImage2D(gl.TEXTURE_2D, 0, internalFormat, format, type, data);
} else if (isTexture2d(data, target, gl)) {
gl.pixelStorei(gl.UNPACK_FLIP_Y_WEBGL, !!data.flipY);
width = data.width, height = data.height;
gl.texImage2D(target, 0, internalFormat, width, height, 0, format, type, data.array);
......@@ -314,6 +322,20 @@ export function createTextures(ctx: WebGLContext, schema: RenderableSchema, valu
return textures;
}
/**
* Loads an image from a url to a textures and triggers update asynchronously.
* This will not work on node.js with a polyfill for HTMLImageElement.
*/
export function loadImageTexture(src: string, cell: ValueCell<Texture>, texture: Texture) {
const img = new Image();
img.onload = function() {
texture.load(img);
ValueCell.update(cell, texture);
};
img.src = src;
document.body.appendChild(img);
}
//
export function createNullTexture(gl: GLRenderingContext, kind: TextureKind): Texture {
......
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Please register or to comment