How to get depth from AVO (Arbitrary Output Variables)?

You can grab a lot of data using AVOs, such as normals, albedo etc:

https://docs.unity3d.com/Packages/com.unity.render-pipelines.high-definition@10.0/manual/AOVs.html

But how to grab depth?

 

Btw, do you know if it's possible to grab this info and apply in real-time? Like every frame.

 

Yes. You can use post processing for that.

 

Noone knows? Where is the documentation, seems we are just left to trial and error with HDRP.......?

 

There's a depth capture example: https://github.com/alelievr/HDRP-Custom-Passes#depth-capture
Maybe it can help you.

 

Can you elaborate on that, please? I'm not much of a programmer, and I want to output directional specular every frame
to subtract it from some materials. But when I tried the depth capture script I've mentioned above, it turned out that it doesn't work in real-time. It only updates when a camera is moved in the editor. So I'm looking for some minimal example of how I can output the directional specular in real-time every frame into a rendertexture.

 

I'm referring to the specular contribution of all real-time lights:

I want this contribution to be rendered into a rendertexture every frame, so I could use this texture inside shaders/custom passes. Is this possible/practical?

I've tested your script and it works perfect and does exactly what I want, but it has a serious performance impact, as you said.
Thanks!

 

Thank you, this is useful. However, when setting Material Output to None, Debug Output to None and AOV Type To Depth Stencil, my rendertexture remains black..? I tried different kinds of RenderTexture formats... The other render options works as expected (albedo etc)

On another note, I found a separate way to do this by making a custom post processing shader. There I can get the depth, but now Albedo is missing ... Lol. Or at least I do not know how to fetch it in this way, maybe it is easy?

Could you advice please?

Code (CSharp):

1.         float4 CustomPostProcess(Varyings input) : SV_Target
2.         {
3.             UNITY_SETUP_STEREO_EYE_INDEX_POST_VERTEX(input);
4.    
5.             float3 sourceColor = SAMPLE_TEXTURE2D_X(_MainTex, s_linear_clamp_sampler, input.texcoord).xyz;
6.    
7.             // Apply greyscale effect
8.             float3 color = lerp(sourceColor, Luminance(sourceColor), _Intensity)*0.5;
9.    
10.             float3 albedoColor = //HOW TO????
11.    
12.     ...
13.        
14.             return float4(color, 1);
15.     }
16.  

 

Ah, yes, this is exactly what I started to do. But I could not figure out how to get Albedo this way, which is important for my denoiser. How can I grab this in the shader?

I used this code to grab depth and normals:

Code (CSharp):

1. #include "Packages/com.unity.render-pipelines.core/ShaderLibrary/Common.hlsl"
2. #include "Packages/com.unity.render-pipelines.core/ShaderLibrary/Color.hlsl"
3. #include "Packages/com.unity.render-pipelines.high-definition/Runtime/ShaderLibrary/ShaderVariables.hlsl"
4. #include "Packages/com.unity.render-pipelines.high-definition/Runtime/PostProcessing/Shaders/FXAA.hlsl"
5. #include "Packages/com.unity.render-pipelines.high-definition/Runtime/PostProcessing/Shaders/RTUpscale.hlsl"
6. #include "Packages/com.unity.render-pipelines.high-definition/Runtime/Material/NormalBuffer.hlsl"
7.  
8. ...
9.  
10.     float3 GetNormalWorldSpace(float2 uv, float depth)
11.     {
12.         float3 normalWS = 0.0f;
13.         if (depth > 0.0f)
14.         {
15.             NormalData normalData;
16.             const float4 normalBuffer = LOAD_TEXTURE2D_X(_NormalBufferTexture, uv);
17.             DecodeFromNormalBuffer(normalBuffer, uv, normalData);
18.             normalWS = normalData.normalWS;
19.         }
20.         return normalWS;
21.     }
22.  
23.         float3 sourceColor = SAMPLE_TEXTURE2D_X(_MainTex, s_linear_clamp_sampler, input.texcoord).xyz;
24.         float depth = SampleCameraDepth(input.texcoord);
25.         float3 normalWS = GetNormalWorldSpace(input.texcoord*_ScreenSize.xy,depth);
26.         float3 albedo= //?????
27.  

 

Thank you for your suggestion. I tried out the shader example in that link:

Code (CSharp):

1.    
2. float4 FullScreenPass(Varyings varyings) : SV_Target
3.     {
4.         UNITY_SETUP_STEREO_EYE_INDEX_POST_VERTEX(varyings);
5.         float depth = LoadCameraDepth(varyings.positionCS.xy);
6.         PositionInputs posInput = GetPositionInput(varyings.positionCS.xy, _ScreenSize.zw, depth, UNITY_MATRIX_I_VP, UNITY_MATRIX_V);
7.         float3 viewDirection = GetWorldSpaceNormalizeViewDir(posInput.positionWS);
8.         float4 color = float4(0.0, 0.0, 0.0, 0.0);
9.  
10.         // Load the camera color buffer at the mip 0 if we're not at the before rendering injection point
11.         if (_CustomPassInjectionPoint != CUSTOMPASSINJECTIONPOINT_BEFORE_RENDERING)
12.             color = float4(CustomPassLoadCameraColor(varyings.positionCS.xy, 0), 1);
13.  
14.         // Add your custom pass code here
15.         return float4(color.rgb , 1);    
16. }

But this only returns black. How to render albedo here?

 

I’d like to use AOV to render out specific passes (alpha, depth, etc.) and use render image sequence where Unity renders out each frame at run-time.

 

Bump

 

Hi PavlosM. I was trying the attached script and it works great for extracting an AOV channel. We want to extract 2 channels, ALBEDO and Normals. I tried adding two instances of the component, but the second channel renders black. Any suggestions on how to extract two channels to separate RenderTextures?

Thanks in advance,

Alejandro.

 

Hi @PavlosM please advise.

 

When are you guys going to make the non beauty passes have anti-aliasing. The Alpha pass in particular is pretty useless because it's so jagged at the edges.