Giter Club home page Giter Club logo

fidelityfx-cacao's Introduction

FidelityFX CACAO

Copyright (c) 2020-2021 Advanced Micro Devices, Inc. All rights reserved.

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

Combined Adaptive Compute Ambient Occlusion (CACAO)

Combined Adaptive Compute Ambient Occlusion (CACAO) is a highly optimized adaptive sampling ambient occlusion implementation. This directory contains the source code for CACAO as well as a sample demonstrating usage and integration of the library. The directory structure is as follows:

You can find the binaries for FidelityFX CACAO in the release section on GitHub.

Sponza Model Issue

In the provided Sponza model for the FidelityFX CACAO sample, at the bottom of some curtains, there is an issue with incorrect normals in the mesh causing ambient occlusion to be incorrectly calculated as light in places where occlusion should be dark. This is a known issue with the mesh, and not an issue with FidelityFX CACAO itself.

Notices

FidelityFX CACAO is a modification of the Adaptive Screen Space Ambient Occlusion (ASSAO) algorithm that was developed by Intel. The original implementation can be found here.

fidelityfx-cacao's People

Contributors

jayfraser avatar rys avatar

Stargazers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

Watchers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

fidelityfx-cacao's Issues

Native Resolution

In the preivous version there was native resolution AO, why has it been removed?

Reduction in Quality between V1.0 and V1.1

First of all, thank you for your hard work and for making this open-source! It is genuinely appreciated by those of us looking to learn from examples, and looking to benefit off your hard work in our engines. :)

Background

I am a contributor to Godot which is an open-source game engine. We have been evaluating new methods for SSAO for a few months now. We were planning on using ASSAO when we found out about CACAO. However, once trying CACAO v1.1 we noticed the quality did not match the promotional images at all. However, we noticed CACAO v1.0 looks great!

The Issues

  • V1.1 substantially reduces the quality of CACAO compared to v1.0.

  • V1.1 restricts itself to only using the downsampled version edit I now see the note in ffx_cacao.h

    // In future it is planned that FidelityFX CACAO will allow SSAO generation at native resolution.
    // However, at the current time the performance/image quality trade off is poor, and further optimisation
    // work is being carried out. If you wish to experiment with native resolution SSAO generation enable this
    // flag. Integrating FidelityFX CACAO into games with native resolution enabled is currently not recommended.
    // #define FFX_CACAO_ENABLE_NATIVE_RESOLUTION

  • V1.1 is blurred much more heavily than the downsampled version in V1.0

  • In V1.1 "adaptive quality" doesn't actually impact the image quality, even when set to "highest quality"

I have included screenshots below to illustrate the issue.

Could these issues be addressed in a subsequent point release or a bug fixing commit? I ask because I am unfamiliar with your development approach/priorities. For us working on Godot, we would love to see these issues addressed and then we could port CACAO directly into Godot. But if not, we totally understand. Thanks anyway!

Version 1.0:
Screenshot (10)
Screenshot (16)

Version 1.1:
Screenshot (11)
Screenshot (15)

Version 1.0 downsampled
Screenshot (17)
Screenshot (18)

Heavy Glitches on R9 390

grafik

Im using the latest AMD driver 20.4.2 and builded from source. Same happends with the release build.

Bug: Disparity in quality levels

Hello,

I found some disparities in quality levels in the code. Could you comment on It whether I am missing something, or It is indeed a bug?

In fx_cacao.cpp on line 1974 the importance map is generated only if the level is Highest (4) which seems correct.

if (context->settings.qualityLevel == FFX_CACAO_QUALITY_HIGHEST) {
  // generate importance map code
}

BUT, in fx_cacaco.hlsl for level HIGH (3) you call GenerateSSAOShadowsInternal with qualityLevel = 3 and adaptiveBase = false which results in:

static const uint g_numTaps[5] = { 3, 5, 12, 0, 0 };
// ... later
const int numberOfTaps = (adaptiveBase) ? (SSAO_ADAPTIVE_TAP_BASE_COUNT) : (g_numTaps[qualityLevel]);

^ number of taps being 0! which I believe should be 12 as in https://software.intel.com/content/www/us/en/develop/articles/adaptive-screen-space-ambient-occlusion.html.

and

if ((qualityLevel != 3) || adaptiveBase)
{

}
else // if ( qualityLevel == 3 ) adaptive approach
{
 // ...
 float importance = g_ImportanceMap.SampleLevel(g_LinearClampSampler, fullResUV, 0.0).x;
 // ...
}

^ you sample from a map that is not generated. Because level 3 (High) is not the adaptive one. That is 4. I believe the if should be qualityLevel !=4.

No data to support your claims?

This claim is made:

(CACAO) is a highly optimized adaptive sampling ambient occlusion implementation.

But there's no data available how much better it does vs Intel's ASSAO or Nvidia's HBAO+ or VXAO.

Will such data be made available on a later date?

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.