Fxaa natural vision code#

#define pattern 2 // Choose a sample pattern ( 1, 2, 3 or 4 ) - I suggest 2 #define sharp_clamp 0.050 // Limits maximum amount of sharpening a pixel recieves - Default is 0.050 #define sharp_strength 1.0 // Strength of the sharpening - You should probably use something between 0.2 and 2.0 - default is 1.0 This is similar to using Unsharp Mask in Photoshop. It does this in luma to avoid color artifacts and allows limiting the maximum sharpning to avoid or lessen halo artifacts. It blurs the original pixel with the surrounding pixels and then subtracts this blur to sharpen the image.

Fxaa natural vision code#

Here is the code I use for LumaSharpen in MPC I just use it as a post-resize shader though. I suggest you add it to the list of Post-resize shaders though, but it also works as pre-resize shader. You can do the same with LumaSharpen too. Press CTRL+9 to open the Edit Shaders dialog again. The settings can be edited in the top of the Vibrance shader code. Also remember to enable Pre-resize shaders. If it won't compile right because you have a very old GPU then choose ps_2_0.Д§) Go to Play -> Shaders -> Select Shaders, and add Vibrance to the list of Pre-resize shaders. return color_saturation.xxxx //Visualize the saturationДЇloat4 main(float2 tex : TEXCOORD0) : COLOR 6) Choose ps_3_0 as the pixel shader version to compile with. color = lerp(luma, color, (1.0 + (Vibrance * (1.0 - color_saturation)))) //extrapolate between luma and original by 1 + (1-saturation)Д¬olor = lerp(luma, color, (1.0 + (Vibrance * (1.0 - (sign(Vibrance) * color_saturation))))) //extrapolate between luma and original by 1 + (1-saturation) This avoids oversaturation of pixels that were already very saturated.ДЇloat4 color = colorInput //original input colorДЇloat3 lumCoeff = float3(0.2126, 0.7152, 0.0722) //Values to calculate luma withДЇloat luma = dot(lumCoeff, color.rgb) //calculate luma (grey)ДЇloat max_color = max(max(colorInput.r,colorInput.g),colorInput.b) //Find the strongest colorДЇloat min_color = min(min(colorInput.r,colorInput.g),colorInput.b) //Find the weakest colorДЇloat color_saturation = max_color - min_color //The difference between the two is the saturation So pixels that had little color get a larger boost than pixels that had a lot. Vibrance intelligently boosts the saturation of pixels #define Vibrance 0.3 //Intelligently saturates (or desaturates if you use negative values) the pixels depending on their original saturation.Д«y Christian Cann Schuldt Jensen ~ CeeJay.dk To sum up - I use this code to support GPU Shaderanalyzer - it has no other function. Therefore I check with #ifndef ( if not defined ) if BUFFER_RCP_WIDTH is not defined and if it isn't then I define all the values match a common screensize like 1680x1050 so that GPU Shaderanalyzer can compile and analyze the shaders. The shaders depend on these values being defined, however when I do debugging and profiling with AMDs GPU Shaderanalyzer those values are not defined because the injector did not run and so the shaders cannot compile. When the injector runs it defines some values depending on your screen resolution.Д«UFFER_WIDTH = The width of the screen resolution you are rendering atД«UFFER_HEIGHT = The height of the screen resolution you are rendering atД«UFFER_RCP_WIDTH = (1.0 / screen width) - This is the width of a single pixel in texture coordinatesД«UFFER_RCP_HEIGHT = (1.0 / screen height) - This is the height of a single pixel in texture coordinates I have only meant for normal users to edit SweetFX_preset.txt, SweetFX_settings.txt and injector.iniД«ut if you're curious to what that code does I'll explain: No, as a user you don't have to edit any files in the SweetFX/Shaders/ subdirectory.