Transverse or lateral chromatic aberration is an optical artefact caused by different wavelengths of light focussing at different positions on a camera's focal plane. It appears as blue and purple fringing which increases towards the edge of an image. In addition to the Wikipedia entry on chromatic aberration, there's a great article here at Photography Life which discusses the phenomenon in great detail.

Although the bane of many photographers, I've created a Core Image filter to simulate the effect which can be used to add a low-fi, grungy look to images. The basic mechanics of the filter are pretty simple: it essentially consists of three zoom filters each with a slightly different offsets for red, green and blue. The technique borrows from Session 515: Developing Core Image Filters for iOS talk at WWDC 2014. 

Let's look at the general kernel I've written for the filter and then step through it line by line:


    let transverseChromaticAberrationKernel = CIKernel(string:
        "kernel vec4 chromaticAberrationFunction(sampler image, vec2 size, float sampleCount, float start, float blur) {" +
        "  int sampleCountInt = int(floor(sampleCount));" + // 1
        "  vec4 accumulator = vec4(0.0);"// 2
        "  vec2 dc = destCoord(); "// 3
        "  float normalisedValue = length(((dc / size) - 0.5) * 2.0);" + // 4
        "  float strength = clamp((normalisedValue - start) * (1.0 / (1.0 - start)), 0.0, 1.0); " + // 5
        
        "  vec2 vector = normalize((dc - (size / 2.0)) / size);" + // 6
        "  vec2 velocity = vector * strength * blur; " + // 7
        
        "  vec2 redOffset = -vector * strength * (blur * 1.0); " + // 8
        "  vec2 greenOffset = -vector * strength * (blur * 1.5); " + // 8
        "  vec2 blueOffset = -vector * strength * (blur * 2.0); " + // 8
        
        "  for (int i=0; i < sampleCountInt; i++) { " + // 9
        "      accumulator.r += sample(image, samplerTransform (image, dc + redOffset)).r; " + // 10
        "      redOffset -= velocity / sampleCount; " + // 11
        
        "      accumulator.g += sample(image, samplerTransform (image, dc + greenOffset)).g; " +
        "      greenOffset -= velocity / sampleCount; " +
        
        "      accumulator.b += sample(image, samplerTransform (image, dc + blueOffset)).b; " +
        "      blueOffset -= velocity / sampleCount; " +
        "  } " +
        "  return vec4(vec3(accumulator / sampleCount), 1.0); " + // 12

        "}")


  1. Because Core Image Kernel Language only allows float scalar arguments and sampleCount needs to be an integer to construct a loop, I create an int version of it. 
  2. As the filter loops over pixels, it will accumulate their red, green and blue values into this three component vector. 
  3. destCoord() returns the position of the pixel currently being computed in the coordinate space of the image being rendered. 
  4. Although the filter can calculate the size of the image with samplerSize(), passing the size as an argument reduces the amount of processing the kernel needs to do. This line converts the co-ordinates to the range -1 to +1 for both axes. 
  5. strength is a normalised value that starts at zero at the beginning of the effect and reaches one at the edge of the image. 
  6. vector is the direction of the effect which radiates from the centre of the image. normalize keeps the sum of the vector to one. 
  7. Multiplying the direction by the strength by the maximum blur argument gives a velocity vector for how much blur and in what direction the filter applies to the current pixel.
  8. Transverse chromatic aberration increases in strength proportionally to the wavelength of light. The filter simulates this by offsetting the effect the least for red and the most for blue.
  9. The filter iterates once for each sampleCount.
  10. For each color, the filter takes accumulates a sample offset along the direction of the vector - effectively summing the pixels along a radial line.
  11. The offset for each color is decremented.
  12. The accumulated colors are averages and returned with an alpha value of 1.0.
The number of samples controls the quality of the effect and the performance of the filter. With a large maximum blur of 20 but only 3 samples, the effects looks like:


But with the same blur amount and 40 samples, the effect is a lot smoother:


The falloff parameter controls where the effect begins - a value of 0.75 means the effect begins three quarters of the distance from the centre of the image to the edge:


Core Image for Swift

There's a full CIFilter implementation of this filter under the Filterpedia repository. However, if you'd like to learn more about writing custom Core Image kernels with Core Image Kernel Language, may I recommend my book, Core Image for Swift.

Core Image for Swift is available from both Apple's iBooks Store or, as a PDF, from Gumroad. IMHO, the iBooks version is better, especially as it contains video assets which the PDF version doesn't.




1

View comments

  1. How would you approach removing chromatic aberration?

    ReplyDelete


It's been a fairly busy few months at my "proper" job, so my recreational Houdini tinkering has taken a bit of a back seat. However, when I saw my Swarm Chemistry hero, Hiroki Sayama tweeting a link to How a life-like system emerges from a simple particle motion law, I thought I'd dust off Houdini to see if I could implement this model in VEX.

The paper discusses a simple particle system, named Primordial Particle Systems (PPS), that leads to life-like structures through morphogenesis. Each particle in the system is defined by its position and heading and, with each step in the simulation, alters its heading based on the PPS rule and moves forward at a defined speed. The heading is updated based on the number of neighbors to the particle's left and right. 

The project set up is super simple: 



Inside a geometry node, I create a grid, and randomly scatter 19,000 points across it. An attribute wrangle node assigns a random value to @angle:
@angle = $PI * 2 * rand(@ptnum); 
The real magic happens inside another attribute wrangle inside the solver.

In a nutshell, my VEX code iterates over each point's neighbors and sums the neighbor count to its left and right. To figure out the chirality, I use some simple trigonometry to rotate the vector defined by the current particle and the neighbor by the current particle's angle, then calculate the angle of the rotated vector. 
while(pciterate(pointCloud)) {

    vector otherPosition;
    pcimport(pointCloud, "P", otherPosition);

    vector2 offsetPosition = set(otherPosition.x - @P.x, otherPosition.z - @P.z);
    float xx = offsetPosition.x * cos(-@angle) - offsetPosition.y * sin(-@angle);
    float yy = offsetPosition.x * sin(-@angle) + offsetPosition.y * cos(-@angle);
    
    float otherAngle = atan2(yy, xx); 

    if (otherAngle >= 0) {
        L++;
    } 
    else {
        R++;
    }   
}
After iterating over the nearby particles, I update the angle based on the PPS rule:
float N = float(L + R);
@angle += alpha + beta * N * sign(R - L);
...and, finally, I can update the particle's position based on its angle and speed:
vector velocity = set(cos(@angle) * @speed, 0.0, sin(@angle) * @speed);  
@P += velocity ;
Not quite finally, because to make things pretty, I update the color using the number of neighbors to control hue:
@Cd = hsvtorgb(N / maxParticles, 1.0, 1.0); 
Easy!

Solitons Emerging from Tweaked Model



I couldn't help tinkering with the published PPS math by making the speed a function of the number of local neighbors:
@speed = 1.5 * (N / maxParticles);
In the video above, alpha is 182° and beta is -13°.

References

Schmickl, T. et al. How a life-like system emerges from a simple particle motion law. Sci. Rep. 6, 37969; doi: 10.1038/srep37969 (2016).


5

View comments

  1. ok. I've got to finish current job, then crash course in programming, and ... this is very inspirational!

    ReplyDelete
  2. This comment has been removed by the author.

    ReplyDelete
  3. This comment has been removed by the author.

    ReplyDelete
  4. This comment has been removed by the author.

    ReplyDelete
  5. This comment has been removed by the author.

    ReplyDelete
About Me
About Me
Labels
Labels
Blog Archive
Loading