Box Fitting for After Effects

Workflow

It’s important to understand how Box Fitting works in order to get the most out of it. The number of iterations specifies how many attempts Box Fitting will try to render boxes before giving up. Each iteration forces Box Fitting to randomly pick a pixel location and, if there is no box already there, it creates a new one. When it creates a new box, it searches outwards from the pixel location in all directions until it either reaches the maximum box size specified, or it reaches the edge of another box. At this point, Box Fitting will draw the box.

As such, it’s important to note that iterations does not mean number of boxes. It may run 10,000 iterations but it will inevitably find pixel locations which already contain a previously-rendered box, meaning that iteration will be terminated and the next one initiated. Typically, the larger the image resolution, the more iterations you’ll want. Otherwise, there will be few boxes rendered.

Box Fitting will automatically render the background transparent, which allows you to drop in any background layer you want underneath it. It also allows you to perform unique layer styles and effects such as adding layer strokes, shadows, etc.

Remember to play with the frame rate to get different results. Box Fitting relies on your composition’s framerate for it’s temporal settings so specifying a frame rate of say 12 frames per second will yield a less frenetic result.

All of the images rendered in this page have been rendered at 1:1 ratio. This means they were not scaled up or down. Why is this important? Because filtered scaling (Bilinear, Bicubic, etc) introduces antialiasing and softens the boxes. If you want a clean result, don’t scale up or down – instead create your composition size at the same size you intend to export.

Parameters

Iterations: Defines the maximum number of iterations to attempt. There is a maximum limit of 100,000,000 iterations. The more iterations there are, the longer the computation. Make sure to check the preview window at 100% for an accurate representation of boxes.

Maximum Size: Specifies the maximum box size. Any rendered box may be smaller than or equal to this size. The smaller the box size, the more visible fidelity, but more boxes will be required.

Size Mapping: Specifies how to map box sizing according to image pixels. Constant means the box sizing will be determined only by the box size parameter. By Intensity means the box size will determined by the intensity of the image RGB pixels. By Alpha means the box size will determined by the transparency of the image RGB pixels. When these features are enabled, the darker or more transparent a pixel is, the smaller its box size. In the image example below on the right, the boxes become smaller in darker areas of the source image.

In the image example below on the right, the boxes become smaller in the more transparent areas of the source image (as dictated by the color black in this example).

Seed: Specifies the seed value for the random number generator. Having the ability to change the seed is important in case you wish to layer multiple instances of Box Fitting together.

Changelog

• v1.0 – First release.