Really Cheap Targa Motion Blur is a 32-bit Windows utility designed to take a numbered series of input Targa files, average groups of them together, and then output the resulting data as a new image, in order to simulate motion blur.

Read How to use Really Cheap Targa Motion Blur before beginning, or if what you're getting isn't what you expect!

For information on the various buttons and input fields, click on their representations below:

This is the base path and name of the numbered file sequence to be used as input images. For example, if you have files c:\frames\foo001.tga, c:\frames\foo002.tga, c:\frames\foo003.tga..., you would enter c:\frames\foo in this field.

If you wish to locate your input files through the use of a standard open-file dialog box, click this button. Select any one file of the numbered file sequence which you wish to use; the input file stem will be derived from this file (the numeric sequence and extension will be stripped off). If the input and output files were previously located in the same directory as each other, the output file stem will also be altered to point to the new location.

This button will automatically alter the Starting Frame and Ending Frame of the input files to reflect the first and last images it finds in the given sequence.

Enter the number of the first frame in the sequence which is to be blurred. Some 3D packages which generate numerically sequential frames default to beginning with frame 0, and others default to beginning with frame 1. You may also start at any arbitrary (positive) integer.

Enter the number of the last frame in the sequence which is to be blurred. This should be set to a number greater than the starting frame number, if you want anything useful to be done.

This field indicates the number of digits used to differentiate each frame from the next. If the images are numbered 000, 001, 002..., enter 3 in this field; if they're numbered 0001, 0002, 0003..., enter 4, and so on. If the numbers are not zero-padded, enter 0 (zero) in this field.

This is the base path and name of the numbered file sequence to be used as output images. For example, if you wish to save to the files c:\frames\bar001.tga, c:\frames\bar002.tga, c:\frames\bar003.tga..., you would enter c:\frames\bar in this field.

It is not recommended to use the same filename stem as the input files!

Enter the number of the first frame in the sequence which is to be generated. This can be any positive integer (or zero); the number of frames generated will be the number of input frames divided by the Frame step value.

Enter the number of input files which should be averaged to produce each output image. This value will vary depending on a number of factors, but it will primarily be determined by the largest motion in the field of view. It is suggested that this value be set high enough to prevent visible "banding" of the blurred objects.

Note that the number of output files will be the number of input files divided by this number, so if you use a frame-step value of 5, you should render five times the usual number of frames (ie, if you normally render at 30fps, the section to be blurred should be rendered at 150fps).

The Even Weighting and Temporal Bias radio buttons permit you to select the kind of averaging which is performed.

The Even Weighting method treats each input frame equally, and produces a blur effect which is more suitable for animations than the Temporal Bias method.

The following image was produced using the Even Weighting method:

The Even Weighting and Temporal Bias radio buttons permit you to select the kind of averaging which is performed.

The Temporal Bias method gives greater weight to each successive input frame, meaning that the earliest image is the faintest, and the latest image is the strongest. This produces a blur effect which is more suitable for still images than the Even Weighting method, as it provides contextual clues to the direction of motion.

The following image was produced using the Temporal Bias method:

When this box is checked, the output files will contain "blurred" alpha channels as well. If the input files do not contain alpha channels, the output alpha channel will be entirely opaque. Be aware that some applications cannot read alpha-channel Targa files; for example, Animation Master can generate them, but cannot use them for rotoscopes.

When this box is checked, the output files will be written as compressed Targa files to take up less space (compression ratios are dependent upon the complexity of the image). Be aware that some applications cannot read compressed Targa files.

When all of the parameters are set appropriately, click the Begin! button to start processing the input files. During processing, the button will change to a Cancel! button, in case you wish to abort the procedure before it has completed.

Press the Quit! button to exit the program.

The status indicator line informs you of what's going on at the moment. During processing, this displays the number of the current input file being processed.

In real-world cinematography, motion blur is produced when an object is moving too quickly for a camera to get a crystal-clear image. In computer animation, the lack of motion blur can make fast-moving objects appear too crisp.

One method of overcoming this is to render multiple intermediate images between each actual frame, and then average the intermediate frames into the actual ones; the more intermediate frames are rendered, the smoother the blur appears.

The "Frame step" field indicates how many images are to be combined into a single frame; this number also indicates how many more frames you need to render. With a frame step of 5, you need to render five times as many frames (so if you normally render at 30fps, you should render at 150 fps in order to keep the same timing).

Larger frame-step values will produce a smoother blur (and, unfortunately, require rendering more intermediate images).

The Temporal Bias method produces a blur which is more suited to still images; the blur fades over time, which provides clues to the viewer regarding the direction of movement. For animations, the Even Weighting algorithm is more suitable.