Anaglyph image

Anaglyph images are produced using color filters or computer image processing techniques to combine images from two slightly different viewpoints into a single image. These images may then be viewed with Anaglyph glasses, which use color filters to moderate the light reaching each eye to create the illusion of a three diminutional image. This is an example of stereoscopy. Anaglyph images have seen a recent resurgence due to the presentation of images on the internet, coupled with the availabilty of low cost paper frames that hold accurate color filters. Practical images, where depth perception is useful, include the presentation of complex multi-dimensional data sets and stereographic images from (for example) the surface of mars, but for the most part, the materials are presented for entertainment. Anaglyph images are much easier to view than either parallel sighting or crossed eye stereograms, although the later types offer bright and accurate color rendering, not possible with anaglyphs. Anaglyphs could be useful as introductory materials for the merchandising of stereo image cards. With the techniques outlined below it is possible to convert stereo pairs from any source into anaglyph images.

Producing anaglyph images

In historical methods using camera filters, two images from the perspective of the left and right eyes are projected or printed together as a single image, one side through a red filter and the other side through a contrasting color such as blue or green or cyan. As outlined below, one may also use an image processing computer program to simulate the effect of using color filters, using as a source image a pair either color or monochrome images.

As the eye is less sensitive to red than to the other primary colors the viewing of an image is somewhat improved if the non-red color is made less intense in brightness to compensate.

Using a digital image processing computer program

Anaglyphs from monochrome images

For the example above the entire stereo card image was first converted to grayscale (shades varying between black and white). The right image was selected and pasted into a new document. This new document was then converted to color (it still looks black and white, with all channels identical). Returning to the original stereo pair document the selection was dragged to the left card image. The red channel was selected and the selection with that channel copied to the clipboard. The new color anaglyph document was selected as the destination window. Using the channels window the red channel was selected and the left image was pasted in. The images were flattened and saved as a jpeg without further adjustment.

Anaglyphs containing color information

Using color images, replace the red channel of the right-eye image with the red channel of the left-eye image. Here's how: select the entire right eye image (if the original is a crossed eye stereogram this will be on the left) and make a new document. Paste the right eye image in. Move the selection to the left eye image (with consideration as above for crossed eye stereograms) and using the channels dialog select the red channel. Copy the red channel from this source image. Select the new document and select the red channel. Paste the left eye image into the red channel. Depending upon the colors this might look rather good as a color image (but not exactly true to the original color). Eye sensitivity balance can be improved by selecting the green channel and reducing it using a linear curve selection (e.g. reduce to 12.5%). Select the blue channel and reduce somewhat less (e.g. reduce by 5%). This action compensates for the eye's lower sensitivity to red and its high sensitivity to green, but may induce bleed through (you may see ghost images on one side or to each side of an object). Without the anaglyph glasses the picture will appear reddish and somewhat dimmer, but the overall effect is improved when the glasses are used. The dimness can be corrected by increasing the overall brightness 15 to 20% and the contrast 10 to 15%. All of these adjustments will depend upon color balance, brightness, and contrast of the original image and the nature of the subject material.

Using color information, it is possible to obtain reasonable (but not accurate) blue sky, green vegetation, and appropriate skin tones. Color information appears disruptive when used for brightly colored and high contrast objects such as signage, toys, and patterned clothing when these contain colors that are close to red or cyan.

Depth adjustment

(The adjustment suggested in this is section is applicable to any type of stereogram but is particularly appropriate when anaglyphed images are to be viewed on a computer screen or on printed matter.)

Those portions of the left and right images that are coincident will appear to be at the surface of the screen. Depending upon the subject matter and the composition of the image it may be appropriate to make this align to something slightly behind the nearest point of the principal subject (as when imaging a portrait). This will cause the near points of the subject to "pop out" from the screen. For best effect, any portions of a figure to be imaged forward of the screen surface should not intercept the image boundary, as this can lead to a discomforting "amputated" appearance. It is of course possible to create a three-dimensional "pop out" frame surrounding the subject in order to avoid this condition.

If the subject matter is a landscape, you may consider putting the frontmost object at or slightly behind the surface of the screen. This will cause the subject to be framed by the window boundary and recede into the distance. Once the adjustment is made, trim the picture to contain only the portions containing both left and right images. In the example shown above, the upper image appears (in a visually disruptive manner) to spill out from the screen, with the distant mountains appearing at the surface of the screen. In the lower modification of this image the red channel has been translated horizontally to bring the images of the nearest rocks into coincidence (and thus appearing at the surface of the screen) and the distant mountains now appear to recede into the image. This latter adjusted image appears more natural, appearing as a view through a window onto the landscape.

Scene composition

In the toy images to the right, the shelf edge was selected as the point where images are to coincide and the toys were arranged so that ony the central toy was projecting beyond the shelf. When the image is viewed the shelf edge appears to be at the screen, and the toy's feet project toward the viewer, creating a "pop out" effect.

Viewing

A pair of eyeglasses with two filters of the same colors used on the camera (or simulated by the image processing software manipulations) is worn by the viewer. In the case above the red lens over the left eye allows only the red part of the anaglyph image through to that eye, while the cyan (blue/green) lens over the right eye allows only the cyan part of the image through to that eye. Portions of the image that are red will appear dark thrugh the cyan filter, while cyan portions will appear dark through the red filter. Each eye therefore sees only the perspective it is supposed to see.

In comics

These techniques have been used to produce 3-dimensional comic books, mostly during the early 1950s, using carefully constructed line drawings printed in colors appropriate to the filter glasses provided. The material presented were typically short graphic novels of a war story, horror, or crime/detective nature - similar in content to some modern Japanese manga. These genres were largely eliminated in the US by the rise of the comic book code authorities. Anaglyphed images were of little interest for use in the remaining comics, which emphasized bright and colorful images, unsuited for use with the viewing and production methods available at the time, which were usually red-green rather than red-cyan.

In fine arts

Some maintainers of internet web sites have added depth information to images of famous paintings, further processing these to produce color anaglyph images.

See also: Anaglyph image, Anaglyph glasses, Batman, Blue, Comic book, Cyan, Digital image processing, Graphic novels, Green, Manga