PyMOL Stereo viewing


Stereo Viewing Title

PyMOL Stereo viewing

PyMOL can display stereo in many ways. Unless special equipment is available, the method is to split the PyMOL 3D view area in two with the command stereo on (and stereo offto turn it off.) The alternative is to use the menu cascade: Display > Stereo.

However, there are 2 ways for human eyes to converge the 2 images into a stereoscopic view: crosseye  (also called divergent) and walleye (also called parallel.)

PyMOL will have a default option that can be switched with commands stereo walleye or stereo crosseye. The alternative is to use a menu cascade, for example: Display > Stereo  Mode > Wall-Eye Stereo

PyMOL display stereo mode menu

If the image looks weird you probably need to change this setting!

regular, convergent, and divergent visualization
Regular, crosseye, and walleye visualization. Black dots represent viewer’s eyes.

How to watch stereo images

While rotation provides useful 3D perception, stereo viewing is outstanding in visualizing complex spatial relationships.But seeing stereo requires skill and can be taxing on the eyes. Emeritus Professor of Chemistry Gale Rhodes wrote an excellent introduction to stereo viewing [Archived 168 times] originally written for Rasmol (a sort of ancestor to PyMOL.) The Wikipedia page on Stereoscopy also provides  detailed “how to” information.

With either instructions the viewer is first likely to see 4 images, that will resolve into 2 flat images on either side that should be ignored, and focus on the center image that will appear in 3D as the result of the merging of the 2 images in the viewer’s brain.

Stereo images in publications used to be more common. But “smart authors” manged to provide an illustration for both crosseye and walleye viewers by presenting a triple image, placing the image for the right eye on both sides of the image for the left eye, i.e. pattern R L R. Crosseye viewers use the 2 images on the left, and walleye viewer the 2 images on the right.

stereo pairs of ethanol molecule

The helper dots (see autostereogram below) at the top can help the viewer merge the images.

This fun practice page [Archived] can help with the process step by step with supporting images. This other tutorial by the stereosite [Archived] can also be a useful guide “because the technique is not so much intellectual as it is physiological and psychological.”

Autostereograms might help

An autostereogram is a three-dimensional image puzzle with a hidden 3D scene unrecognizable until viewed properly (see images within autostereogram Wikipedia page.) Autosterograms are meant to be seen in the walleye fashion. Crosseye viewers would see an hollow version of the 3D scene. Autostereograms often have 2 helper dots or simple pattern at the top. The viewer can defocus the vision to try merging the 2 dots which may be less confusing than staring at the complex drawing below. The web site has “how to” and practice images:

Instructions to see stereogram

– Stereogram viewing instructions and practice
– Stereogram viewing practice image

A useful answers page is provided to show the 3D hidden scene by hovering the mouse over images. The practice image scene is that of a coloring book image named “1. Good Catch“.

Other practice images and tutorials

Here are a few more options of stereoscopic images:

– Wikipedia article on Stereoscopy contains stereo pairs from long ago.
– The Johnson-Shaw Stereoscopic Museum Fan Site and vintage stereo pairs.
Riverside exhibit museum image of the aftermath of the 1906 San Francisco earthquake [Archived].
– The Stereoscopy blog basic tutorial [Archived]
– The UK-based Stereoscopic Society tutorial [Archived]
– The Library of Congress
– Videos: 50 amazing 3D Stereograms / Magic Eye; Relaxing stereograms
– Video: How to reveal hidden 3D scene in autosterogram with PhotoShop (How to un-hide a “Magic Eye” picture in Photoshop)
– Depth perception and deficiencies explained by Optometrists



Image Credits: top image background wallpaper by Pixabay psychelops. Ethanol structure provided by Prof. Brian G. Fox, rendered in PyMOL.