The following text provides some further informations on stereoscopic images, especially different viewing methods. It is partially based on a text written by Larry Berlin, 1999
|
|
Stereoscopy enables us to create and view three dimensional images. In this case, the 3D impression is not caused by means of perspective or lighting and shading, as in "normal" (flat) images; it is the real, first class depth sensation, gained by using both of our eyes. Each eye is provided with a slightly different view of a given scene, just the same way we are accustomed to when looking at the real world.
For Stereoscopy to work it is important that the two images strictly obey to the rules of perspective and that for every object to be depicted threedimensionally, there are corresponding, alsmost identical image elements, one at each half image. When our eyes are presented with such corresponding image elements, they try to converge — the halfimanges fuse, as we call it. From this convergence of eye axes, and from the slight variations in convergence when looking at several objects, there arises the sensation of depth essential to stereoscopy. And from the minute incompatibilities and discrepancies between the corresponding image elements to be fused, there arises an increased sense of presence or feeling of reality often witnessed as accompaning the depth impression of stereoscopic images.
The classical way to create stereo images is using a special, two-eyed stereoscopic camera. If no such camera is at hand, we can simulate it by subsequently taking two images while laterally shifting the camera. But, indeed, stereoscopy is older than photography, and before beeing able to create stereo photographs, stereoscopic drawings were created by means of geomety and perspective. Nowadays, with computational power beeing freely available, this can be done in an automated end elaborated way.
In order to gain the depth impression, the stereoscopic image has to be presented accordingliy, so that each eye is provided with its own half image. With printed images or slides, this can be achieved by optical devices like mirrors or lenses. But maybe the simplest method ist, to squint or look cross-eyed at the image. For this to work, the stereoscopic half images must be arraged in cross-eyed order, i.e. the left halfimage has to be mounted on the right side of the stereoscopic pair. This may seem silly, but after a little bit of training is convienient and easy to get acquainted to.
|
— Try it! — |
A good starting point is to use the tip of a pen. Place yourself in front of a stereoscopic pair suited for cross-eyed viewing. Place the pen half way in between and look at the pen. Now, the viewing rays or axes of your eyes should meet or cross at the pen's tip, and as a consequence, the left eye sees the half image mounted on the right side of the pair behind the pen and vice versa. Finally, the only thing left to learn is shifting the focus back from the pen's tip to the images behind without changing the direction you are looking at.
Quite a different viewing method uses anaglyphic images: the two halfimages are presented superimposed, but the left image is colored green and the right one is colored red. For stereoscopic viewing, you need special spectacles, where the left glass is colored red and the right glass is colored green. Because you won't see much of a red image through a red tinted glass and vice versa, each eye only receives one half image, thus creating a stereoscopic impression. This method works very well on computer monitors. There are several possible combinations, e.g. red/green or green/red or red/cyan or blue/red etc.
A more expensive method for viewing stereoscopic images on a monitor utilizes LCD shutter glasses. The monitor must be operated in interlaced mode and the two halfimages are to be presented line interleaved, while a special electronical device switches the LCD glasses in sync with the monitor's vertical image frequency.
But, finally, I should mention another aproach, that, in spite of beeing somewhat complicated and "low-tech", deliveres very high quality stereoscopic viewing, maybe the best possible quality you can get out of a computer monitor: Use a conventional camera and take color slides directly from the monitor. Choose highest possible resolution and color depth. Optimally adjust brightness, contrast, color balance and exposure. Look at the slides with two magnifying lenses or with a special stereo slide viewer!
PNS and JPS files are a convenient way to provide stereoscopic images on the internet. Instead of having to create a bunch of different sizes and formats for the different viewing methods, or worse yet only provide one type of stereo image, you can use JPS or PNS files and be assured that everyone will have optimum access to your images.
The file is a standard JPG file that contains a stereo pair in a cross-eyed format. (PNS is analogous, but based on the PNG image format and uses lossles compression). It is recommended not to use progressive JPG and avoid large amounts of jpeg compression as it can degrade the appearance. The changed file extension from JPG to JPS allows some Plug-in for web browsers to recognize that the file contains a stereo image. Such plug-ins are able to process that image for presentation in one of many viewing methods, as chosen by the user.
The major benefit is that everyone has the full sized, full color stereo image available to view.
It gets converted to anaglyphic or freeviewing versions by software in a plug-in or an applet.
The software usually allows the viewer to change the viewing size as well as explore all the viewing types.
One file, two pictures, any stereo viewing method you desire!
There are two primary ways to use JPS images on a web page. One is linking directly to the JPS file. For this to work it is necessary that the user installes a plug-in for JPS stereo images and that the web server serves *.JPS files as MIME type image/x-jps. The latter is done by updating the server configuration and adding a new record for this MIME type. Some major Internet Service Providers have already done so. This is only a small change and there is no charge for adding this mime type. But in many situations changing the server configuration is not an option. But without the right MIME type the visitors end up downloading garbled text. The reason this happens is that the server assumes any file type not recognized to be a text file. Another drawback is that plug-ins usually are available for the windows platform exclusively.
The other method to include stereo images on a web page circumvents such problems by using some JAVA Stereoscope Applet. Applets require JAVA runtime environment, but that is available for most platforms in one configuration or another. JAVA may impose performance problems for some older machines and generally needs some startup time. Besides that, the usage for the web page visitor is identical: There will be a small preview embedded in the web page content, but the image in full resolution has to be explored in a separate window.
Stereoscopic images are always special and don't integrate in the same smooth way with
other web page content like common "flat" images do — not very surprising, as they really
open into an additional dimension.
You should note, that some plugins and some JAVA Applets provide a standalone stereo viewing application to be used locally as well. Personally, I have installed the «depth charge» plugin and the Stereoscope Applet by Andreas Petersik on my machine. All the ressources mentioned here are available for free.
|
|
|
|
|
| |||
| StereoPOV | by Ichthyostega  |
home images usage download | |