Introduction

I'm a big fan of 3D photography. To me it makes the experience much more lively than 2D. I'm not sure how to describe it, but perhaps a good comparison would be riding along in a tour bus compared to exploring at your own pace and choosing which side of the street you want to be on. For years I kept an eye on 3D cameras and their prices were out of my budget, but I had a nice suggestion on 3D and Stereo Photography ; sadly dpreview.com is set to close in 2 days. 😞

I look forward exchanging about 3D photography. 😃

Your welcome! (and I hope will be more than just the two of us 😀)

For a pleasant experience of three-dimensional immersion, viewing devices are needed first of all.
It is also preferable to train to use the cross viewing for the 3D experience without specialized devices. This also requires a minimum of knowledge about how different 3D formats can be converted into others.

Sharing your own achievements is welcome. 🙂

A curious grasshopper.

a nice capture of a grasshopper

This is a still taken from a video clip - two cameras both recording at 4K 30fps - parallel view

Well synchronized and with pleasant depth. 👍

Parallel viewing. Stereo image taken with 2 cameras on bar.

Hi,
I am not new to stereo photography, but have not done much for many years. Yes, a stereoviewer of some type is the best way of viewing 3D, but very few people will have anything. Cross viewing is good for some people, but I cannot do it, but I am able to view parallel L/R images with my eyes (as long as the images are small).

I have many stereo images I would like to share, but I am wondering which format is best for everyone. The following is a Left/Right/Left set of images. View the 2 on the left for parallel viewing and the 2 on the right for cross viewing.

I can see them all, in any format, that's why I prefer them in the largest possible screen. In small parallel format, can be see in 3D, but they are not impressive.

For example, I download the 3D pairs to the PC and insert them into a stereoscopic player where they can be converted and played from any 3D format to any other 3D format. 🙂

I have not looked at stereo imagery for a long time, so I did some more investigation. I can actually see images of upto 12 cm across each image (24cm for both) using my eyes only on a good monitor. I can see images twice this size using a Pixi viewer. But eyes only seems to be just as good, it seems to be a cleaner image and I can take in the whole 3D image. A mirror stereoscope is the best device I have used, but sadly I do not have one any more.

Assuming that you accept some suggestions, I allowed myself to make some changes to your image.
On the big screen, a more complex misalignment and a significant violation of the stereoscopic window in the lower half of the image is immediately evident.
So I corrected the misalignment, the integration in the stereoscopic window and expanded it in height so that it fits completely in the height of the 4K screens (2160 pixels high).

What do you think about how it looks in the new version?🙂 (your image is well collected and is deep👍)

The differences are easy to see in the Red-Cyan montage.

Thanks for the positive feedback.
But let me start by saying that I used to teach photogrammetry to university students. One of the exercises we did was to produce a set of stereo images and have them anaylsis the results (get some 3D measuremants). Some students did struggle to see 3D images at all, but there was a large difference in how much of the 3D effect could be seen by each student.
But this is now done using software such as PhotoModeler, and all the fun has been removed.

What you are seeing at the bottom of the image is caused by your brain not being able to resolve the large amounts of X-parallax differences (because the foreground is much closer than the rest of the image). Funny thing is, I can see the foreground in perfect 3D (maybe I am just lucky). I am not sure what you did, but your adjustments made the 3D image easier to see. Sometimes just seperating the left and right images will make them easier to see in 3D (especially when using mirror stereoscopes), and maybe this what you effectly did.

Now most of my stereo images do have large amounts of depth, and this will be a problem with many of them. Below is an extreme example, and I can not resolve the X-parallax at the bottom (the bench and red flowers). The best solution in this case would be to crop the image so the bottom part is removed.

Now if I seperate them some, then they are easier to view with the Pixi and I can almost see the entire image in 3D.

thanks
Eric

So, will be a long story...
1. How do you take your stereo pairs? (with two cameras, with a dual lens camera, with shifting a single camera, with a single camera and a mirror splitter?)
2. There are very well defined rules about how to take the stereo pairs and about how to mount them to make a stereo pair. Must to take in consideration the "stereoscopic window" and the rules for alignment.
3. If the rules are broken, the stereoscopic images are difficult to look at. Viewers bear differently the ocular stress due to misalignment (too much convergence, divergence, vertical misalignment that forces the eyes to look in different vertical directions, each).
4. The depth and convergence of the images must be thought in advance depending on the size at which they intend to be seen.
5. Images with very deep depth seen on large screens can induce divergent gaze forcing at the farthest objects in the image, which should be avoided.
This situation can lead to settings that violate the rules regarding the stereoscopic window.

The objects in the image located between the eye and the stereoscopic window must be chosen so that they do not touch/intersect the edges of the window.

An example is in the fractal image called "out of screen", in which the side walls of the interior of "cube" are positioned very close to the stereoscopic window, and the central object is positioned between the eyes and the window so that on a large screen it gives the impression that you can put your hands behind him so that he can be "held in your hands".

Your images are harder to see at large because they are severely misaligned vertically and the stereoscopic window in the bottom half of each image is also violated.

So, I made some corrections on your last image, but the stereoscopic base used was much too large considering the distance to the closest objects.

In the first correction, I only corrected the vertical alignment.

The second image represents an intermediate version of correcting the framing in the stereoscopic window.

In the third image, I mostly corrected the framing in the stereoscopic window, leaving only the intersection in the lower right area.

Yes, I see now there there are some misalignment. I did a very quick job of creating the stereo setup. But again I can see these in perfect stereo so did not notice the slight misalignments. Did not know about these rules for stereo viewing, I was more interested in getting a good depth of field.

I cannot remember how these were taken. I use either two cameras on a bar (for action shots), or shifting a single camera (usually at least 3 different base distances). Much of my stereo images are used for measurement so having large base distances helps with intersection angles but makes 3D viewing more difficult.

You probably see them in parallel at very small sizes. Exposed full screen, the last one cannot be seen in any format. On my screen, the vertical gap is ~22mm, to which is added the excessive horizontal separation, a situation to which the eyes cannot adapt.🙂

For measurements, an extended stereoscopic base helps with accuracy, especially at long distances.
In 1992 I published a methodology for determining the three-dimensional coordinates of some points in space inaccessible for direct measurements in underground spaces (caves), using pairs of stereoscopic images, so I understand what it is about.🙂

Yep, your right. There is large y-parallax in that stereo pair (I get about 8mm on my monitor). But my problem is I can see it very well in 3D even at a large size. And this probably means this set was taken using a single camera. I will have to be much more careful aligning the L/R images.
Nice to see that you have put the stereo images to good use. I have also worked on some projects using stereo pairs, but normally its multiple (single) photos from different camera stations.