FMF - Friends of Minerals Forum, discussion and message board

[Elise]

In the Smithsonite thread, Gerhard posted a cross-eyed stereogram which was extraordinary https://www.mineral-forum.com/message-board/viewtopic.php?p=10549&highlight=#10549

I collect early 1900s stereograms of various themes, mostly Norwegian, but I am also fascinated by mineral stereoviews. I have one book of them and also can recommend the 1987 American Mineralogist article by Wendell Wilson and Dr. Chamberlain on the art. I can't focus the cross-eyed ones; I have to use a viewer or I get seasick, never-the-less, maybe we can start a thread of mineral stereograms and I'll see how any of my viewing scopes work on a computer screen image!

Best wishes,
Elise

[Peter Megaw]

I am traveling without access to my library, but there is also a cute book of crystal drawing sterograms that are used for teaching crystallography...perhaps one of you knows who produced it?

[Duncan Miller]

[Jesse Fisher]

The article referred to by Elise was in the Mineralogical Record, not the American Mineralogist (Wilson, W. E., and Chamberlain, S. C., 1987, Mineral Stereography, Mineralogical Record 18(6), 399-404). They give a pretty good description of how to take and view such photo pairs. Basically, the specimen needs to be mounted on a rotating platform, directly on the axis of rotation, and photos taken between 5-10 degrees of rotation. They also give an example of a stereo camera, which will simultaneously take two side-by-side images at slightly different angles, thus avoiding having to rotate the object being photographed. These things were popular during the 1950s. I haven't seen one in a while but I'm sure they're some still floating around out there on e-bay.

[Vinoterapia]

I know that Fuji is selling a 3D camera, although I have not seen any independent review about such gadget.

Regards.

José Luis.

[Elise]

[Gerhard Niklasch]

Many thanks for the interest in my humble excursions into the third dimension.

I guess once a proliferation of 3D movies results in widely available and affordable 3D monitors (one technology has them alternating the left and right image at a fairly high frequency, synchronized to electronic shutters in goggles, so that each eye only gets to see the image intended for it), the viewing will become trivially easy... but for now, we'll have to kludge around with putting the images side by side.

First, let me explain why I (only) post crossed-eye stereographs, which is also the convention followed on mindat (where e.g. Modris Baum has submitted a huge number of micro stereos, mostly from Mont Saint-Hilaire: https://www.mindat.org/gallery-11235.html ).

We don't expect everyone to possess a stereo viewer, let alone one that lends itself to use on a computer monitor. Without one, parallel-view stereo pairs are severely limited, since most people cannot bring their eyes to diverge. Thus the physical separation of corresponding points in the left and right image must not become larger than the inter-eye distance, and must preferably be kept a little shorter. The inter-eye distance varies among people, it can be less than 60mm. So each partial image would have to be kept narrower than that. But I have no control over other people's monitor resolutions, so in terms of pixel counts, the images would need to be kept very small indeed! With crossed-eye pairs, it's no problem to have each image as wide as 10cm or so, significantly wider than the inter-eye distance.

Moreover, forcing the eyes to look approximately parallel (as if converging to a target in the far distance) while still focusing on the distance to the screen is a bit harder than the opposite trick, focusing on the screen while having the eyes converge to a point approximately a third to half as far away from them as the screen (though it requires practice either way). Personally I can't quite focus on parallel-view pairs at all (e.g. on the rotating models displayed on the webmineral pages). So I don't feel comfortable producing pairings which I can't really see properly myself.

As to how I'm taking mine:

Either parallel or converging shots work, but the lighting must be exactly identical, so rotating or translating the specimen without also rotating or translating the light source(s) -- along with anything that might reflect or scatter light towards the specimen! -- won't work too well. Varying the camera position whilst keeping the specimen and light sources motionless is much simpler.

I do have a vintage (1950s) stereo camera, but with its fixed inter-lens distance of about 63mm, it's only suitable for scenery on human scales, not for close-ups.
(Nor for landscape, where you may want to take parallel shots from vantage points up to a few meters apart, and pay attention to which way the clouds are moving.)

Many of my stereo pairs are simply taken through the eyepieces of my stereo microscope, which takes care of reproducible viewing angles.

Non-microscope close-ups are done either hand-held (well, "elbow tripod") from above or sideways with a small table tripod, and then I typically take three or four shots in quick succession, shifting the camera by a few cm (less than the normal inter-eye distance) from each shot to the next. Afterwards I select which pairing works best. Depending on the geometry, the depth may appear a bit reduced or exaggerated. Shooting four pictures instead of one takes next to no extra time.

Much time, however, goes into the postprocessing. The two partial images must be lined up as precisely as possible. They must match to better than half a percent in scale, so if my hand-held shots came from camera-target distances varying by just a few mm, that needs to be compensated. They must not be rotated around the viewing axis against each other by more than a fraction of a degree, which almost always requires compensation. Corresponding points must appear at the same vertical positions in both sub-images, to within a pixel or two. Moreover, the brightness, contrast, and color balance need to match fairly closely (this is somewhat less critical).

It's less of a problem if one image is less sharp than the other - the brain routinely compensates for this, and indeed a stereo pair often looks a bit crisper than either of its partial images alone would at the same scale.

Crystal faces reflecting a light source into one of the two lens positions only, or much more strongly than into the other, can sometimes provide a striking effect, but most often they're a massive nuisance or even a show stopper - the pair would no longer blend properly. This must be taken into account already when setting up the shots.

Finally, there's no substitute for experience. Experiment a lot and find out for yourself what works! Vast numbers of raw shots end up in the trash can, only a small fraction finally turns into presentable pairs. (One more reason for using a digital camera rather than the old film-based stereo gadget...)

[Duncan Miller]

Hello Gerhard

Thank you for your quick and detailed response to my query about how you take your stereophotos. The advice to keep the specimen and illumination stationary is particularly helpful. This morning I experimented by simply altering the angle of view by turning the head of the mini-tripod a few degrees left and right. The result was acceptable, while not a work of art, and a lot better than I have achieved before by translating or rotating the specimen. Hopefully this might encourage others to try their hands at this. It is not as difficult as it first appears.

Duncan Miller

[Gerhard Niklasch]

Glad you're finding this helpful, Duncan!

Returning to the seasickness aspect, Elise:

To some extent - just like gaining one's sailor's legs - this is a matter of training too.

Until then, whenever you look with crossed eyes (or with parallel eyes onto a non-crossed pair), it helps to hold your head perfectly still and merely let the eyes wander.

Any head motion, with the eyes at anything other than their normal convergence angle, will result in eyes and inner ear sending conflicting feedback about the motion to the brain. To our tree-living distant ancestors, that was a Bad Thing. Not only did it interfere with coordinating their movements, it also suggested that at least one of their senses wasn't functioning as it should, and the most likely cause for that would have been mild food poisoning... whence those who developed a reflex to empty the stomach were more likely to become our ancestors than their more robust cousins.

Hope this helps!
Cheers, Gerhard