Needed: 19th Century stereo camera information #historical


fonstad@...
 

I'm hoping members of this group might be able to help out.

My name is Mark Fonstad; I'm a professor of geography at the University of Oregon. Amongst various things I do, I am interested in repeat photography to better understand environmental change. I have done some of this in the Colorado Plateau region (mostly in southern Utah), and I hope to do more.

Many of the 19th century photos taken of the American West were, as many of you know, in the form of stereographs/cards. In addition to my normal repeat photography tools/approaches, I'd like to be able to do repeat stereophotography, especially of close-in scenes like the one:

https://www.flickr.com/photos/usgeologicalsurvey/16308810373/
http://gotbooks.miracosta.edu/gonp/3Dcanyons/pages/esca14.htm

This pair was likely taken by Jack Hillers in 1872 in Tantalus Creek in southern Utah. What I would like to learn is specific information about the camera that he and the other main photographer on John Wesley Powell's 2nd expedition, E.O. Beaman, may have used to make these stereoscopic images. In particular, information on the spacing between the lenses and the lens focal lengths (even better would be the specific camera type), so that I can try to replicate these to some degree.

I don't really know how to find out this information, or who to ask as a starting point. I don't know if the stereo cameras in the 19th century were fairly standard or if there were a great variety of camera geometries. I thought I might ask if anyone in this group had any ideas on this question, or if you might be able to point me in the direction of others who might know about this.

Sincerely, Mark


Zane Healy
 

Finally an interesting post in this group! :-)  (I’m not very interested in 3D digital photography.)

I’m not sure how much data you’ll be able to find on specific photographers of that period.  Though most of my efforts find such information has been limited to the Civil War period.

Some states or towns have had books published on them.  I have one for Massachusetts, and one for Portland, Oregon.

The following information is from the “Collector’s Guid to 19th Century U.S. Traveling Photographers”, by Robert O. Brown (photobook@..., he’s in Forest Grove, Oregon, and also did the book on Portland photographers), mentions John K. Hillers.  I think you’re looking for information on him.  It sounds like he may have taken up Photography on John Wesley Powell’s 2nd expedition.  Halfway through the expedition, just as they were preparing to move down the Colorado river, E.A. Beaman quit, and took his darkroom with him.  Major Powell’s brother Clement had been working with Beaman, and when he left, John K. Hillers started working with him.  James Fennemore from Salt Lake City was brought in to finish Beaman’s work, and trained Hillers as his assistant.  Fennemore became ill in the spring of 1872, and left the expedition.  At that point Hillers became the appointed photographer.

The Bibliography leaves something to be desired, the above is from a book by Ralph W. Andrews, either “Photographers for the Frontier West:  Their Lives and Works, 1875 to 1915”, Superior Publishing Company, 1965.  Or “Picture Gallery Pioneers, 1850 to 1875”, Bonanza Books, 1964.  My guess is the second book, due to the time frame.  The information on Hillers is on page 19.

A couple businesses that might be of interest to you are Pacific Rim Cameras in Salem, and Blue Moon Camera and Machine in St. John’s (Portland).  Pacific Rim may have something appropriately period.  I bought my 5x7 Wet Plate stereo camera from Blue Moon, they also carry dry plates.

There are plenty of books on cameras of that time period, but very few mention very much on 3D cameras.  My guess is that it would have been either a “Half Plate” or “Full Plate” camera.  The lenses themselves would have been roughly 2 1/2 inches apart.  If you’re interested, I could point out a few books that have photo’s and very basic information on 3D cameras of that period.  Since Dry Plates were invented in 1871, I think that you’re looking for info on Wet Plate photography.  The prints themselves would be Albumen prints.

Other resources:
https://stereoworld.org/ - National Stereoscopic Association (NSA)
http://www.3dpdx.org/ - Cascade Stereoscopic Club (CSC), Milwaukie, Oregon

Zane




On Jan 19, 2020, at 6:05 PM, fonstad@... wrote:

I'm hoping members of this group might be able to help out.

My name is Mark Fonstad; I'm a professor of geography at the University of Oregon. Amongst various things I do, I am interested in repeat photography to better understand environmental change. I have done some of this in the Colorado Plateau region (mostly in southern Utah), and I hope to do more.

Many of the 19th century photos taken of the American West were, as many of you know, in the form of stereographs/cards. In addition to my normal repeat photography tools/approaches, I'd like to be able to do repeat stereophotography, especially of close-in scenes like the one:

https://www.flickr.com/photos/usgeologicalsurvey/16308810373/
http://gotbooks.miracosta.edu/gonp/3Dcanyons/pages/esca14.htm

This pair was likely taken by Jack Hillers in 1872 in Tantalus Creek in southern Utah. What I would like to learn is specific information about the camera that he and the other main photographer on John Wesley Powell's 2nd expedition, E.O. Beaman, may have used to make these stereoscopic images. In particular, information on the spacing between the lenses and the lens focal lengths (even better would be the specific camera type), so that I can try to replicate these to some degree.

I don't really know how to find out this information, or who to ask as a starting point. I don't know if the stereo cameras in the 19th century were fairly standard or if there were a great variety of camera geometries. I thought I might ask if anyone in this group had any ideas on this question, or if you might be able to point me in the direction of others who might know about this.

Sincerely, Mark


bojano54
 

Hi Mark,

In The Ribbon of Green: Change in Riparian Vegetation in the Southwestern United States (University of Arizona Press 2007, page 41): Hillers used a stereoscopic camera which produced stereoscopic 4-by-5-inch image on a 8-by-10-inch sheet of glass. [The camera could use] a normal lens with a focal lenght equivalent to a 50-millimeter on a 35-millimeter camera.

Greetings from France,

JMH



Zane Healy
 

I find it fascinating that a book that’s not on Photography would have that info, when it’s so rarely found in books on period photography.  Though the sizes sound a little off, and I’d love to know the author’s source of information.  The example that Mark provided looks to be closer to 5x7 aspect ratio (for each image in the pair), but I don’t think that makes sense either.  Does anyone know when the sizes of 4x5, 5x7, and 8x10 started being used?  I think they came about in the late 1880’s or the 1890’s.  My first 8x10 camera is an 1890’s Rochester Optical Company 8x10 dry plate camera.  A beautiful camera, but not practical.

I’m able to find mention of stereo cameras in the 1870’s being either 4x7, 4x8, or 5x8.  There doesn’t look to be that much difference in the cameras between 1855 and the 1880’s.  


1870’s and 1880’s

If anyone has better insight into this, I’d sure appreciate knowing.  Part of why I find 4x5/8x10 or even double 5x7’s hard to believe has to do with contact printing.  This means the negatives would need to be close in size to the prints.

Zane




On Jan 20, 2020, at 2:34 PM, bojano54 <jean-marie.hering@...> wrote:

Hi Mark,

In The Ribbon of Green: Change in Riparian Vegetation in the Southwestern United States (University of Arizona Press 2007, page 41): Hillers used a stereoscopic camera which produced stereoscopic 4-by-5-inch image on a 8-by-10-inch sheet of glass. [The camera could use] a normal lens with a focal lenght equivalent to a 50-millimeter on a 35-millimeter camera.

Greetings from France,

JMH




Mary Paul <maryp@...>
 

Gisela Will The Keystone Re-creation Project:
Gisela Will is a German stereo photographer who has been asking stereo-photographers to recreate Keystone Stereo card views.
When people travel somewhere they have a Keystone stereo card of the area they try and recreate the same modern stereo view as closely as possible.
In 2013 Gisela presented a show of these recreations, some of which were from the US western parks. 
People have been doing more of these recreations around the world so there probably are more of the Western US out there now.
I don't have Gisela's e-mail or know if she is part of this group.
Mary in Berkshire, UK


Antonio F.G.
 

On Mon, Jan 20, 2020 at 10:20 AM, <fonstad@...> wrote:
What I would like to learn is specific information about the camera that he and the other main photographer on John Wesley Powell's 2nd expedition, E.O. Beaman, may have used to make these stereoscopic images. In particular, information on the spacing between the lenses and the lens focal lengths (even better would be the specific camera type), so that I can try to replicate these to some degree.
Some guys in this list have already told something about the likely focal length, stereo base and plate size.

But it is possible to calculate all this data EXACTLY from the photo pairs themselves, which is an interesting mathematical problem.
First you need to measure the real X,Y,Z positions of several features in a scene. It should be an scene that has not changed significantly in 150 years, such that the features could be recognized. A survey instrument like this can provide X,Y,Z positions respect same arbitrary point (i.e. a nail fixed in the soil). The bearing is also arbitrary.

Now it is possible to set the equations of the projection of each X,Y,Z point of the scene (measured by survey) into the corresponding xp,yp point of the ancient photo plate. The unknowns are the Focal Length (FC), the Optical Center of the old plates (xp0,yp0), the positions of the L,R cameras (XL,YL,ZL,XR,YR,ZR) and the pointing angles of each camera (PHL,PHR,PVL,PVR,RTL,RTR). A maximum of 15 unknowns, so we need a minimum of 4 surveyed points as each one provides 4 equations: xp,yp of L and R photo (xpL,ypL,xpR,ypR).
This is solved by a non-linear least squares procedure (iterative). It is best to have more points than 4 to reduce the errors.

   Regards
       Antonio


fonstad@...
 

Hi Antonio (and others),
Yes  that type of solution has occurred to me (I have a small amount of photogrammetry background). And I think it would work for a few pictures Hillers and others took in places with close-up bedrock scenes, where there would be little change in the past 150 years. If a few of these worked, I could use the computer camera information for the other photos where there aren’t non-changed points in the scene. With modern structure from motion software, it is easy for me to make 3D point clouds of a modern scene from which I could get the needed 3D information for the process you describe. Doing all this would depend on figuring out the best photos, going to Utah for fieldwork and doing some surveys, but it is a possibility. And could be repeated for other photographers, potentially.


fonstad@...
 

I meant to write “computed camera information”


John Rupkalvis
 

Mark Fonstad  ~

Here is one site that includes some of the 19th century stereo cameras, although many of the ones shown are from the 20th century (mainly 1950s era):


However, there were a large variety of stereo cameras made in the 19th century, many of which were one-offs (only one of a particular type being built).  Often the photographers themselves designed and built their own cameras.  While most were designed specifically for stereo, some were conversions of monoscopic cameras.  

Some of them were "box camera" designs, using glass plates or metal plates before the use of flexible film became common, and these looked similar to the 1950s stereoscopic film cameras.  In fact, some of the 1950s cameras were patterned after or even manufactured by companies that made plate cameras in the 19th century, like for example, the French Verascope.  

Most of the 19th century stereo cameras were bellows-type.  Some of them used a pair of bellows side-by-side with a double lens board (a single board with two lenses mounted in it).  Others, also with double lens boards, used a single bellows with a septum divider inside.  

In the 19th century there were a variety of different lens spacings (interaxials), sometimes determined by the center distance of the plate size.  However, most were close to typical human interpupillary distances, being around 60mm to 70mm (there was no standard back then, so there was some small amount of variance even among these).   

I hope that this is helpful.

John A. Rupkalvis
stereoscope3d@...

Picture


On Mon, Jan 20, 2020 at 8:20 AM <fonstad@...> wrote:
I'm hoping members of this group might be able to help out.

My name is Mark Fonstad; I'm a professor of geography at the University of Oregon. Amongst various things I do, I am interested in repeat photography to better understand environmental change. I have done some of this in the Colorado Plateau region (mostly in southern Utah), and I hope to do more.

Many of the 19th century photos taken of the American West were, as many of you know, in the form of stereographs/cards. In addition to my normal repeat photography tools/approaches, I'd like to be able to do repeat stereophotography, especially of close-in scenes like the one:

https://www.flickr.com/photos/usgeologicalsurvey/16308810373/
http://gotbooks.miracosta.edu/gonp/3Dcanyons/pages/esca14.htm

This pair was likely taken by Jack Hillers in 1872 in Tantalus Creek in southern Utah. What I would like to learn is specific information about the camera that he and the other main photographer on John Wesley Powell's 2nd expedition, E.O. Beaman, may have used to make these stereoscopic images. In particular, information on the spacing between the lenses and the lens focal lengths (even better would be the specific camera type), so that I can try to replicate these to some degree.

I don't really know how to find out this information, or who to ask as a starting point. I don't know if the stereo cameras in the 19th century were fairly standard or if there were a great variety of camera geometries. I thought I might ask if anyone in this group had any ideas on this question, or if you might be able to point me in the direction of others who might know about this.

Sincerely, Mark


Georg Klein
 

Gisela Will contact
https://hamburg.stereoskopie.org/kontakt


John Rupkalvis
 

Mark Fonstad  ~

You mentioned photogrammetry.  This part of your background might be useful for the determination of what the stereo baseline was on the particular camera used for the referenced stereo cards.  

Most photographers in that era used one specific camera for all of their stereoscopic work.  So, what you want to do is find a stereo card (or better, several stereo cards) by the photographer in question, that has some object, or better, objects, of a likely known size, such as a person, at a likely known distance, and "reverse engineer" the metrology.  

Yes, it still is somewhat of a guessing game, but knowing certain facts can at least lead to an "educated guess".  

As you pointed out, one of the things that is very helpful is to know the lens focal length.   Older camera designs often used a single focal length lens, that focal length usually being based on the corner-to-corner diagonal of the image on the film or plate.  Although enlargements were made in the latter part of the 19th century, most stereo cards were made as contact prints (same size as the original camera negative image).    

A modern digital camera, especially one with a wide-range zoom lens, can be very useful in determining the parameters that are most necessary for "matching" an original stereo card.  You want that card, or an exact size duplicate of it, in front of you when you are setting up to duplicate the scene.  Check the near point (closest object) and far point (furthest object) in the image, and use the zoom lens plus the subject distance to match as closely as possible the comparative sizes of those two objects on the viewscreen.  

Before you go out in the field to do the actual photography, shoot some tests of an easy-to-measure grid, which will then allow you to calibrate your viewscreen to match the stereo card proportions.  You can even use this information to make a transparent overlay grid to tape to your camera viewscreen, the squares of which you will have precalculated to match the actual sizes/distances on the original stereo cards.  

Having a laptop or netbook or tablet computer with you will let you check your work (comparing to the actual stereo card image pair) at the location, so that you can modify and reshoot if necessary.   

John A. Rupkalvis
stereoscope3d@...

Picture


On Tue, Jan 21, 2020 at 11:21 AM <fonstad@...> wrote:
Hi Antonio (and others),
Yes  that type of solution has occurred to me (I have a small amount of photogrammetry background). And I think it would work for a few pictures Hillers and others took in places with close-up bedrock scenes, where there would be little change in the past 150 years. If a few of these worked, I could use the computer camera information for the other photos where there aren’t non-changed points in the scene. With modern structure from motion software, it is easy for me to make 3D point clouds of a modern scene from which I could get the needed 3D information for the process you describe. Doing all this would depend on figuring out the best photos, going to Utah for fieldwork and doing some surveys, but it is a possibility. And could be repeated for other photographers, potentially.


Antonio F.G.
 

On Tue, Jan 21, 2020 at 01:21 PM, <fonstad@...> wrote:
 I think it would work for a few pictures Hillers and others took in places with close-up bedrock scenes, where there would be little change in the past 150 years.
Finding those places may be a challenge in its own right. I bet those nice water canyons are now wider and deeper than they were 150 years ago


With modern structure from motion software, it is easy for me to make 3D point clouds of a modern scene from which I could get the needed 3D information for the process you describe.
If I understand well, with computer graphics you can reproduce the images that two cameras would take to a cloud of 3D points, IF YOU KNOW THE POSITION OF THE CAMERAS. But you do not. What you know are the 2D position in the photo plate of each of those 3D points. It is the inverse of what computer graphics does.

Regards
    Antonio


John Rupkalvis
 

Using the principles of photogrammetry, it is often possible to determine the position of two different cameras.  Each camera or camera lens will have its own point-of-view, which has a specific "sight-line".  If you plot a "map" of the scene, using objects "seen" by both cameras/lenses, you can re-create those two sight-lines on your map.  Using triangulation, along with "relative size" proportions, you may be able to locate the position (location along the sight-lines) that the cameras would have been on your re-created map.  This is similar to the procedure that forest rangers use to locate a fire on a map from the position of the sighted smoke plume as seen from two different stations (observation towers).  They use topographic maps of the area in question.  You essentially create your own miniature topo map of the area seen in the stereo pair.  From that, you will be able to deduce where each camera or camera lens was at the time that the original stereograph was made.  The key is to find and locate two or more different objects that were at two or more different distances from the original camera(s)/lens(es). 

John A. Rupkalvis
stereoscope3d@...

Picture


On Thu, Jan 23, 2020 at 2:23 AM Antonio F.G. via Groups.Io <afgalaz=yahoo.com@groups.io> wrote:
On Tue, Jan 21, 2020 at 01:21 PM, <fonstad@...> wrote:
 I think it would work for a few pictures Hillers and others took in places with close-up bedrock scenes, where there would be little change in the past 150 years.
Finding those places may be a challenge in its own right. I bet those nice water canyons are now wider and deeper than they were 150 years ago


With modern structure from motion software, it is easy for me to make 3D point clouds of a modern scene from which I could get the needed 3D information for the process you describe.
If I understand well, with computer graphics you can reproduce the images that two cameras would take to a cloud of 3D points, IF YOU KNOW THE POSITION OF THE CAMERAS. But you do not. What you know are the 2D position in the photo plate of each of those 3D points. It is the inverse of what computer graphics does.

Regards
    Antonio


fonstad@...
 

I'd like to thank everyone for the suggestions made over the past few days. While I don't have exactly the information I'm ultimately looking for, I do have a better idea of how to get towards that information, and from more than one angle. I'll leave you with an image of Jack Hillers during the Powell Expeditions examining one of his glass plate negatives.