More Panoramic Fun: A Preliminary Report on the Horizon S3 Pro

[OlDoinyo]

The Horizon line of swing-lens cameras, made until recently by KMZ of Krasnogorsk, were a relatively common type of panoramic film camera from the 1950s until about 2012, when production may have ceased (I am not sure about this point.) NOS examples of their later products can still be had for non-astronomical prices, and emboldened by my success with the Spinner, I decided to obtain one:

 

 

Although limited to a 120 degree field of view, this camera is less of a toy than the Spinner, has far more options to control exposure, and offered the possibility of better image quality. But what about the UV bandpass of the lens? A pinhole test was awkward to stage, but I managed to get one (posted here as a composite:)

 

Not as good as the Spinner's but far from terrible--at least as good as the Autocord's, and perhaps approaching that of the old Asahi 35, which passes down to 345-350  nm. So far, so good. But what to do about a filter? This camera takes weird little bespoke filters which only the factory made, containing small glass disks. So I ordered a couple of 15mm ZWB1 disks (my wife teased me about ordering a package containing black dots) and with my son's help and a 3-d printer, a modified filter was improvised:

 

 

Next step: load up the trusty Delta 3200 and go shooting! Unfortunately, the camera chose this roll of film to start manifesting a nasty light leak around the perimeter of the rotating turret (I sent it in for repair based on this finding.) The following images are thus somewhat spoiled, but I present them here anyway because they tell us a great deal about the possibility of this camera, and I trust the leak will be a fixable problem. The images were taken with the camera on a monopod, though handholding would be reasonable for some; typical exposure was 1/60 or 1/125 second at f/11. Some have complained that the camera's lens goes soft at f/16 but when I tried that I did not observe this in UV.

 

The least spoiled of the images is this panorama of Medano Creek in Alamosa County, Colorado:

 

 

Exposure looks good at a working ISO of 50, and detail in the distant trees looks good enough that there seems little trouble with focus shift.

 

A panorama of Cottonwood Pass in Gunnison County turned out similarly:

 

 

Two yellow road signs at the right margin turned out black as expected:

 

 

There is no horizontal fall-off of sharpness toward the side as there would be with a fixed-lens camera. There is a small amount of vertical fall-off.

 

In the fading light, I tried a longer exposure (1 second) of the Twin Lakes reservoir area in Lake County.

 

 

The restroom building in the midground looks reasonably sharp, the acuity being limited by film grain as much as by the optics:

 

 

And finally, lest anyone doubt that these are UV images, one may examine these roadside Helianthus petiolaris specimens excerpted from a different frame:

 

 

In the future, when the camera is fixed, I would like to try this again, perhaps ultimately with finer-grained film. Aside from the light leak, everything worked reasonably well, and the pictures are definitely a step up in quality from those obtained with the Spinner, though they lack the full-circle sweep of those.

[dabateman]

Interesting camera. 

You get 24x54 images and the 28mm lens focus is fixed. 

 

I look forward to more photos once you sort out the light leaks. 

 

Fun that you don't need an IR blocker as the film can't see IR. That leaves you open to more fun UV photos. 

[Andy Perrin]

These are nifty. Pity about the leak, but I look forward to the improved photos later. 

[Andrea B.]

I think the film shows the contrasts between the lighter distant mountains and the UV-dark foreground areas maybe a little better than we get with digital? It's very striking in these photos.

[OlDoinyo]

15 hours ago, Andrea B. said:

I think the film shows the contrasts between the lighter distant mountains and the UV-dark foreground areas maybe a little better than we get with digital? It's very striking in these photos.

I don't know about that--both my digital and film images are tonally tweaked on an individual basis, and it may just be a function of what  curve I chose rather than anything intrinsic to the equipment. I have not done any controlled comparisons. It is possible that differing sensitivity curves of film versus a digital sensor might play some secondary role in how the pictures work up. The atmosphere was somewhat smoky due to wildfire season when these were taken, so that may also be a factor.

[dabateman]

Based on some of the Kodak film curves I have seen. I think film is much more UV sensitive than our digital sensors. 

So that should help the film. Also not being sensitive to IR also really helps film.

[OlDoinyo]

18 hours ago, dabateman said:

Based on some of the Kodak film curves I have seen. I think film is much more UV sensitive than our digital sensors. 

So that should help the film. Also not being sensitive to IR also really helps film.

In UVA, non-overcoated film is certainly more sensitive than any digital camera I have personally owned. Essentially all of my digital UV shots are extended exposures on tripods, whereas a good many of my film UV shots were handheld. With some of the newer cameras where you can crank the sensor gain up to insane levels like ISO 204800 and beyond, that may no longer be true; I don't know.

 

In UVB and below, I suspect film has more of an advantage, as consumer digital sensors often roll off steeply below 340 nm unless they are specially modified. As I possess neither the filters nor the lenses to test this, I can only speculate for now. That said, I am unsure how much to trust manufacturers' UV sensitivity curves--are they real data, or just careless extrapolations from the visible? Most customers, after all, cared only about the part of the graph between 400 and 700.

[OlDoinyo]

Postscript: The Second Roll

 

I sent the camera out to be examined--and the repairman swore he could find no evidence of a light leak! But I still have no credible alternative theory of what might be going on. I shot a roll of slower film (not UV) and found no evidence of the leak, as before. So I loaded up another roll of Delta 3200 and had at it once more. This time, the leak (if that be what it is) only showed up in just over half the frames (the first six frames did not show it, and only one of the following frames showed it really strongly.) Perhaps this is progress of a sort. The variability in how intense the leak appears seems out of proportion to the variability in exposure, suggesting that the recorded leakage might be primarily occurring between, rather than during, exposures. The leak probably does not consist of UV, and the film has no sensitivity beyond 690 nm, so it cannot be IR either. Most likely it consists of orange and red light. The answer may be keeping the camera partially or totally covered between exposures, and I shall try that. Meanwhile, a couple of frames worth posting were salvaged from the unaffected exposures:

 

 

The first two frames show UV's ability to handle shooting into bright light sources well without much halation or flaring. The UV reflectivity of snow and white water are well exhibited in the second two frames. All of the frames suffer a bit from smear bands (vertical patches of blurring in places.) This is more often seen in slit-scan cameras due to mismatched camera rotation and film feed speeds. But the Horizon is a swing-lens camera, whose film does not move during exposure. Perhaps these are just unfortunate examples of camera shake.

[dabateman]

Could it be some type of glare or reflection from your filter holder?

The edges seem quite sharp on the sides of the over exposure area. 

 

[OlDoinyo]

A Vancouver photographer, Goran Basaric, provided this example in a Photrio post, from one of his cameras. I must say that it looks like the same thing. In his case, it was a leak, and was remedied by having the turret light seals replaced. The explanation at least makes some logical sense to me. 

 

Any problem originating anywhere in the rotating lens carriage, if it produces a fault, would be expected to smear the feature in a horizontal direction. That is not observed here.