Panoramic Slit-Scan Camera Idea

[holmburgers]

Hey everyone,

So I've had this idea tumbling around in my noggin for some time now, but I'd like to actually see it in action some day. I could certainly use your ideas and input for the ideation stage.

In a nutshell, imagine a long, narrow box that would hold an entire roll of film (let's assume 35mm for now). The roll would be laid out flat in this dark box. The lens would then be in a track that runs along the length of the box, with a slit aperture between it and the film... much like a Widelux, or similar designs. The picture would be exposed by uncapping the lens and sliding the lens across the length of the film; the slit allowing a thin sliver of light to scan the film. Shutter speed could be altered by different sized slits or by the speed of the lens transport.

Your resulting image would be a continuous image the length of your film.. the ultimate panoramic!

This is the basic idea, and the only conditions that a prototype would need to meet.

The uses of a camera like this are yet to be determined. But in a horizontal position, set your 3 best buds on a love seat and scan away; you've now got a group shot that extends the length of a roll of film, each person captured head on. In a vertical position, take your favorite child (or least favorite, depending on their enjoyment of having dad or mom take their picture) and scan from head to toe, assuming you have a short kid...

It's really hard for me to imagine what the resulting pictures would look like, and that's a huge part of the impetus!

As for making this... I've got some ideas, but nothing concrete; just a starting place. First, you'd need a lens mount which could be taken from a junk camera or an extension tube or even a rear cap. The focal flange distance would of course have to be determined and your box & track based off of that. The hardest part in my mind would be an accordion, curtain or venetian-blind like material that would allow the lens to move from end to end and keep the light out.

Down the road, one could of course get quite fancy with motorized transport and the like... curvable boxes, and so forth.

What about a prototype box? Wood, mat-board, some found object?

 

[domaz]

You would need a fast moving motor to do this with modern films. Ever seen how fast a Speed Graphic focal plane shutter goes at 1/125 for instace? Now imagine your lens assembly moving on a rail that fast. That's going to require a serious electric motor that can accerlerate quickly and keep up the speed evenly through the entire track. A stepper motor could do this but it would need to be a big one.

As far as the venetian blind- why not just use a lens with a built-in shutter? So basically you mount the shuttered lens on a rail that runs over the length of the box. When the exposure starts the shutter goes on "B", there is a permanent slit in the back of the lens assembly that keeps the coverage of the lens confined to a narrow slit- just like a focal plane shutter does. The lens runs over the entire length of the film at an even speed and wa-la panoramic photo.

 

[holmburgers]

As for the speed of the lens, I don't think this should be a problem. I can always use ND, a smaller slit or even utilize the aperture of the lens. Or Velvia 50.

The built in shutter wouldn't change the fact that this track needs to be sealed on either side of the lens as it moves across the camera.

 

[MattKing]

holmburgers:

It is a really interesting idea, but I don't think it will work.

I think that unless you had something like a step motor synched to a shutter from a movie camera you would just end up with a long blur.

Unless you have an infinitely narrow image at the film plane, each image would tend to overlay the previous - essentially a continuous multiple exposure.

To do this you would need some mechanism that stops the lens mount, opens a shutter, exposes a defined slit on the film, closes a shutter, and then advances the entire mechanism a sufficient distance to ensure that when the cycle repeats an entirely new section of film is exposed.

Sort of like a movie camera where the film stays stationary and the lens and shutter move.

 

[Steve Smith]

This is almost (but not quite) the way photo finish cameras work for horse and car racing. The film passes the lens (with a narrow slit) at the same speed as the horses or cars. The background is the same narrow view along the whole length of the film with normal images of the competitors.

http://en.wikipedia.org/wiki/Photo_finish


Steve.

 

[holmburgers]

Thanks Matt,

I'm pretty sure that it will work as this is the same principle that the Widelux, the Cirkut and all other slit-scan cameras utilize.

The slit should be thin, but not infinitely narrow as you say.

Imagine an vase of flowers, for instance. A lens directly in front of it will obviously project an image onto the film plane, directly behind it... now move the lens to the side one inch and that vase will still be in the same place on the film plane. It's just like utilizing shift on a view camera, the side of the lens is now imaging the vase as opposed to the center. But the image of the vase has not moved on the film plane, it's just that it has changed in relation to the lens. The slit should take advantage of the sharpest part of the lens, but a slit isn't essentially necessary.

I'm not saying I couldn't be wrong, but unless you have a good idea of how slit-scan works, I'm fairly confident that my principle is sound. The movie camera analogy is different in that the film is moving and yet the image remains the same, thus nothing but blur.

Steve; thanks for that note!

edit: I should add that the moving film will result in blur except if a subject moves in sync with the moving film (like the race-track photo-finish cameras.... http://people.rit.edu/andpph/text-photofinish-race.html)

 

[Dan Fromm]

Its wonderful that you're imaginative. Its not so wonderful that you haven't met current art.

See http://www.surplusshed.com/pages/item/i1433.html and http://www.surplusshed.com/pages/item/i1475.html . Per the USAF data sheets I have lying around, the Zeiss lens is a Planar and the Fairchild isn't. So much for Roger's puffery. Either version of the K-18 will do what you want and there are kits or instructions for converting K-18s into panoramic cameras for terrestrial use. If it matters to you, the Fairchild lens is probably better than the Zeiss.

I know, I know, buying one will deprive you of the joy of making your own. Would you rather tinker or take pictures?

I didn't look to see whether Roger is still selling the lenses without cameras. If he is, don't yield to the temptation. One of my neighbors, now, alas, dead, had a drawer of the lenses. Putting them in proper shutters and adapting to a more normal camera isn't easy, costs much more than just getting the right lens for the camera.

 

[holmburgers]

Dan, alas, that design is quite different than what I have in mind.

To achieve what I described in the first post, those cameras would have to move in relation to the film trasnport, along their own track; not the lens. But thanks for sharing the links because certainly it's in the ballpark, but fundamentally different.

Building a film transport would be much more difficult for me than a sliding lens on a track.

The idea is to have a long bar with film in it and a lens that tracks along this. The film will remain stationary, the lens will move.

 

[MattKing]

My head hurts .

Steve:

For those photo-finish cameras, isn't it critical that the speed of the film be close to the speed of the horses? And isn't it the case that stationary objects are rendered as blurs?

holmburgers:

Doesn't "shift" move the image on the film plane? Wouldn't your example vase move, unless you used swings instead?

I always thought that the Circut and other rotating cameras worked because of the curved film and/or rotating (rather than sliding) lens.

I don't know for sure - so I ask!

 

[Mainecoonmaniac]

Here's one from the Lomo folks for $100.
http://microsites.lomography.com/spinner-360/about

 

[Steff1]

Just bought one of the spinners. Great fun!

 

[holmburgers]

Matt, I understand and share your confusion; this is certainly an exercise for one's brain.

As good of an explanation as I can come up with is the fact that you can take a picture of a mirror, as though you were in front of it, but with no reflection of the camera in the mirror, by utilizing shift. But that's no explanation, is it... :/

I'm trying to reason this out and I'm hitting brick walls in my brain... I think that you're right that shifting will move the image, and thus you need a slit to isolate a region. But then I start thinking about the fact that this image is flipped and my brain starts to malfunction, I hit Cntrl+Alt+Delete and nothing happens.

But the lomography spinner is another example of what I think proves my theory. And yet again, I can't quite put it into words.

However this is a camera-building thread, so let's stick to that.

I'm glad we're sorting it out... we need a slit-scan expert STAT.

Must..... reason......it......out......

 

[richard ide]

As per your first post; I am glad you are not using concrete.

 

[williamtheis]

actually the equivalent has already been done. there was an artist at Laguna Beach Festival of the Arts who runs 70mm or 220 through a Hasselblad continuosly with a slit lens photographing a moving subject (cars going by, marching bands, etc)....

there is also a scheme to rotate the camera on a turntable for stationary subjects.

BTW: this is what the Betterlight scan back does in panorama. it parks the sensor in the middle of the Field of View and then rotates the camera at the right rate to make the 360 degree panarama

 

[holmburgers]

There are many analogues, but do any of them have stationary film with a moving lens?

MattKing has definitely placed doubts in my mind as to whether it would work like I had envisioned. I've got to devise some kind of test...

I'd love to see this Laguna Beach artist though, that sounds great!

And luckily I'm not a materials engineer.... I have very few concrete ideas........ cue: *laughter and applause*

 

[MattKing]

There are many analogues, but do any of them have stationary film with a moving lens?

MattKing has definitely placed doubts in my mind as to whether it would work like I had envisioned. I've got to devise some kind of test...

I'd love to see this Laguna Beach artist though, that sounds great!

And luckily I'm not a materials engineer.... I have very few concrete ideas........ cue: *laughter and applause*

How about testing using a shoe box and about a foot of film?

 

[holmburgers]

Good idea. I was also thinking about doing something with ground glass or tracing paper and drawing what I see; crude but would at least help me visualize it.

I'll see about doing something over the weekend.

Cheers!

 

[Dan Fromm]

Hey everyone,

In a nutshell, imagine a long, narrow box that would hold an entire roll of film (let's assume 35mm for now). The roll would be laid out flat in this dark box. The lens would then be in a track that runs along the length of the box, with a slit aperture between it and the film... much like a Widelux, or similar designs. The picture would be exposed by uncapping the lens and sliding the lens across the length of the film; the slit allowing a thin sliver of light to scan the film. Shutter speed could be altered by different sized slits or by the speed of the lens transport.

So you want the lens to translate during the exposure, not swing as in, e.g., a Kodak Panorama camera or a Widelux or in the aerial cameras I pointed out to you.

I don't get the point of it, but that's just more proof that I'm not imaginative.

Why don't you simulate the effect by setting a camera down first at one end of a yardstick on top of a tripod, then at the other, and taking an exposure from each position with the camera pointing straight ahead. Nice simple feasible proof-of-concept prototype. If the two shots or narrow sections of them give you the effect -- with the middle missing -- you want, wonderful. If not, reconsider what you're trying to accomplish.

When you think of the mechanics, an electrically-driven translating lensboard shouldn't be too hard to make. But, as I've already said, I can't grasp why you want to do it.

 

[holmburgers]

I want to make it because I like to keep on the tradition of photographic invention. What if George Lawrence had never decided to make a camera that could be flown by a kite? We wouldn't have this picture...

http://upload.wikimedia.org/wikipedia/commons/b/b3/San_Francisco_in_ruin_edit2.jpg

What if F.E. Ives never invented the Kromskop? We might not have color film!

What if Edwin Land never invented a film that could develop into a print instantly! We might not have digital cameras!

.... (i'll stop my work immediately)

Ok, sorry, got carried away.

But seriously, why not make something if you've got an idea and you think the result could be worth the effort to achieve it?

I have learned something more about this in the meantime. I took out my view camera and made a slit by masking a filter with some painter's tape. Just two pieces that left about a 5mm gap for the lens to peer through. The funny thing is that in either case (in front of lens or behind lens), a full image appeared on the ground glass.

Looking back it's obvious that a slit that close to the lens would act just like a waterhouse stop, but one important design feature is that the slit has to be as close to the image plane as possible to achieve a well defined beam of light.

And since the slit has to move with the lens, the lens mount will have to protrude back and hold the aperture slit recessed in the cavity.

So this discovery delay's any test with my view camera as I was going to observe how the slit moved across the ground glass.

 

[holmburgers]

A Test

Last report for now, but I found a mailing tube, cut it to 140mm, secured it to the back of my 135mm Raptar's lens board and on the end of the tube nearest the film plane was my slit.

Matt was right in that it will basically double-expose the image and I was totally wrong about not needing a thin slit. The wider the slit, the more area that is being exposed in a single moment. Imagine a diagonal line in your subject. If it's a wide slit that encompasses the diagonal it'll end up as a neutral gray because the whole line will sweep across the film. But if it's a thin slit, a smaller section of the diagonal line will be seen by the film at a given moment and thus it'll register as a line (but thicker depending on the width of the slit).

It's a very weird thing to observe; when you shift the lens left or right the slit will move at one rate and the image will move at another. So I think it'll take your subject and stretch it very interestingly. Perhaps it could "smushed" back in post production or in analog terms, by an anamorphic lens of some sort.

And indeed, when you swing the lens, the subject remains motionless on the ground glass and the slit scans across it, like shining a flashlight in the dark. This is what the Widelux, Horizon, et al., do; pivot around the lens' nodal point.

So with a thin slit, the image will be discernible as your subject, but with a yet to be envisioned blurry appearance.

Fascinating stuff!

 

[MattKing]

Thanks for the test.

My head hurts less now .

It was/is a great idea - very fascinating.

 

[Steve Smith]

Steve:

For those photo-finish cameras, isn't it critical that the speed of the film be close to the speed of the horses? And isn't it the case that stationary objects are rendered as blurs?

Yes but if you keep the film still and move the lens instead then the opposite may be the result.


Steve.

 

[2F/2F]

There are many problems with the idea, but here is just one of them. Take a picture of an object at a distance that would likely be used for panoramic shots. Then, move the distance of a length of film (about four feet) to the right or left and take another picture, with the lens parallel to where it was for the first shot. Compare the two shots, noting how extremely similar they are in the objects that are pictured within the frame. In other words, this will not widen your angle of view any noticeable amount unless you are quite close to your subject. To make a moving-lens camera work to significantly widen the angle of view, you must curve the film plane and have the movement of the lens follow its path.

 

[holmburgers]

2F/2F, you're quite right and I guess saying "the ultimate panoramic" was a bit misleading on my part. All the subjects I'm imagining are as close as 5 feet or so.

As for curving the film plane; the ultimate version of this theoretical camera would/should have that ability. Y'know, the top o the line version!

What other problems do you see?