Slit camera project
Constructed 11th July 2008
Last modified 30th March 2010 <![endif]>
This site under construction.
more details to come
50mm lens exposed at f2 film length 280mm April 2008
28mm lens f2.8 film length 180mm 04/05/2008
28mm f2.8 film length 180mm Panning motor turned off, camera stationary 07/05/2008
28mm f2.8 film length 180mm 12/05/2008
Videos of camera setup:
Mpeg 2 video link approx 16mb (right click and “save target as” if streaming in browser doesn’t work)
Firstly apologies for the crude design of this site. It is work in progress and as developments take place notes are written here. This is my notebook.
By way of introduction, I am a Program coordinator of a photography program at Northern Melbourne Institute of TAFE (NMIT) and a practicing photographer since the late 70’s. (Gosh, don’t calculate my age)
I have been intending to experiment with slitscan cameras for years now. On a visit to RIT in 1997 I was inspired by work done by Andrew Davidhazy http://people.rit.edu/andpph/ and saw one of his modified cameras in his office. Building on my long term experience in radio control aeromodelling, I took an interest in fitting and turning and did a course at NMIT, Heidelberg campus, Melbourne, AU. I had purchased my own metal lathe, and the course filled in the gaps in my knowledge to get me going. I recently purchased a HM50 mill from Hare and Forbes..Now I had the tools to manufacture the parts I needed for this project.I also have to say that aeromodelling (particularly in electric flight ) over the last 15 years has also kept my interest in electronics alive. In recent years I have been introduced to micro controllers, first starting with a Picaxe system which I would highly recommend to anyone starting out with microcontrollers. Their help manual is so full of information on interfacing to sensors and motors. Then I began to use PIC microcontrollers, revisiting some assembler programming and now C programming. The application of a microcontroller to moving mechanical parts is what really makes this project exciting. Once the mechanicals are built there is much tweaking, fine tuning and expansion of application through the software in the micro.
I know that panoramas can be easily made by shooting separate images digitally and stiching them together in software. But I see this project as having possibilities for photographing crowds, seascapes as well as other scenes that would be hard to do with separate images over a number of seconds. The thing that appeals to me is that these images are one shot done in a single rotation of the camera. Also some of the unusual effects capableof this technique will be interesting. In the future I hope to take things learnt from this project and develop a rotating slitscan camera using a CCD or CMOS chip. If anyone has information on wiring a chip and handling the data I would appreciate a link to it.. I havent found anything on the net yet....
The camera is mount is adjustable so that the camera lens nodal point can be positioned precisely over the centre axis of rotation.
Slit size Originally .5mm but now 1.35mm to allow longer effective shutter speeds for low light photography.
By rotating a unmodified camera (clockwise when seen from above) , viewing the image with tracing paper at the focal plane, measuring the movement of the image coresponding to degrees rotated, it was found that a 50mm lens requires 320mm of film to produce a undistorted 360degree image. A 28mm lens needs 184mm of film. After initial tests such as the beach image above which was taken on 280mm of film, this shows why this beach image is a bit unsharp and distorted.
The second image above, taken with a 28mm lens, 180mm in film length produces a much more usable image. The proportions of this latter image are more suitable for printing. Not as long and thin.
Now that usable images are coming from the camera, I am scanning on a Epson 1650 scanner. There is some fine tuning to do here I suspect.
Camera rotation – 360 degrees in 2 seconds (This will be variable by program selection to give a range of shutter speeds and effects)
Camera: Olympus Zuiko OM1 ( My dream camera in the 80’s. I still used mine professionally until about 5 years ago)
Olympus film winder basically gutted of all internal electronic components except for motor and gearbox and convenient battery holder.
Shutter, frame counter and standard 35mm frame stop mechanism all disabled by removing vital parts after removing baseplate of camera.
Uses pic 16f628A microcontroller for motor control and timing. Uses hardware PWM (pulse width modulation) for film transport control. Programming using Boostc compiler which is inexpensive, good support available and I think, easy to use.
Slit width is now 1.35mm (For some of the images above I started with a .5mm slit - Worked fine but the effective shutter speed was too a bit too short for what I wanted).The slit is made by placing two razor blades on the mirror side of the focal plance shutter. I simply glued these in place with 5 minute epoxy resin. This gave may enough working time to adjust the slit size and eveness of slit width before the glue sets. I also decided to put the blades on this side so that their edges on the film plane ( a few millimeters back) will be a bit soft (out of focus) This is an attempt to minimize banding (uneven exposure) on the film should this occur.
50mm Zuiko f1.8 lens and 28mm f2.8 lens (need to remember to put a toothpic or glue in the auto aperture stop down lever on the back of lens. With the camera shutter disabled there is no auto stop down. Early experiments reminded me of this when images were consistently overexposed!!)
Exposure is made by removing the high tech lens cap from the camera just before exposure and putting it back afterwards. Because of the fine slit there isn’t much of a problem here because if you leave the lens cap off you only overexpose a tiny width of film.
Composing: Good thing this camera model has a mirror lockup. Framing and focusing of the image is made with the mirror down and at full aperture. Before taking the image, the mirror is locked up to allow light to reach the film through the slit and the aperture is manually closed down to the desired fstop.
The film cassette must be wound tight before loading so that there is minimum friction or binding as the film is wound through. This will cause excess banding through uneven transport speed past the slit.
Instructions for using the slitscan camera:
Program options 0-7 (In the end these programs will be used to select different effective shutter speeds by changing film transport speed and rotation speed, also calibration for different focal length lenses)
0 – not yet used
1 – 172mm film strip shoot for 28mm lens, 7 second self timer delay, 360 degree pan in 1.58 seconds
2 – not yet used
3 - not yet used
4 - not yet used
5 – not yet used
6 – not yet used
7 – not yet used
The camera on/off switch needs to be turned off then on to activate a change of program
The slitscan camera assembly is mounted on a standard ¼” tripod head.
The head should be leveled using the tripods spirit level.
Load your camera with film, making sure the cassette is wound tightish (not so tight that you scratch the film!). This should help provide a smooth transport of film. Reloaded cassettes may be on the tight side and that will be a problem as far as binding of the film in transport. Do the usual 3 winds with the lever to make sure you have moved past the exposed film leader. Lens cap should be on the camera lens at this point.
With the mirror down on the camera and the lens wide open, compose and focus the camera with the lens cap off. Replace the lens cap when done.
Take a incident or reflected lightmeter reading using a handheld meter or another camera. For the ISO of the film being loaded into this camera, read off the aperture corresponding to 1/80th of a second when using the 28mm lens. This is the aperture you need to manually set on this camera lens prior to exposure.
Effective shutter speed is calculated by dividing the slit width by the length of film moved in 1 second
For a 28mm lens the horizontal field of view is 75.4 degrees over 36mm so for 360 degrees:-
(360 divided by 75.4) X 36 is the length of film required to record 360 degrees. (1 full turn)
= 4.775 X 36
= 171.88mm length of film required for 360 image capture.
For the 28mm lens, 172mm film is transported in 1.58 seconds
172mm : 1.58 sec as x : 1 sec
Therefore 172X 1 divided by 1.58 = distance per second
Effective shutter speed = slit width divided by film transport per second
= 1.35 divided by 109
= .01239 (find reciprical)
= 1/80.71 th second. That's near enough to 1/80th second effective shutter speed!!!
Now for the actual exposure. Make sure the mirror is now locked up (otherwise light wont get to the film!) aperture is set. Turn slitscan power switch on. Momentarily press the fire button on the slitscan. In program 1 there is a 7 second delay. You can either duck below the camera or get into position in the shot!
After exposure, replace the lens cap!!!.
For further shots repeat the above procedure.
If you run out of film you will hear the motors straining. Please turn the power switch off as soon as possible at this point to prevent damage to the motors. (I haven’t put in a fail safe mechanism at this point)
When the film is finished. Turn the rewind release lever on the front of the OM1 and then rewind the film. Make sure the lens cap is on for this part. I have learned my lesson!!!
Lessons learnt so far!!
- Large apertures need to be used so far, banding is most evident if you stop down past f5.6 due to increased depth of focus of the slit on the focal plane. You could use neutral density filters on the lens to do this in bright conditions.
- At the end of the film REMEMBER to put the lens cap on before your rewind the film. I got some strange lines on the film - after some contemplation realized that in the excitement of the moment rewound the film while the slit is still exposed!! After many years of shooting digital and going back to not allowing ANY mistakes with film this is a big lesson. Take it slow and be methodical with this camera!! Also on a number of occasions forgot to lock mirror up during exposure, resulting in no image!!!
Some Slitscan links
Commercially made cameras
http://www.roundshot.ch/xml_1/internet/de/intro.cfm (Digital version)
Next project ideas
(requires late version of flash player very impressive)
Comments and contributions are most welcome. Email Alex Zattelman