Five or six weeks ago there was a thread on slide scanning and copying and I mentioned that I was constructing a slide copier. Well, here it is.
I have 5000+ slides that I want to digitize and using my flat bed scanner was not going to cut it as it takes about 16 minutes to scan 4 slides at 4800 dpi. Not counting inserting and removing the slides it would take approximately 350 hours just for the scanning (!). Slide scanning services would cost about $1700 (4000 dpi)for JPEG; $800 more for TIFF files.
Since I already owned a camera and a macro lens I decided to do it myself. My Nikon D90 is able to do 3000 dpi (4288 pixels/36mm*25.4mm/in). How does that compare to 4800 dpi? Since there may be some people who are interested in just how good, or not good, I also decided to do a comparison of camera vs scanning quality and this will be the subject of another thread.
The device looks cumbersome but the whole point of it is to lock in the relationship and alignment of the camera with the slide holder. If you have machinist’s skills you could simply construct a tube with threads at one end and a slide holder on the other. Or start with a screw-on lens hood? The commercial models, of which there are only one or two, either do not provide a sharp image (based on reviews) or need a different macro lens to work properly.
Photo 1 shows the device with camera and lens and Photo 2 is a close up of the camera mount.
Photo 1
Photo 2
The camera mount component looks somewhat complex but I did not want any free play or adjustment capability at this end. The slide holder is a combination of a portion of an old slide duplicator unit and an old bellows unit [originally for an Exakta camera – anyone out there ever own one?]. The bellows has been discarded and only the rack mechanism has been retained. The slide holder component has all the adjustments to align it with the camera lens. Adjustments can be made in 3 planes and 3 rotations and everything can be locked down once collimation has been achieved. Collimation was done by inserting a mirror in place where the slide goes and making adjustments until I could see the mirror reflection centered. I thought I would be able to see my eye but it seems I am seeing some component located inside the camera (AF mechanism?).
To copy a slide insert it into the holder, face the device towards a sunlit window (not a valid option for UK residents) and take a photo. For consistent exposure I decided to use a flash. I have some old flashes, and because they are not recommended to be connected to a digital camera, I bought an optical slave so it could be fired remotely. I use the camera’s built in flash to set off the main flash. I experimented with various flash-to-slide distances until I got a good exposure. Four to six feet is a good distance for a flash unit with a Guide Number of 60 (ISO 100).
Camera setting parameters:
Utilize shutter delay to minimize vibration.
Use a remote release to also minimize vibration.
Set AF to multi point, not single point. If single point is used and that part of the slide has no contrast you will lose focus. You can also try manual focus but not all slides may be in the same exact focus plane. DOF is razor thin.
I used manual setting but you can also try shutter preferred. A starting point for exposure of ISO 200, 1/200 second and f/11 worked fairly well.
If using an optical slave the on-camera flash can be set to its minimum output (1/128 power on a Nikon D90) to save the battery from running down prematurely.
If using the on-camera flash a light hood should be used to prevent lighting up the front of the slide.
Slide copying tip:
Brush each slide to remove dust. I found that all slides had dust even though they had been in their carousel tray and box their entire life.
Preliminary results are very good. 3000 dpi provides an image suitable for monitor viewing and small prints. The scanner appears to provide better resolution (4800 dpi claimed) but is only noticeable at 400% or better. I will provide a more detailed comparison (with photos) in another thread in a couple of weeks.