Jan. 19th, 2016

By the time I retired a couple of years ago I had accumulated over 50 years of photography some 6.000 slides and an unknown number of negative films. All these treasures were boxed up in our basement and we hadn't looked at it for years. And then I dicovered Panoramio, a photo sharing platform, which allows to geotag photos and place them on Google Earth. I decided, that this had to be my thing and I started scanning and uploading my slides. Scanning was done with an available Canon 3200F, slide by slide by slide. Looking back the quality was lousy and it took ages, but what the heck, I had the time ...

After I finished the slides I started with the negatives. Again, the quality was lousy and it took ages again and in the back of my head nested the nagging idea, that there must be a better, faster and foremost sharper way to scan my negatives. However, I was not even sure, if the mediocre results of my scans (meanwhile by a Canon 5600F) were due to the quality of the negatives or the scanner just couldn't do better. That question was settled last year, when I tested another cheap scanner, which was not really a scanner but simply took shots of the negatives. The results were much sharper than those of my Canon scanner, however, the scans/shots were extremely noisy. So I had traded one problem for another, but at least I knew, that there is still potential.

Of course I did review the market various times in order to find out, which scanner would solve my problem. The result was discouraging: poor quality or far out of budget and all rather slow. But that cheap and noisy scanner I tested last year did it differently and having that in the back of my head I made up my mind, to find out, if I could build a device like that with little money. And, yes, I know, a lot of people have tried and failed already. On the other hand ... it's fun:-). Let's wait an see what is coming out of this.
The Demonstrator, a fine piece of german engineering!

First I wanted to find out, if this is feasible at all and decided to set up some kind of a demonstrator based on my Tablet PC as a light source, the negative holder from my old Canon 3200F and my EOS 70D with the 15-85 mm zoom lens and of course a tripod.


This was the first useable shot (AGFA), after I found out, that the tablet needs to be out of the focal plane lest every pixel of the tablet backlight is visible on the photo. And since I didn't have extention tubes, the reproduction filled only a small part of the frame and therefore the final pic was only about 1700x1100 px after crop.

This is the result after processing in Lightroom. It is not perfect but given the circumstances quite encouraging. Compare to the pic in my photostream. Next step will be my B-Model with LED Video Light with diffuser and extention tubes to fill the frame.

To be continued ...

Jan, 21st 2016

From the demonstrator I learned, that in principle it can be done, however, as always, the devil is in the details. While working with the B-Model I encountered a whole bunch of nasty problems. The B-Model consists of a simple cardboard box with a cutout for the negative holder, a battery powered LED video light and a diffuser (for the time being a sheet of white paper until the ordered frosted glass will arrive), EOS 70D with 85mm, f/1,8 and extension tubes on a tripod. See photos.




  • Paper does not work well as a diffuser because it’s not homogenous and leaves a cloudy image, hope I can fix that with the frosted glass.
  • Given the extremely shallow DoF the film must be very nicely flat. Especially my old Kodacolor II films are heavily bent along the longitudinal axis, which makes it even difficult to feed them into the holder. Therefore some areas of the image are always out of focus. To fix that I may have to build another negative holder. The problem does not exist with my old AGFA films.
  • Focusing is a mess. To fix that I may have to invest in some more expensive gear, which I wanted to avoid. Moreover Camera and lightbox must be fixed to each other.
  • The scans of the Kodacolor II come off extremely blue. The backlight has a color temperature of 5.400 K, nevertheless I have to adjust the camera WB to 2.500 K to get acceptable colors. Again, this problem does not exist with the AGFA negatives.
  • The workflow will be rather complicated.
  • But the worst one is repeatability or rather non-repeatability. Each and every photo needs it’s own set of corrections, which is extremely time consuming.
Before or if I continue I have to give it some thought regarding the options.

By now my investment in this project israther limited:

• LED Video Light 14€
• Frosted Glass diffuser 5.90€

I don’t count the extension tubes; they have been on my wish list anyway for quite a while.


To be continued ....

January 26th, 2016

It can be done …. but not by me ;-)
Being an engineer you have to know, when to call it quits and limit your exposure. My little project seems to be a problem with an unknown number of unknowns. But what did I learn? First and foremost, that my negatives are much better and sharper, that I thought. I have to thank Norbert here, who volunteered to provide a sample scan of one of my negatives on a Plustek 7600 scanner. And secondly, that my B-Model provided excellent results regarding noise, much better than all my previous scanners and level with Norbert’s Plustek. Other findings were:
  • Focussing is a mess, mostly due to the extremely shallow DoF.
  • Frosted glass performed much better as a diffuser then my plain sheet of white paper
  • Illumination of the diffuser from the back is extremely sensitive to geometry inaccuracies (whereby geometry is a general problem also with respect to axes etc.) and distance of the light source.
  • Camera color temperature setting may have to be established for every film. Color management in general seems to be a big problem.
  • The box needs to be plain white, grey or black inside. Any colors inside the box compromise the result.
To make this a success, I have to dig much deeper and probably invest much more money with an open result. Too risky. If I really want better scan results I’m afraid I have to invest in new hardware. Nevertheless, arriving at this conclusion was fun in a way and provided some new and valuable insights.