OT: Digitising collections of microfiche - Re: Looking for opinions...

[Fred Cisin]

On Wed, 28 Mar 2018, Bob Rosenbloom via cctalk wrote:
> This thread reminded me of a DYI scanner I had read about. Found it with 
> google:
> http://retrocmp.com/projects/scanning-micro-fiches/235-the-homebuild-automatic-micro-fiche-scanner

I like his enthusiasm and the fact that he actually got it together and 
DID it.

Hmmm.  If you gut the reader, and photograph the illuminated portion of 
the fiche, rather than project and photograph the screen, image quality 
will be a lot better.  Ambient light will cease to be a significant issue.

I don't think that ANY of them are color, so with bitonal, noise should 
not be that much of a problem, unless that is being compounded by the long 
exposure?  Steadiness for a 3 second exposure seems like more of an issue. 
Although I generally agree with stopping down the lens, the midpoint of 
the aperture range is not necessarily the sharpest (often over-simplified 
to "2 steps down from wide open"), and the difference between f10 and f5.6 
is not likely to be as significant as other issues.  THAT would drop the 
exposure to less than a second. I don't know the sensitivity of the CCDs 
to long exposure, an exposure as long as 3 seconds gets into the "reciprocity 
failure" range.

If it can be set up to let the PC monitor the image, rather than create a 
JPG and transfer that, then it should be possible to make some software on 
the PC to find the edges of the frames and move the positioner to center 
them.

I'm not familiar with an "Isel" CNC positioner.  We're not talking much 
mass to need something heavy enough that it disturbs everybody in the 
house, and requires a 5 second settle time.  I'm thinking more in terms of 
floppy head positioners.

BUT, he built one and made it work.  I haven't done that.