Flat Bed Scanner + Enlarger = Film Scanner?

I have a 6x6cm enlarger and an HP FB scanner. I am thinking of adding a
switch to the scanner to turn off the light tube and use the enlarger to
project the image on to the scanner. In theory, it should work, right?
Before I risk trashing the scanner, I thought I should ask if anyone has
done this before and what's the quality of the scan.
I know there are MF film scanners but they are expensive and I won't use
them enough to justify the expense.
GH

 

I have a friend who did something very similar. He had an 8 x 10 view
camera. Used the camera to project image on flatbed, scanned, and it
worked great. I have seen the impressive images.

This may only work on some scanners, but it will indeed work on some.

 

"> This may only work on some scanners, but it will indeed work on some.


I love "Yogi-isms" ;-)

 

I tried exactly this with an old scanner and a 35mm
enlarger recently but it didn't work. It could be made
to work if you projected the image onto a fresnel
lens on the scanner glass but it would be difficult
to remove the hangover of fresnel rings from the
image.

The idea that flatbeds cannot give good results with
film has been overtaken by technology. Modern
flatbeds such as the Epson 3200 are not far behind
dedicated film scanners for 1/10th the price. A new
Epson is imminent. These scanners have masks for
35mm, MF and 4x5 film.

Graham

 

Me too.

Yogi was being interviewed by Bryant Gumble.

Bryant asked Yogi to try some word association.

Meaning Bryant says something and Yogi says the first thing that comes to
his head.

Bryant said "Mickey Mantle."

Yogi replied "What about him?"

:-DDDDD

Best, Dan

 

Two points to consider.
1. The flatbed is (approximately) focused on the glass surface plane. You
would probably need a means to focus the enlarger image on the CCD (remove
scanner lens).
2. The light source of the scanner is concentrated on a narrow line across
the bed, in order to keep exposure time low (not all scanners allow the
scan software to change exposure time). That light slit moves with the
scanner head. A stationary lightsource will have significant light
fall-off away from the center.

Bart

 

You'll get severe vignetting with a standard scanner. The problem is
getting the light that leaves the enlarger lens to enter the scanner
lens anywhere other than near the centre of the image.

Draw the light path for an an enlarger. Light from the enlarging lens
reaches the paper at an angle to the vertical, except at the very centre
of the image. That's fine for enlarging, because the paper doesn't care
what angle the light reaches it from (within reason).

Now look at the light path for a scanner. Although the path is folded
several times by mirrors to keep the scan head relatively small, the
optical path is equivalent to having a lens something like 10 inches
below the glass surface, looking up at the glass through a narrow slit.
At the edges of the page, the light has to leave the paper at an angle
in order to enter the scanner lens. This isn't a problem when scanning,
since the light illuminates the paper and the paper scatters the light
in all directions.

If you put a scanner under an enlarger, you need some way to bend the
light from the enlarger lens, which is spreading out, back towards the
scanner lens. As someone else suggested, a Fresnel lens of just the
right focal length would work, but if the Fresnel lens is too close to
the focal plane you'll see the rings. A large plain lens would work
too, but the diameter needs to be larger than the area of the scanner
platen you are using, and it needs to be fairly strong.

There is one type of scanner that can be made to work. The Canon LiDE
scanners don't use a small CCD plus a lens to capture the image, they
use a large CMOS sensor that's the full width of the page, plus a long
array of tiny lenslets to project the paper image onto the sensor. If
you remove the lenslet array completely, you can project the enlarger
image *directly* onto the CMOS sensor. Since there's only one lens
involved (in the enlarger), you don't need to worry about redirecting
light from one lens into the other.

However, there's now another problem: the Canon sensor is B&W, and
colour sensing is done by using red, green, and blue LEDs in sequence to
illuminate the paper. If you're going to use an enlarger as the image
source, you need to disable the scanner light source, but rapidly switch
between red, green, and blue light in the enlarger in sync with the
colour that the scanner believes it has turned on. A conventional
enlarger won't do this; an enlarger head with R, G, and B strobe tubes
might be able to, or you could use a rapidly-rotating colour wheel
synced to the scan head. Or you could just make 3 greyscale scans with
a colour wheel selecting red, green, and blue filters one for each pass.

In short, an interesting experiment, but a lot of work for what you'd
get.

Dave

 

Thanks for the explanation. I didn't know a scanners uses a lens and a
small CCD array. I just checked my scanner and it does have a lens. I
always thought all scaners use contact image sensors (close to your
Canon example), but with 3 rows to get RGB.
GH

 

There seem to be several different arrangements. Most flatbeds use a
small CCD trilinear array (3 rows of sensors, with red, green, and blue
stripe filters in front) and white light illumination. The Canon CIS
scanners are full-width sensor (probably CMOS) with a very interesting
lenslet array and colour LED illumination for colour separation. The
lenslet array appears to be GRIN lenses, which produce an *upright*
image not an inverted one. (That's necessary for all the individual
images to overlap to form a single image).

Some film scanners seem to have a trilinear array and white light, while
Canon seems to use a single-row small sensor and multiplex between 4
light sources (RGB and infrared). In most modes the LS2000 (at least)
moves the light source and lens/CCD while leaving the film stationary,
but there's a preview mode that moves the film past a presumably
stationary scanning head.

Dave

 

Yes, you'd need a scanner with a CIS (contact image sensor)
array, rather than the more usual CCD+mirrors+lens array.

However, I know of no CIS array that has sufficient resolution
to come anywhere close to film's capability -- the best I know
of are around 1200 dpi.

CIS arrays are usually built with an integrated R/G/B LED
lighting system.


rafe b.
http://www.terrapinphoto.com

 

Virtual Drum Scanner

This does exist comercially and its expensive. i'm assuming its because of its limited production.

The Hasselblad Flex X5:


however, if you're truly interested in making your own, take a look at this:


a fellow in japan made a 100 megapixel camera out of a scanner. if you replaced the lens with an enlarger, it should work the same way.