Comparison - Epson Flatbed Scans With Glass vs. Without Glass

Flatbed scanner users have always wondered how much detriment is

caused by the piece of glass found in the optical path of flatbed

scanners when scanning their film. There have been many statements

made about this but I had never seen any postings of actual

comparison images. I had been contemplating experimenting to see

how much of a gain could be obtained by removing the glass. Other

projects took precedence and I also wasn't real excited about

dismantling my scanner, so the project had been "on the

backburner." A friend, Ian S. (name hidden to protect the

innocent!), recently emailed me and said he had removed the glass on

his Epson flatbed scanner. Ian does incredible panoramic work and

found glassless scanning added benefit for his particular workflow

(including sharpening, etc.) and encouraged me to give it a try. I

have no doubt it is a good solution for him. His panoramic scans

are gorgeous. For my workflow, I did not feel the benefit was worth

the effort. I can see a very slight increase in resolution (e.g.

the fins in the color image), but nothing like I had hoped to

achieve.

<p>

I thought others might like to see the results, so links are

provided below. Both images were shot using a 35 mm Nikon F3 HP on

a tripod with either a Nikon 24 mm 2.8 AFD or a 50 mm 1.4 AIS. The

color image was shot on Kodak E100S and the b/w image on Tmax 100

developed in Xtol. The film images are tack sharp when printed in

the "fume room." I used my Epson 3200 with the original Epson

holders for these scans (my 4870 isn't out of warranty yet!). FWIW,

Ian sent me a comparison scan made on his modified 4870 and

the "level" of benefit appeared about the same to us both. I did

not manipulate these scans (including resampling) before uploading.

EpsonScan was used without any adjustments to the software, so that

is why the scans look a bit flat and show the scratches. The custom

cut "plate" that I created to replace the glass bed retained all the

characteristics to provide correct scanner startup calibration.

Both of the "before" and "after" scans produced nearly identical

histograms/black points/middle points/white points.

<p>

<a href="http://www.photo.net/photo/3111674">Full

image used for the color scan comparison</a>

<p>

<a href="http://www.photo.net/photodb/photo?

photo_id=3111679&size=lg">Crop of color scans comparing with and

without glass</a>

<p>

<a href="http://www.photo.net/photo/3111581">Full

image used for the black and white negative comparison</a>

<p>

<a href="http://www.photo.net/photodb/photo?

photo_id=3111670&size=lg">Crop of black and white scans comparing

with and without glass</a>

<p> Doug<p>

<a

href="http://home.earthlink.net/~dougfisher/holder/mainintro.html">Do

ugs MF Film Holder for batch scanning of 120/220 medium format film

with flatbeds</a>

How do you keep the slide/negative from falling into the inner workings of the scanner?

Doug,

 

Questions and observations:

 

1 - How did he suspend or otherwise hold his 35mm

slides and negatives flat without glass?

 

2 - Microtek makes flat-bed scanners with glassless

scanning for film. The i900 is one example.

Unfortunately, I've seen no test of these

against a quality flatbed. Your test of the

Epson with and without glass, is a very nice

conceptual test!

 

3 - I've seen it recommended to USE glass even with film scanners, especially for medium format due to problems keeping the film flat. In short the problems of film curling and sagging, etc) can offset the disadvanges of glass!

I have never thought the culprit for low performance on the Epson scanners was the glass. Creo scanners and other hig end production scanners give great results and they have a glass too.

 

For me, the culprits are:

- no focusing

- low end optics

- low end sensor

- low end, very simple mecanism

 

In a word, you get what you pay for, as usual.

Doug,

 

Finally, someone actually did it!

 

But I would not stop there. One of the reason of getting rid of the flatbed, beside eliminating the glass, it to be able to darken the inside of the scanner so that contrast can be increased. I did not see that in your report.

 

Also, was the height of the film holder adjusted to the focus point of this particular scanner? Some Epson 3200 scanners, because of mfg tolerance, may have their focus point just below the glass bed! In other word, the optimum focus point of the Epson 3200 and possibly the Epson 4870 may differ from one scanner to the other.

The old Epson 2450 we have here has good focus with the stock Epson holders. Placing a negative directly on the glass didnt increase measured resolution at all.

Jay and Fast Primes - I created a "plate" out of the same type of material I use for my holders that I installed in place of the standard glass which is designed to allow all of the Epson and third party holders to be used in normal fashion (but of course without any glass in the actual film scanning location/area).

<p>

Jack - Yes, "flocking" the inside could provide benefits. Anything you can do to minimize stray light and internal reflections is a good thing. Especially with the 2450 and 3200 scanners since they have the relatively large light source (as opposed to the newer moving light tube system). In regard to the film suspension height, my plate provided the exact same suspension height for the holders, so it was an "apples to apples" comparison. My understanding is that Ian accomplished the same thing with his 4870. That is why I made note of the fact that we both saw what appeared to be about an equal amount of raw optical performance change. This helped me feel that my results were probably pretty representative for most flatbeds using similar technology.

<p> Doug<p>

<a href="http://home.earthlink.net/~dougfisher/holder/mainintro.html">Dougs

MF Film Holder for batch scanning of 120/220 medium format film with flatbeds</a>