OK, this might not be the typical DIY discussion, but believe me it is DIY.

I have a service manual that I obtained in microfiche medium. One of the reasons being, I wanted to confirm if that medium can be utilized.

I washed the gelatin fiche and dried it in the same manner I would do for film negatives.

Then did some tests scanning it on a flatbed scanner that can scan films. As the microfiche can't fit inside the film holder, I just placed it on the glass surface.

After the required color inversion (scans like blueprints) and desaturation, I ended up with a normal positive page.

But it lacks definition and details. Looks either like lack of focus or lack of scanner resolution.

The scanner is an Epson Photo scanner (V100 Photo) and has a native resolution of 4800x2400. I scanned at different resolutions, applied sharpen filters and reduced size.

A small version of one page is attached. I know it is hardly readable - the original 1200 x 1600 version is somewhat better - I can almost make out the words.

The size of each page in the fiche is more or less the size of a 35mm film frame (if somewhat smaller). So, the scanner should be able to make out details, provided they are there.

I also have a 4000dpi film scanner that performs excellent but don't want to cut the fiches into pieces to use that.

I am thinking it's a matter of focus error. The film isn't normally touching the glass plate (as it is placed inside the holder) and I am guessing that the scanner adjusts focus depending on the type of document to scan.

I am also thinking of using a piece of paper with a cutout the shape of a page and place the fiche over that. Try with different number of pages stacked until I (hopefully) get the best (decent) result.

While I go about this, I wonder if anyone else has done this before and has useful suggestions to offer.

Maybe your local library has a microfiche reader hooked up to a printer, I know mine does

Aspect ratio is 105mm x 148mm (A6). If you input that ratio it might make for a cleaner image. Since the film is always magnified by a reader, I think you need the microfiche specialized copier to convert to standard paper copy.

Most likely the scanner is out of focus, you'll need to raise the film off the bed.

Examples here:
http://www.normankoren.com/Tutorials/Epson_flatbeds.html#Focus_spacing

and here for some hardware:

http://www.betterscanning.com

http://www.betterscanning.com/scanning/usinginsert.html

You might also try oil mounting the microfiche, google "oil mounting" or "wet mounting" for scanning negatives.

Sasi;

If I recall correctly the V100 is a predecessor to the 4990. I've done some research on operating a 4990 and finally bought one. So I gathered a couple of forum links on the way.

1.) This one is based in the UK http://www.photo-i.co.uk/BB/viewforum.php?f=3 They have had several discussions on the 4990 and maybe the V100.

2.) This forum thread discusses a little bit about focus and color. I don't recall going back to the original board location. http://www.rangefinderforum.com/forums/showthread.php?t=15946

3.) My purpose for buying the Epsom 4990 was microfilm scanning. The 4990 works better with 35mm than 16mm. Of course, Mr. McScan here wanted McService out of the machine like yesterday. I'll calm down one day and really get into it.

4.) I think the focus issue is the cause of the lack of detail. It seems that this series of scanners require adapters to get the proper focal length on the film scans. The film needs to be about 1 to 2 millimeters off the bed.

Good Luck!!:thmbsp:
Walter

Have you verified the microfilm image is sharp with a magnifier. If it is a copy of another film, that may be as sharp as it gets.

The scanner is an Epson Photo scanner (V100 Photo) and has a native resolution of 4800x2400.

Epson specs your optical resolution at 3200 dpi. Anything higher and you get software interpolation, which will yield artifacts and softening. You should get your best results at even divisions of 3200 if you scan at lower resolutions (1600, 800, 400...), 1200 should not be an optimal scan.

OK, I took a closer look at your posted scan... it looks as if there are not enough pixels for sharpness. If you would like to email me an original scan, I'll be happy to do some post-processing and see if I can make a helpful suggestion. (My email can be found in the "v-card" download in my user profile.)

Also, I would ask if you have tried putting a white surface behind the microfilm and scanning it as a positive?

I did some focusing tests on a Canon flat bed scanner i had when I was using it as a fast way to get a photo of something onto ebay. And even a couple mm can greatly effect the focus on a scanner. Also since a Microfish is basically a clear cut through the emulsion on the film you may be getting diffraction from the tiny openings created by the letters and lines when a light is shined though it.

I would as suggested try a positive scan with a couple layers of white paper on top of the film. And something like a book over that to hold it very flat.

Thanks to everyone for their suggestions.

I tried to switch to 3200 native resolution (thanks for the pointer) and the results immediately became slightly better, especially regarding noise.

I also tried variations of scanning as a reflective surface with a white background and as film with scanner doing backlight.

I did the tests both with the fiche flush on the bed and using pieces of paper acting as spacers. I increased the height by adding a piece of paper every time I scanned and did a reflective and film scan mode each time.

The obvious result is that if I scan the fiche in reflective mode, the more I raise it from the scanner's glass the worse focus gets. After 4-5 pieces of paper I can clearly see that focus is poor.

Using film mode scanning, the results are rather mixed. I was expecting that raising the film would / should improve results as the film holder keeps some distance from the glass. But while I do get improved focus with every spacer piece of paper I add, I do get increased flare and chromatic abberation as the height of the fim increases.

@Ozmoid, thanks for the post processing suggestion but I don't expect post processing to improve the current results. I tried scanning a 3200 dpi and reduce that to 1600 after sharpening (tried several different options) but the original is out of focus.

I will try a few more distance settings, at least to reach a conclusion at what distance I get the best results. I am not hopeful, however, that these results are going to be adequate.

I concluded the tests with the flatbed scanner. I had the best results with backlit film mode scanning when the cutout window in the paper was almost the size of the scanned image on the microfiche. Best result regarding focus were given with 6 pieces of 80gm/sq.m paper as spacers.

Unfortunatelly, these results were mediocre. Much better than my original scan but in absolute terms, a failure.

So, I took the brave decision to cut a strip from the bottom of the fiche and scan that using my dedicated film scanner (Canon FS-4000).

The attached picture is of my first attempt to scan one of the pages on the fiche, desaturated and reduced to 24bit depth. The original scan was done at 4000dpi, the scanner's native resolution and 42 bits. I only utilized infrared clean in medium strength, as the fiche films appear to be quite beat up.

At least this time, with this scanner, the fiche will become useful.

I hope (now I really do) that some better cleaning of the fiche and some post processing will allow for better and more readable results.

It is worthwhile to point out that when I did a comparison between that V100 flatbed and the film scanner with slide film, I realized that the film scanner was substantially better at details. Considering that a page of paper on that microfiche occupies less than 1/8 of the 35mm film frame, it is no wonder that the resolution of the flatbed scanner is inadequate for microfiche scanning.

Sasi;

It appears to my eyes that the original has ink bleeding into the paper. The printing process would appear to be the first issue with the scan. Unfortunately, nothing simple can be done to help that. I tried a few tricks with your attachment with photoshop. The software only worsened the scan. The only thing that photoshop could do would be a manual erase of the bleed area. That is not a quick fix at all.

At least you don't have document blurring from the auto-feed scanner that made the film! Or double exposure, that was one nasty section of 35mm film that crossed my eyes for a month.

Walter

Sasi;

It appears to my eyes that the original has ink bleeding into the paper. The printing process would appear to be the first issue with the scan. Unfortunately, nothing simple can be done to help that. I tried a few tricks with your attachment with photoshop. The software only worsened the scan. The only thing that photoshop could do would be a manual erase of the bleed area. That is not a quick fix at all.

At least you don't have document blurring from the auto-feed scanner that made the film! Or double exposure, that was one nasty section of 35mm film that crossed my eyes for a month.

Walter

After scanning a few strips off the microfiche on the film scanner, I can say that the results are moderately good. The fiche itself is in nasty condition (heavily scratched and with dirt that won't go even after two baths), but the film scanner can extract good detail out of the text portions. There are some diagrams and b/w pictures that have little contrast and reproduction is poor.

All in all, it appears that the microfiche was well done considering it's 20 years old and it's just the bad condition as well as the inability of a flatbed scanner that produced the first results.

Focus appears to be of paramount importance. As the microfiche strips are not the size of a 35mm film, they don't engage well in the film strip holder. Only the middle portion of the strip is adequately supported by it and the edge portions are awfully out of focus. That means that I have to shift the strip to the left and right to get well focused results.

It appears to be a tedious excercice but I might as well do it as I can't get this service manual in any other form (it's a Technics RS-9900US sm).