Digitizing medium format film with m4/3

[RAH]

So there is a flatbed photo scanner that has a second moving CCD element on the lid?

OK, you're right, of course. The scan element is underneath, and as @newmikeman says, the light moves in tandem with it from above (I'm pretty sure).

But anyway, the point is that when you are doing a transparency, the scanner knows it and from my understanding (and as evidence all those holders that you get with it), it expects the item to be raised off the bed, similar to a slide. It may have enough DOF to still be in focus, of course, although, as I have said, I have read that it doesn't. But @John King has had good luck, so that's very telling, of course.

 

[BDR-529]

But anyway, the point is that when you are doing a transparency, the scanner knows it and from my understanding (and as evidence all those holders that you get with it), it expects the item to be raised off the bed, similar to a slide. It may have enough DOF to still be in focus, of course, although, as I have said, I have read that it doesn't. But @John King has had good luck, so that's very telling, of course.

I saw somebody mentioning Epson V700 Photo in previous page and according to some user stories I found, this is expensive enough to have two fixed focus points. One for full scanning area (i.e. A4 document directly on the glass) and second for negative holder which also activates the light source on the lid.

That being said, I don't really see any difference between sample photos users have uploaded with or without negative holder. I believe that bigger issue here is that pixel peepers just don't realize that film is never "sharp" in the same sense as digital images due to it's grain structure. If I remember correctly my Canon flatbed boasted with 9600dpi in one direction and scans of even sharpest 35mm slides looked just awfull because grain became so visible that whole image was made up from random shaped blobs.

There are, or at least used to be flatbed photo scanners with AF but I don't know if these are still around.

 

[RAH]

With 20/20 hindsight, I wish I had taken a chance and gotten the V700 scanner for the museum I volunteer at. Considering @John King 's findings and what you have said, it probably would have worked fine. I wish I could find the sources I saw earlier (like 6 or 7 years ago) that said it wouldn't work well, but haven't been able to (Google is so riddled with so many ads nowadays that it seems impossible to actually find anything with it, and DuckDuckGo wasn't much help either).

I agree with you that the results of scanning film are often really poor looking. In fact, I think that was another reason I backed off the idea of buying the more expensive Epson scanner. I was afraid the results would look terrible but then I wouldn't know whether it was because of the holder issue, not using the "wet" process, or just the way they look. Using a camera seemed the more foolproof method (at least for this fool!).

 

[BDR-529]

I agree with you that the results of scanning film are often really poor looking. In fact, I think that was another reason I backed off the idea of buying the more expensive Epson scanner. I was afraid the results would look terrible but then I wouldn't know whether it was because of the holder issue, not using the "wet" process, or just the way they look. Using a camera seemed the more foolproof method (at least for this fool!).

I have been "scanning" 35mm slides and negatives with 20MP m4/3 camera and inexpensive Olympus 30mm f/3.5 macro lens. When I compared the results with what I got from old Canon flatbed scanner I never went back to insanely slow scanner. With camera setup it's just: click-new slide-click-new slide-click....

It is possible to get a lot more pixels out of flatbed scanner but I found out that 20MP is pretty much the limit where film grain starts to become visible anyway so more pixels didn't add any details or sharpness, grain just became larger. I used mainly ISO 200 Fuji back in the day so this might not be true for super fine grain film.

I also made an interesting test after I realized that said lens can focus so close that 20MP covers just half of 35mm film which essentially created a 2x magnification. Same result here: I didn't really get additional details or sharpnes because the size of film grain was the real limit here (and also the sharpness of lenses I could afford back in the film era).

 

[Bytesmiths]

How does the half pint of beer fit in?

It used to be a full pint, until he figured things out.

 

[Bytesmiths]

I did search for the word "wet" and got no hits. As I understand it, the absolute best way to scan negatives is using a method called "wet scanning."

I used to have a 5,400 spi drum scanner. (Color Getter Falcon) You really need to use wet mount on anything much more than 2,400 spi, and it doesn't really improve things much under 2,400 spi.

The main thing it does is eliminate interference fringing that come from having an inconsistent spacing between the film and the glass or acrylic holder. If you're just shooting it with a camera in free space, there's no advantage to wet mounting.

It comes with its own challenges. You have to make sure you have it spread evenly, without any bubbles, and of course, the neg/tranny needs to be very clean. And speaking of "clean," you have quite a clean-up when you're done scanning.

 

[Bytesmiths]

9600dpi in one direction

Whenever a scanner is spec'd with different resolution in the x and y axes, beware. The higher number comes from stepping the sensing head at a fraction of its native resolution, which doesn't actually result in more pixels that are sharp.

 

[RichardC]

Whenever a scanner is spec'd with different resolution in the x and y axes, beware. The higher number comes from stepping the sensing head at a fraction of its native resolution, which doesn't actually result in more pixels that are sharp.

Is this because scanner pixels are rectangular?

 

[BDR-529]

I also made an interesting test after I realized that said lens can focus so close that 20MP covers just half of 35mm film which essentially created a 2x magnification. Same result here: I didn't really get additional details or sharpnes because the size of film grain was the real limit here (and also the sharpness of lenses I could afford back in the film era).

And here's a sample from this test. I did also use Olympus 30mm f/3.5 macro to zoom in as close as possible (original negative was on LED lightable). This setup did create 2x zoom so essentially I did take a 20MP MFT image from the center of a FF negative which equals scanning the full 35mm negative as 80MP image. I did crop a 1000*800 pixel rectangle from the outcome so that you can see it in native resolution even here.

Point is that added resolution did not improve anything compared to normal case where I zoom in so that 20MP shot covers the whole 35mm negative because what you see here is just film grain. Since the size and shape of this grain is random and it's not in a neat predictable matrix as digital noise from sensor, even Topaz can't do much to it. (Sample is skin which looks absolutely horrible, especially on face if scan resolution is increased). Film here was actually ISO 400 Fuji so this setup could squeeze out more details from ISO 100 film if it has really fine grain. It also looks like film grain of one colour is not evenly distributed but somewhat clustered close to each other.

Subscribe to see EXIF info for this image (if available)

 

[Bytesmiths]

Is this because scanner pixels are rectangular?

No, the sensors are square. (Well, actually, they are probably circular.)

There is a linear strip of sensors that have a fixed dimensional relationship to each other. So you cannot "oversample" with such sensors. (Well, I guess you could shift the sensor strip in its long dimension by half a pixel, a la Olympus's "High Resolution" mode, but I'm not aware of any scanner that does such a thing.)

So that's one axis. The other axis is controlled by a stepper motor that either moves the sensor strip past the item being scanned, or moves the item being scanned past the sensor strip. This is where sub-sampling can happen, as the stepper motor could move things at less than the natural pixel pitch of the sensor strip.

This is fraught with problems, and its not clear how sophisticated scanners are in this regard. The sensor array has a fixed "aperture" that overlaps when sub-sampling. Simple sub-sampling results in no effective increase in resolution at all! The sub-sampled result will have more pixels, sure, but they will be fuzzy, and generally no better than up-sampling in Your Favourite Photo Editor™.

I haven't looked at the scanner market in some time, and so could be wrong about this, but I don't think the super-high-quality scanner market is big enough to provide the R&D budgets necessary to do the sort of magic Olympus does with High Resolution mode. So I don't think any of those claims of 2,400 spi really hold any water, except in marketing literature.

So I'd welcome updated information anyone might have regarding the reality of resolution on scanners that claim to have more than 600 spi resolution in one dimension. (I believe 600 spi is the maximum pitch available in sensor strips.)

So how do drum scanners achieve 5,400 spi? They don't have a strip of sensors with a fixed dimensional relationship to each other. Rather, they have a tightly focused spot of light that is picked up by an extremely sensitive photomultiplier tube. But it takes a lot of power to squirt enough light through a seven-micron aperture and still be picked up by a PMT! So the highest-resolution drum scanners used some 500 watts of light to do so! They were beasts, and my "compact" Color Getter Falcon (with a 500 watt xenon carbon-arc light source) weighed 100kg (220 lbs) and filled a 48"x30" table top!

 

[BDR-529]

No, the sensors are square. (Well, actually, they are probably circular.)

There is a linear strip of sensors that have a fixed dimensional relationship to each other. So you cannot "oversample" with such sensors. (Well, I guess you could shift the sensor strip in its long dimension by half a pixel, a la Olympus's "High Resolution" mode, but I'm not aware of any scanner that does such a thing.)

I have an old CanoScan 8600F which claims to have optical resolution of 4800dpi * 9600dpi (extended).

Since pixels are - as you mentioned - square, resolution should be the same on both directions. Scanner should also move the CCD strip by steps which are exactly as long as the distance between adjacent photosites and this gives true optical resolution of 4800dpi * 4800dpi.

I'm pretty sure that this extended resolution along y-axis only is achieved by half stepping the CCD strip. (It might be actually easier to move the CCD strip carrier forward by full steps but yank just the strip sideways by half step at each position and take a second shot. 9600dpi was listed for vertical but can't figure out which way the scanner should be looked at)

I tested this extended 4800dpi * 9600dpi resolution only few times because it takes forever and didn't really find any improvement when I scanned 35mm negatives mostly for the same reason I demonstarted above. ISO 200 and ISO400 film has so large grain that additional resolution just made it bigger. Eventually I came to conclusion that it's some sort of marketing gimmic like the "interpolated 19200dpi * 19200dpi resolution" that was also promised for said scanner. It's pretty much equivalent to "digital zoom" which doesn't exist either in real life.

 

[Bytesmiths]

I have an old CanoScan 8600F which claims to have optical resolution of 4800dpi * 9600dpi (extended)

I think that's a bit of a wish, more than a testable specification!

Can you imagine the jitter involved with moving a sensor strip by 0.0002"? Then there's "skew," whereby one side of the sensor strip moves a few ten-thousandths of an inch further than the other side of the sensor strip. I'd guess such a device would use a lead screw, rather than belt drive, but lead screws suffer from backlash, causing positioning to be uneven.

It would be interesting to run MTF tests on such scanners, to see how much resolution they really have!

I recall reading of some scanner tests that all came to the conclusion that anything much higher than 600 spi was wishful thinking. But that would have been years ago.

But perhaps they've conquered all these problems since I last went deep on the topic a dozen years or so ago. There are $500 UV 3D printers that claim such positioning resolution, so perhaps I'm wrong.

 

[John King]

I think that's a bit of a wish, more than a testable specification!

Can you imagine the jitter involved with moving a sensor strip by 0.0002"? Then there's "skew," whereby one side of the sensor strip moves a few ten-thousandths of an inch further than the other side of the sensor strip. I'd guess such a device would use a lead screw, rather than belt drive, but lead screws suffer from backlash, causing positioning to be uneven.

It would be interesting to run MTF tests on such scanners, to see how much resolution they really have!

I recall reading of some scanner tests that all came to the conclusion that anything much higher than 600 spi was wishful thinking. But that would have been years ago.

But perhaps they've conquered all these problems since I last went deep on the topic a dozen years or so ago. There are $500 UV 3D printers that claim such positioning resolution, so perhaps I'm wrong.

Jan, the print head on my Epson R3880 is positioned by three lasers. This is so accurate that an ink cartridge that runs out can be changed mid-print without a trace.

This is printing at 2880x2880 dpi, according to Epson.

Similar technology on my slide scanner, which scans at 5400dpi.

 

[RichardC]

Sometimes I need to scan a highly textured photo.

A solution to the stippled finish SHOULD be to make a number of scans, rotating a little each time, then lining up in photoshop layers with blend mode set to darken.

The images scanned at 90 degrees never quite line up with the ones scanned at 180 degrees on my Epson 700 photo - they are distorted just a little - which is why I wondered whether the pixel shape or pixel layout was not quite square or perhaps rectangular (they do seem to quote different resolutions on the x and y axis).

I've considered getting a higher end scanner - but don't really do enough to justify the enormous cost of a GT10000 or whatever the modern equivalent is.

This is how GAS starts.

 

[John King]

Epson V700 review here:

https://www.photoreview.com.au/revi...eviews/scanner-reviews/epson-perfection-v700/

 

[RichardC]

Epson V700 review here:

https://www.photoreview.com.au/revi...eviews/scanner-reviews/epson-perfection-v700/

That looks just a tiny bit more advanced than mine. I think I might have misquoted my model number

 

[John King]

That looks just a tiny bit more advanced than mine. I think I might have misquoted my model number

Richard, there were very similar looking Epson scanners of the same era. Most were a lot cheaper.

 

[RichardC]

Richard, there were very similar looking Epson scanners of the same era. Most were a lot cheaper.

Mine WAS cheaper - probably why they filled it with rectangles.

 

[John King]

@RichardC Richard, I bought mine as a "kit" with my R3880. With some A2 paper, I got little change from AUD$3,000 about 10 years ago. Not cheap! IIRC, the scanner made up over $700 of that, paper about $300, and the printer a bit under $2,000. At the time, the printer alone was around AUD$2,300. A LOT more expensive than it was in the USA, even allowing for exchange rates at the time.

You are probably right about the "rectangles".

My dedicated slide scanner, Konica Minolta Scan Elite 5400 MkII cost me around AUD$1,000, and that was a stock run-out. It's a beauty. I keep a PC running Win XP Pro just for it (and other occasional use). No drivers for it after XP. Vuescan costs a lot, with nothing like the functionality of the factory software.

 

[RichardC]

I knew there was a 7 in it. A Perfection V370 photo. Thanks to you I have now got scanner GAS.