Your Scanner Needs Cleaning!

By Ctein

No doubt some of you digital printers have been gloating at the hoops Mike has been jumping through to get himself established for traditional darkroom printing, as well as the myriad comments from all and sundry about the efforts they have to go through to control their printing environments. Problems with stray light, problems with light proofing, problems with safelights. All to make those unruly photons be where folks want them and not where they don't. How nice that digital printing is not heir to the same problems, don't you think?

Think again.

The truth about shmoosh
When was the last time you had your scanner cleaned? I'm not talking about the exterior surfaces like the platen on a flatbed scanner or the film carrier for a film scanner. I'm talking about the interior optics—the back side of the platen, the first surface mirrors inside the scanning head, the relay lens that's between those mirrors or between the mirror and the CCD array (see figure 1). If you've got a scanner that's more than a handful of years old, I can pretty much guarantee that stray photons are degrading your scans. And it's almost certainly crept up on you so gradually that you never even noticed.

Here's the thing. Even though all those internal optical surfaces are protected from direct exposure to the environment, the scanner isn't airtight. Dust creeps in, especially with a film scanner, which can't really be well sealed at all. In addition, the plastics, lubricants, and other synthetics inside the scanner outgas. You know that smudgy haze of plastic volatiles that appears inside your automobile windshield and can make it really, really hard to see when you're driving into the sun? Scientists call this "shmoosh." And, yeah, your scanner is like that. Or it will be. It just takes time.

Fig. 1. This is what the typical light path in a flatbed scanner and a film scanner looks like. The exact arrangement of mirrors and lenses will depend upon the scanner, but this gives you the idea. Shmoosh will show up on the underside of the glass platen, on the surfaces of mirrors, and on exposed lens surfaces.

What's the effect of that veil of deposits and dust that's coating mirrors and lenses? If it's prints you're scanning, it'll be something like figure 2. In mild cases, the veiling will be uniform and you may not even realize it's happened and that you're losing shadow detail. In severe cases, the deposits can be nonuniform, especially on the underside of the platen, and you see tonal variations in your scans that aren't in your originals but appear in the same place every time you do a scan.

Fig. 2. Shmoosh degrades shadows. At the least, it veils them; in severe cases, it may be a irregular enough that you don't even get uniform density, as seen in the lower third of this photograph.

In film scans, you may get the same kind of degradation as in figure 2, but the highly directional transmitted light and the high density range in some films (color slides and high-contrast black-and-white negatives) means you're also likely to see haloing, like in figure 3. When the highlights start bleeding, it's time for cleaning. Actually, it's way past time.

Fig. 3. Film scans can not only suffer a loss of contrast in dense areas but also have haloing when there is shmoosh in the light path. In positive scans, like what's shown here in the diagonal upper half, the halos will be light intruding into dark areas. In negative scans, it will be dark bleeding into highlights.

No help from me
Now here's the bad news. I won't tell you how to do this yourself. There are so many ways to go wrong and ruin your scanner. If you don't get it reassembled exactly the way it was when you took it apart, components may be out of alignment and you won't get sharp scans. The glass platen in a flatbed scanner is pretty durable, but the interior optics in any scanner aren't. The mirrors have their reflective coatings on top of the glass, not behind it, to avoid ghost images. Those coatings are very easy to scratch.

If you're good at mechanical assembly and disassembly and know exactly how to clean first-surface optics, go for it. If you don't, you need to hand this off to a trained professional. Don't email me asking me for tips or advice. I'm not going to be the "enabler" on this one.

Unfortunately, because I can do all the stuff myself, I don't know the names of good repair people to send scanners to for proper cleaning. Do any readers have recommendations?

Ctein

Ctein's regular weekly column appears on Thursday mornings. We're on summer hours, so for now that means late mornings.

Comments

[David Dyer-Bennet]

I killed my Nikon LS-2000 cleaning it.

I hadn't recognized the symptoms of gradually-increasing haloing, until one day I woke up and realized This Was Not Right! Looking at prices and features, I decided to risk cleaning it myself (I just blew off the front-surface mirrors with air).

I still don't really know what happened; I reassembled it and did a test scan and it WORKED, much better than before. And then the next morning it didn't work, and I never got it to work again.

So this wasn't even a case of ruining the alignment, or scratching a mirror, or one of the obvious problems. I don't know what the real problem was (I went in and wiggled connectors and all that afterward, didn't help).

In any case, I sold the carcass for parts on Ebay, so don't go telling me what I did wrong, it's no use to me now :-) .

I replaced it with a Coolscan 5000, AND got the slide feeder for it.

Which is now old enough that I should probably send it to Nikon for official TLC, before they forget how.

[gromit]

Hey! Lock and Dam #1!

-g

[Bill Rogers]

Soooo ... don't scan. Use a digital SLR. Don't you wish all problems were so easy to solve?

[Stan B.]

I'm sure this does happen- to other peoples' scanners. They're usually very mean, generally unreasonable and probably most deserving of it.

[davide]

Any advice from the community on where to get artixscan or polaroid scanners cleaned?

[ctein]

Dear Bill,

And you're gonna loan me your time machine, so I and my clients can retroactively produce digital versions of what we've already photographed on film, right?

Thanks, really appreciate your generosity!

pax / Ctein

[Paul Ewins]

I've got an Epson 2450 and from time to time have wondered whether it would be possible to improve its film scanning performance by replacing the glass with a thick piece of aluminium with a cutout to suit the film scanning area (something like 4"x10" from memory). There would probably be a bit of shimming to get the right height, but apparently they aren't always at optimum when they come from the factory.

[Gabor Samjeske]

When my (film)scans showed that same kind of haloing for all my scans and not only the ones taken with old Leica lenses, I got in doubt about the famous "Leica glow" and thought that the scanner might be dirty ... So I disassembled my Coolscan 4000ED and was able to clean the mirror but not the optics. Disassembling was easy but assembling took me a long time to avoid scratching the mirror... I would not do it again but bring my scanner to Nikon service and have it professionally cleaned.

[Paul H]

I cleaned my old 3170 a couple of times - I think a flatbed is a little less daunting than a dedicated scanner. The last time I did it I went through and 'matt blacked' all the shiny surfaces following an article I saw somewhere on the web. That last bit of work did seem to improve things.

The V700 is much better all round, and doesn't seem to out-gas as bad as the 3170 either.

[Nick]

I would wonder how much extra scanner lifespan one can buy by cleaning. Even the most skilled and careful cleaning is going to take its toll on those mirrors over time, and some of the crud may not be removable at all. Perhaps the inside of a scanner is a better environment than the laser labs where I work (not impossible; environmental control is shockingly poor at my place of employment), but the quality of even our "protected" first-surface silver mirrors tends to degrade over time, and after a few years, no cleaning can rescue them.

[Ken Tanaka]

This is actually one of those little issues I've pondered for years. Clearly, from the description:
"If you're good at mechanical assembly and disassembly and know exactly how to clean first-surface optics, go for it. If you don't, you need to hand this off to a trained professional.
...as well as the testimonials from others here I'm not inclined to tackle this job with my Nikon 5000 or my Epson 750 any time soon (i.e. ever). Nor am I inclined to spend/send them for "professional" cleaning.

Nope, I'm inclined to take two measures. First, I'm going to declare that the haziness is an artistic interpretation (hey, there are PS plug-ins that do that, too) for some images. But secondly, and more importantly, I'm going to continue to back away from film. (I'm already 95% there.) Film's PGR (Pain/Gain Ratio) has already exceeded 2.0 for me, and this just sends it to 2.25.

But thank you for exposing this dirty little topic, Ctein. It really is something I pondered.

[David Bennett]

Which scientists call it 'shmoosh'? They have not found their way to the internet, whoever they are.

I suspect this is a term coined by a practical scientist named Ctein.

[David Bennett]

As an aside on how quickly Google picks up a term, you can see that The Online Photographer is already a source for 'shmoosh'

Shmoosh on Google

[David Dyer-Bennet]

I think what Bill was suggesting was using "slide copying" techniques with a digital SLR to digitize film. It's much faster than a scanner, the resolution depends on your camera, there's no ICE, the brightness range ain't half bad these days.

But if I'm wrong and Bill DOES have a time machine, I've got a few projects I'd like to borrow it for too!

[Mark Roberts]

Very timely subject for me as I'm just preparing to start some scanning with my Minolta Scan-Multi II. But I am completely amazed (and not a little disappointed) that the collective wisdom of the followers of TOP haven't yet offered any source for scanner cleaning services...

[ctein]

Dear Nick,

A good question, and I think it would depend entirely on how badly one's scanner surfaces collected crud. In my experience, cleaning the surfaces will extend the life of the device by at least a factor of three, but everyone's mileage will no doubt differ.

The situation inside the scanner is nowhere as bad as it is in your laser lab. None of the reflective or transmissive surfaces that acquire garbage (save for the underside of the platen) are anywhere close to image planes, focal planes, or nodal points in the optical system. As a result, slight scratches, nicks, and bits of crud that can't be removed don't noticeably degrade the image quality. I only did a half-assed cleaning on my scanner, because to go the whole anhydrous alcohol/distilled water/etc. route would've required completely disassembling the optical head to remove the components, and I had no faith that I would be able to get it back together again properly. So I gave the surfaces a cursory cleaning. They're 95% better than they were before, but they are terribly far from perfect; as a darkroom printer and amateur astronomer, I'd have fits if I saw that much crud on any of my lenses or mirrors in those venues. In the scanner, the performance is almost like new.

~ pax \ Ctein
[ Please excuse any word-salad. MacSpeech in training! ]
======================================
-- Ctein's Online Gallery http://ctein.com 
-- Digital Restorations http://photo-repair.com 
======================================

[ctein]

Dear Ken,

I love the idea of a PGR. How you calibrate that? Is there a unit standard reference in a bell jar under dry argon somewhere in Paris?

~~~~~

Dear DDB,

Oh you're no fun. My interpretation was a lot cooler.

Slide copying techniques, done with even the very good cameras and lenses, produce crappy results compare to a dedicated film scanner. Yes, they are a lot faster and easier. Faster and easier aren't part of my artistic niche; quality is what counts. In terms of color and tonal accuracy and bit depth, total density range captured, and resolution, I would have to spend at least $30,000 on digital camera equipment to approach what I can do with a $2000 scanner. If I just threw away the scanner every two years and bought a new one, I'd be better off economically!

The situation is different for flatbed scans. A very good camera copy setup can frequently rival a mid-line scanner, for all practical purposes.

(An aside: no, I don't want to hear from any readers telling me how wonderful their "slide copy" film scans are. No, they're not. They're quick and very easy, but they are not "good" save for a very small value of that word.)

~ pax \ Ctein
[ Please excuse any word-salad. MacSpeech in training! ]
======================================
-- Ctein's Online Gallery http://ctein.com 
-- Digital Restorations http://photo-repair.com 
======================================

[Paul logins]

I agree with Bill Rogers. Got rid of film,enlarger,trays,chemicals,
etc.,etc.,etc. years ago. I shoot with digital cameras, adjust pics
on computer software, and download to iPad for displaying my
pics to others. If I ever get an once in a lifetime photo,I'll blow
it up to 13x19 on my inkjet printer for hanging on the wall so
all my friends can admire how great a photographer I am.

[Doug Fisher]

If you have an Epson flatbed, I put up a page a while ago with .pdf files and tips to help disassemble these scanners at:

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

The information posted there should help many Epson owners get inside the case to clean things. Most people stop after cleaning the underside of the glass. One word of caution though - use cleaning fluid sparingly. Do not let the white calibration strip on the underside of the case near the top of the glass get wet/discolored or you could experience problems with start-up calibrations. Underside cleaning of the glass bed alone can do wonders. As Ctein pointed out, cleaning the optics can help even more but there is a real risk of damaging things. Based on feedback from others though, I am not sure the average authorized Epson repair center tech will get it right either :( Some are very competent, others are not.

An additional tip is to turn your bathroom into a clean/low dust room where you can disassemble things. Clean it well from top to bottom to remove dust. Then run a hot shower to steam the place up and thus "sink" dust left in the air. Keep the door closed and let the room temperature settle down to normal. Enter once quickly with your scanner and tools to do the disassembly. Lint-free towels are a must to keep the underside of the glass dust free before reassembling. Eyeglass cleaner from Wal-mart is a good cleaner to use and cheap in the small size (plus you supposedly get free refills for life!).

[David Dyer-Bennet]

I find use for a wide range of qualities of scans when dealing with my old materials. They're mostly not useful as art, but as some sort of record, and sometimes it's more worthwhile (it seems to me) to put decent scants of 20 photos online rather than first-rate scans of 2.

Of course my Coolscan 5000 is not the ultimate in scanning, either; I've had occasion to resort to drum scans from a lab now and then.

[David Goldenberg]

Ctein,
If you can't offer suggestions for cleaning a scanner, what about a good test for scanner shmoosh? I was wondering if a slanted-edge MTF measurement might be a way of monitoring scan quality over time. If that is reasonable, is there an easy way to make a good slanted edge target for a film scanner? I have heard of using a razor blade in a slide mount, but I wonder about the reflection from the metal. One could photograph a target, or expose a piece of film directly with a razor blade on top of it.

Any thoughts?
David

[Joey Harrison]

I wonder if you could thwart creeping 'smoosh' by simply draping your scanner with a lint-free cover -- a sort-of scanner cozy. Could be a profitable enterprise for some techno-geeky Etsy member.

[Mike Johnston]

Joey,
A square of silk works nicely for that purpose. Available at fabric stores.

Mike

[Howard Shaw]

Nothing to do with scanning but I'm curious about the curved lines, particularly of the canal side at left, in Fig. 2. What kind of camera/lens produced that effect?