Sensor cleaning

ChrisR · Post by **ChrisR** » Mon May 02, 2011 11:30 am

Thanks foir the reassurance re IPA.
In the UK we can't buy uncontaminated ethanol. the closest is "methylated spirits" which apart from a vague spec, has a purple bitter taste added.

I'd cleaned the sensor before, but the most recent attempt made a real mess. I only had half a dozen pads though. I suppose they'll wash out, perhaps in acetone. They need to be narrower too - folded perhaps.
The fibers coming off the pads seem to be the source of the longer ones on the sensor. I took a lot off the inside "fur" of the mirror box with fine tweezers.

OzRay · Post by **OzRay** » Tue May 03, 2011 1:24 am

Whatever the failings of Olympus DSLRs, sensor cleaning is never an issue; it's rarely, if ever required.

Cheers

Ray

Charles Krebs · Post by **Charles Krebs** » Tue May 03, 2011 11:06 am

One reason I made up the "filter-sealed" T-mount is that I picked up a 3Ti to try some HD video (and I like the articulated screen, and my 50D needs to go off for a shutter repair

). It is one thing to spot out dust in a still image... out of necessity, I have gotten pretty efficient at doing this

. But video is a different matter entirely. An amount of dust that I would grudgingly deem "acceptable" for stills would drive me crazy in video. Since I intend to keep this body on a microscope, sealing it up this way while still new seemed like it might be an interesting experiment. In time it might also yield a partial answer to something that I have wondered... are all these "microscopic" dust particles from the outside environment, or are some actually "generated" by internal camera components? (Initially I removed a couple of specks that seemed to be there out of the box... spots that would never have been noticed in normal" photography).

The two obvious concerns are flare and possible image degradation because of the filter. I am a firm believer that any glass in the image path that it not absolutely needed be avoided for these reasons. But so far I have not encountered a lighting situation where flare has been an issue. I have not yet done an A:B comparison in regards to image quality, which is something I really need to do. It is never a good idea to draw any conclusions without doing such direct comparisons, but the image quality looks excellent with the filter in place.

Again I want to caution about my low-tack "glue dots" on the toothpick technique. But it is worthwhile to develop some method of "grabbing" individual spots that works for you. Very often I will have the sensor spotlessly clean. Then, after some usage, one or two offensive dots appear. I know full well that if I were to do a full wet sensor clean to get those few particles it would be necessary to spend much more time cleaning up the edge and corner dust I actually introduced while cleaning. It's nice to put the body under the scope and just remove the offending specks.

ChrisR · Post by **ChrisR** » Fri May 06, 2011 8:03 am

I found 4 more new pads, not sure ofthe make. Using Eclipse fluid, or IPA, they do much the same as the "Tech Pads". That is, they push the rubbish around the sensor without picking it up. So until I do some origami with lens tissue and try again, assuming it works, this is part of any image, 110 shot stack:

Pass the Prozac.

Mitch640 · Post by **Mitch640** » Fri May 06, 2011 8:06 am

Never seen anything like that. Looks like you got a coating of bacon grease on there. And you should go side to side, not up and down. LOL

ChrisR · Post by **ChrisR** » Sat May 07, 2011 3:01 am

Those are tracks from spots in stacking, not smears from swabbing.

I called Nikon. "You must never touch the sensor with anything, including any wet cleaning.". They will clean it for £30, taking 3 days.

I tried a local camera shop ( of the rapidly-vanishing type) and came out with a much less fluffy pad, a book of lens tissues and a SpeckGrabber

Eventually I got a lot of the rubbish off the sensor with the better pad, wrapped in lens tissue. IPA alcohol worked quite well when it had almost all evaporated from the pad.

Using Charlie's suggestion of the dissection scope, I found all sorts of stuff stuck on hard, and a collection of very very small scratches and chips, which were a bit surprising as I don't think I've ever done anything unreasonable to my sensor. Lighting is crucial, you have to move it around to see things. I don't see how the rather expensive sensor-illuminator ringlight products could be used for that.

Charlie - you might try the SpeckGrabber and instead of the wall-glue. It deserves high praise. Hairs and dust specks are extremely easy to lift off. Small dirty blobs, which are I think the toughest to deal with, do respond to a stippling action. In some cases they spread about a bit , but repeated action with a cleaned tip does gradually remove them. Stippling removes any smudges you add, too.
I used a pad of lens tissue, with a damp (IPA) patch and a dry patch. Wet cleans the tip, and dry removes the IPA. Waving my arm around also made it evaporate. Ethanol would be better, I think. My Grabber came with some surgical type wipes for the purpose.

As Charlie says, this has to be done at high magnification or you just can't see. I have some strong "readers", +5 dioptre, £1, which are pretty good though.
I shall be ordering the heavier duty version of the SpeckGrabber, sometimes a little force works better.

It would be so much easier if this could be done "live". An obvious approach would be to set up Live View with a large screen. I think I get about 20x with zoomed-in Live View. Say that were on a 30 inch screen, that's about 4000/20 = 200 pixels in 30 inches, or 6.6ppi. I daresay HDMI conversion, or something, would reduce that.
But with a suitable lighting arrangement, perhaps a bright led torch(flashlight) with a pinhole, in one hand, and a SpeckGrabber in the other, microsurgery could be performed down the throat of the live patient.

I must be pointed out that the risk of the camera closing its shutter and dropping its mirror while the Grabber were in there, would be high, and the results could be catastrophic, so don't do this at home.
In normal sensor cleaning mode, there's no bright display to drain the battery, and the only way to shut the camera is to turn it off.
In Live View mode, the battery drains quite fast, and an unintentional knock on any of a number of the buttons on the camera would shut it.
I emphasize - I haven't tried it yet.

http://kinetronics.com/store/speckgrabber_product.html
http://kinetronics.com/store/maintSGP.html

Edit:
A couple more thoughts:
>For inspection, a doctor's Otoscope (The things for peering down ear canals)could be a good tool, perhaps withthe addition of a loupe's magnifier.
>I've seen the "tacky" plastic described both as "copolymer" and "polyurethane". Whether it can be the same thing. I'm not sure. But I used to play table tennis with a bat which was tacky - you could pick up a ball with it. It was the material doing it, not some sort of glue. Newly scrubbed it was at its tackiest. The material is available for refurbishing t-t bats. (Do you call them "paddles" Sateside, or was that only Mr Gump?)

ChrisR · Post by **ChrisR** » Sat May 07, 2011 3:13 am

Whatever the failings of Olympus DSLRs, sensor cleaning is never an issue; it's rarely, if ever required.

I thought the same of mine for a long time. Presumably it's down to what happens to the camera. Sticky particulates would always be a problem?

The camera shop owner told me a of a lady customer with a Canon 400D with a dirty sensor. She'd never even considered tying to clean it herself, so it just went back to Canon. They said the surface had been attacked by something. The repair cost meant the camera was scrap.

rjlittlefield · Post by **rjlittlefield** » Sat May 07, 2011 8:53 am

ChrisR wrote:Those are tracks from spots in stacking, not smears from swabbing.

I'm confused. I see what look like big broad smears ("bacon grease" indeed!), plus some bright white spots. The big broad smears run in several different directions -- north-east, north, north-north-east, even a few a bit west-ish (at right of frame). Some of the different directions overlay each other. The smears look completely different from any stacking tracks I've ever seen. So I'm baffled. Can you point out more precisely what's what in this image?

--Rik

ChrisR · Post by **ChrisR** » Sat May 07, 2011 10:31 am

Rik, now you mention it, I wan't very analytical of my trash

The bacon I believe is tracked hairs
Light spots and grey textures are part of the subject - slightly shiny piece of rock.
The stack must have been processed in the reverse direction to normal for some reason, because the output image has smearing round the edges. I thought I'd cropped it off, but on reflection, maybe not entirely. So there's a mixture of smearing (expected), dust trails, and subject banding. I think - but I'm not sure I'm sure.
I thought it looked odd but threw the image across the room into the bin with its mysteries. Shameful - wouldn't make a research scientist.

One way to help illustrate is probably to rerun the stack the "right" way, which is happening as we speak.
Meanwhile, here's a crop, full width of the frame, from the long edge of the frame, and right to the bottom this time. DMap showing less dust, and the Pmax.

(The "bottom" stacking edge-smear band width may look unusually narrow, relative to its width at the sides, because the camera was inverted.)

Original size

rjlittlefield · Post by **rjlittlefield** » Sat May 07, 2011 11:08 am

Ah! Many thanks for the further images and explanation. I think we're seeing several different kinds of streaks, and it was the combination that confused me. The streaks that are exactly perpendicular to image edge are probably due to extrapolating pixels, for input images that after alignment did not cover the entire frame. The radial streaks do look like dust trails. And finally is some streaking in the subject itself. Judging from the number of radial trails, this is one nasty problem!

--Rik

OzRay · Post by **OzRay** » Sat May 07, 2011 3:03 pm

ChrisR wrote:
Whatever the failings of Olympus DSLRs, sensor cleaning is never an issue; it's rarely, if ever required.
I thought the same of mine for a long time. Presumably it's down to what happens to the camera. Sticky particulates would always be a problem?

Other than the benefits of the exceptional dust filter (SSWF) above the sensor, the fact that it does sit at some distance above the sensor means that if there were any dust particles attached, they'd be far less likely to be visiible in the image. I've shot and changed lenses in all sorts of horrible conditions and still haven't had to clean the SSWF.

I'm also wondering whether dust particles on the rear element of a microscope objective can have a much more profound impact on the image than some dust specks on a sensor. Dust on the back of a lens has always been a much greater issue than on the front element.

Cheers

Ray

ChrisR · Post by **ChrisR** » Sat May 07, 2011 3:42 pm

I can imagine dust on a lens scattering light about, but the stuff on the sensor is almost in focus. I read that sensors' fliter packs are between about 1.5 and 3mm thick, Canons being about 2mm.
So at an aperture of f/20 , a point on the sensor filter would be about 100 microns across, plus the width of the "point". Oouch!

The Canon 400D I referred to above had been used on a boat apparently. Airborne salty water droplets?
( I asked when I went back to buy another lens..

)

Charles Krebs · Post by **Charles Krebs** » Mon May 09, 2011 12:48 pm

Chris,

Using Charlie's suggestion of the dissection scope, I found all sorts of stuff stuck on hard, and a collection of very very small scratches and chips,

I think it is a very good idea to use something like the "SpeckGrabber" (or my harebrained method

) to remove any obvious particles that can be "grabbed" before running a wet swab (just as you would remove any particles on a lens surface before using lens tissue). No matter how much the cameras sensor-cleaner vibrates, or how much you blow air on the sensor there are always particles that cling beyond the capability of these methods. I neglected to pull out the obvious particles one time, and now under the stereo microscope I can see a nice noticeable scratch where a participle was dragged across the surface with the cleaning swab. Fortunately it only shows up very faintly under the most extreme microscope conditions.

Since I don't have a stereo with epi-illumination (or a ring light) I use two of those inexpensive Ikea lights (JANSJÖ) as close to the front of the stereo objective as possible. Works pretty well.

I'll look into the SpeckGrabber product you mention. It nice to have a product to use (and be able to recommend) that is made specifically for this use, and does not have the possible pitfalls of the procedure I mentioned earlier.

ChrisR · Post by **ChrisR** » Mon May 09, 2011 3:49 pm

SpecGrabber tip - the cheaper one.
The part from a cheap single AA torch shows possibilities. The hole it goes through is where a laser once lived.

g4lab · Post by **g4lab** » Mon May 09, 2011 4:15 pm

The tip of that thing looks like some sticky polyurethane that they make
clean room floor mats out of. They are washable but particulates stick to them and the clean the bottom of your shoes when you walk on them.

They are occasionally also used by operating room doors.

http://www.americanfloormats.com/permat ... icky-mats/

http://www.stickymatusa.com/