907X funky corner at f/32

[astrostl]

I was just checking out my sensor before doing some landscape shooting. I noticed that the top left corner had a lot of issues.

The sensor already looked immaculate under a light, but I hit it with Hasselblad-recommended Photosol ewipes to see if I could get rid of this. The ewipes streaked like crazy, so I did a quick wet cleaning with Eclipse and a VisibleDust MF swab. Also wiped down the lens elements to make sure that wasn't introducing the issue. Still happening.

While I never intend to actually shoot at f/32, I can see some darkening in that spot down at f/11 where I actually shoot. Any ideas?

 

[docholliday]

Those look like smears on the sensor and a few dust spots. Since they are in the upper left of the image, that would be the lower left of the sensor itself. You need to use an oblique light and a high magnification loupe. I use a ringlight and stereo microscope to clean mine and you can always see any streaks left from cleaning solution that's absorbed water from the air. Eclipse is very hydroscopic. When it gets old, it'll absorb moisture which will leave streaks behind after cleaning. The small, sharper dot is dust. The larger, softer dots are something that's evaporated and left junk on the sensor cover glass or are oil spots that didn't get dissolved completely.

I quit using Eclipse because of it's short lifespan once opened. The Visibledust swabs work well when I use the Sensor Clean solution (the middle solution). The VDust Plus is alcohol based and is just as bad as Eclipse about leaving junk on the sensor.

 

[astrostl]

Thanks for the detail doc! The issue was present prior to my cleaning. I also really focused on the corners (not just the impacted one) when hitting with ewipes and Eclipse (all brand new and just opened). I'm also really not shy handling sensors and would describe it as a "scrubbing" at least as much as a cleaning. I have lit loupes up to 60x but no stereo microscope, that's an interesting idea. What light and scope do you use? I'm not sure the dust spots can be dust because they exactly survived multiple wet cleanings, but I am worried about oil.

I have VisibleDust Sensor Clean AND Smear Away on the way, will report on how that goes. But this is a bit of a mystery given how hard I've hit it. Makes me concerned that there is something actually between the glass, filter, and sensor itself.

 

[astrostl]

Well, good news! I just went over the real element OF THE LENS again and it looks like that was the source. This is the XCD 30 and its rear element is both small and bulbous, and it seems that the issue was in one of the creases. I still can't get the sensor cleaned perfectly so continued tools and tips welcome there, but it's WAY better than before. Thanks again!

 

[onasj]

The key to sensor cleaning without losing your mind is to use a truly lint-free cloth. I recommend the woven cloths from Capture Integration. The PEC-PAD cloths are not really lint-free, despite their advertising. In order, I use:
1) Rocket blower
2) Gel stick (get the less-sticky kind for "sony sensors", which is red I think, as opposed to the blue gel)
3) Wet clean with Eclipse (methanol) plus the lint-free woven cloth wipes from CI.

 

[docholliday]

Glad to hear that you got it! The closer a rear element is, the more likely anything on it will affect an image. The front element is a whole 'nother story - it can be disgusting and never affect an image!

I use an older Zeiss Stemi with an LED ring illuminator. I have the scope for other purposes (hot-air rework and microsoldering), but it just work perfectly for sensor cleaning.

PEC-Pads are designed for cleaning film substrates (along with PEC-12), not exactly for sensors. The old Copperhill method got people started using them, but I've never used them for sensors or optics. They are lint-free - as long as you don't cut them, disturb the heat cut edges, or stretch the grain (like when wrapping on a paddle), that's when they start being a fiber distributor. I use the VD MF swabs with Sensor Clean or fiber optic polishing/cleaning cloths, which are probably what comes in the CI cleaning kit.

Gel sticks or "gummy bears on a stick" work ok when new, but start to leave residue on the sensor that'll show up as little ghosts after they age a bit. The residue can be a nightmare to remove as it'll smear like crazy, no matter what solvent is used. Those things need to be replaced often, no matter how much you clean them afterwards. If you do rinse them, use distilled water - not tap as the minerals will cause more headaches.

As far as methodology goes, my method is:
1. Examine the sensor with the scope. If only one or two dust spots exist, I leave it alone. You'll never have a perfectly dust free sensor.
2. For dust, I use Chemtronics Ultrajet All-way duster to blow *above* and across the sensor, then examine again. The All-Way will never expel liquid at any angle.
3. Use fiber optic polishing swabs, dry, to pick up or move stuck dust, then blow again. The swabs are Sanyo Huby-340 BB-001.
4. If a wet stain exists, I use a VD swab with Sensor Clean for a quick swipe. I put on anti-static finger cots at this point to keep skin shedding down from recontaminating the sensor.
5. For a stubborn stain or spots, I'll use the swabs or pads and Eclipse (if I have a fresh bottle), Sensor Clean, or reagent grade methanol. I've cut the paddles from VD swabs to different sizes so that I can wrap the fiber optic polish pads around them to give a nice handle.
6. To spot clean stubborn stains, I use the Sanyo swabs with methanol and roll the swab over the spot, "scrubbing" if necessary.
6. Once the main sensor path is clean, I'll use more of the BB-001 swabs to clean the remnants left by swab/pad pushed into the corners of the sensor so that they won't fall back onto the sensor.
7. Before reassembling the back/body, I'll open the optical path through the camera body and blow through with Ultrajet to remove any dust trapped inside.

 

[docholliday]

As far as lenses go, I use Kimtech Kimwipes with Hilco C-Clear (or Zeiss cleaner in the field).

For stubborn stains or deep cleaning, I use the Zeiss Microscopy "Optical Formula L" which you can't buy, but must be mixed yourself. The composition is 85% n-hexane and 15% 2-propanol, analytical grade. This is more commonly found as petroleum ether/ligroin and 100% purified, isopropyl alcohol. It's very hydroscopic, so mix and keep in a sealed Boston round. It'll absorb water from the air like crazy, which will ruin the batch and make it streak. It's also very flammable, so don't use anywhere near open flame.

With either the fiber optic wipes or Kimwipes (gently), they'll make a lens element sparkling clean. Simply wipe gently and it'll evaporate in seconds. I'll sometimes wet the edge of a Kimwipe and drag across the element with no pressure which'll do the job perfectly.

I'll use the same solution(s) with the Sanyo swabs to clean the hard to get, recessed edges of lens elements as it's least likely to runoff or streak.

 

[Krosavcheg]

Wondering how Isopropyl alcohol would be for cleaning a sensor?
Diluting with distilled water?

 

[vieri]

In the field, I use a rocket blower to keep the sensor acceptably clean; if a wet clean is needed, I use Sensor Clean with their MF sized pads, and - when necessary - Smear Away (which you have to then remove with Sensor Clean). For rear lens elements, I use Zeiss wipes, which I also use for front lens elements & filters. The whole solution is very portable (two small bottles, quite a few pads and a lot of Zeiss wipes), and does a good job to keep the sensor & lenses clean enough. As Doc said, it is impossible to have a perfectly clean sensor all the time, especially if you change lenses; the idea for me is to keep it clean enough that software-based dust spot removal won't become a real nightmare in post-processing

Best regards,

Vieri

 

[docholliday]

Wondering how Isopropyl alcohol would be for cleaning a sensor?
Diluting with distilled water?

NO! Isopropyl alcohol would be horrible and leave massive streaks. It's hydroscopic, just like the methanol in Eclipse, and would leave streaks like crazy. You'd also have to get very, very pure 100% 2-propanol, heavily filtered, and you don't want to see what the cost of that is. Mixing with distilled water would make it even worse. You would have to have Type I water that is deionized AND demineralized. Remember that water, even "distilled" can contain trace minerals and bacteria. That matter can cause microscratches in the surface of the filter.

Ever noticed that old wipes with alcohol in them, like early Zeiss wipes, would streak like crazy? That's because they've absorbed moisture and the solution is imbalanced. The newer wipes (and most good wipes) don't use alcohol and if they do, it's a very small amount - as well as being filtered and purified to remove any harsh contaminants. The wipes with alcohol are for eyeglasses, where there is no cement and different coatings, not for precision optics.

At least modern lenses use a UV cured cement instead of solvent cure. That makes them more durable than before, but they can still be dissolved when careless.

You want a solvent that is mild, yet active, but fast drying. Ideally, n-hexane would clean well, but may not agree with some of the seals/adhesives around the sensor & IR stack. For the Zeiss Optical Formula "L" and normal optics, the n-hexane evaporates very, very quickly which is the reason for the small amount of 2-propanol. Too much of it and it'll streak. Not enough and the solution will dry up before you can actively use it on the lens. Optical cements can withstand the solvent well, unless you do it for hours at a time and pile the solution onto the cements.

BTW, there are two types of Zeiss wipes! The white packet ones for eyeglasses and "general optics" are usually available everywhere like Walmart. The older, black foil (or current white packet that say "cleaning tissue") packeted ones were designed for camera lenses and come with the cleaning kit. They have completely different smells and are saturated with completely different solutions! Note that the "cleaning tissue" ones are priced as more premium as well:

 

[Krosavcheg]

NO! Isopropyl alcohol would be horrible and leave massive streaks. It's hydroscopic, just like the methanol in Eclipse, and would leave streaks like crazy. You'd also have to get very, very pure 100% 2-propanol, heavily filtered, and you don't want to see what the cost of that is. Mixing with distilled water would make it even worse. You would have to have Type I water that is deionized AND demineralized. Remember that water, even "distilled" can contain trace minerals and bacteria. That matter can cause microscratches in the surface of the filter.

Ever noticed that old wipes with alcohol in them, like early Zeiss wipes, would streak like crazy? That's because they've absorbed moisture and the solution is imbalanced. The newer wipes (and most good wipes) don't use alcohol and if they do, it's a very small amount - as well as being filtered and purified to remove any harsh contaminants. The wipes with alcohol are for eyeglasses, where there is no cement and different coatings, not for precision optics.

At least modern lenses use a UV cured cement instead of solvent cure. That makes them more durable than before, but they can still be dissolved when careless.

You want a solvent that is mild, yet active, but fast drying. Ideally, n-hexane would clean well, but may not agree with some of the seals/adhesives around the sensor & IR stack. For the Zeiss Optical Formula "L" and normal optics, the n-hexane evaporates very, very quickly which is the reason for the small amount of 2-propanol. Too much of it and it'll streak. Not enough and the solution will dry up before you can actively use it on the lens. Optical cements can withstand the solvent well, unless you do it for hours at a time and pile the solution onto the cements.

BTW, there are two types of Zeiss wipes! The white packet ones for eyeglasses and "general optics" are usually available everywhere like Walmart. The older, black foil (or current white packet that say "cleaning tissue") packeted ones were designed for camera lenses and come with the cleaning kit. They have completely different smells and are saturated with completely different solutions! Note that the "cleaning tissue" ones are priced as more premium as well:
View attachment 183728

Makes sense. I'm in Japan where sourcing chemicals is a true pain.
I got 100% isopropanol fairly cheaply and use it diluted to clean lenses on occasion. It does streak, indeed, though before it dries.

Haven't found a good sensor cleaning solution. Used Eclipse before it dried out (just like you mentioned before), now got a 2 part cleaning kit from DigiClean, but it streaks insanely (something I haven't noticed with isopropanol).

I used heptane in past to clean moulded silicone part for pacemakers in my past employment, but since then haven't seen anywhere to procure.

 

[docholliday]

Makes sense. I'm in Japan where sourcing chemicals is a true pain.
I got 100% isopropanol fairly cheaply and use it diluted to clean lenses on occasion. It does streak, indeed, though before it dries.

Haven't found a good sensor cleaning solution. Used Eclipse before it dried out (just like you mentioned before), now got a 2 part cleaning kit from DigiClean, but it streaks insanely (something I haven't noticed with isopropanol).

I used heptane in past to clean moulded silicone part for pacemakers in my past employment, but since then haven't seen anywhere to procure.

For lenses, "regular" 100% isopropanol is fine. But, for sensors, where the surface is more delicate, I would want 100% that is heavily purfied and filtered for contaminants. Actually, I'd never use isopropanol on any glass...it just creates a mess.

So far, the most consistently reliable solution is the Sensor Clean (VisibleDust). Mostly because it's not alcohol based and therefore lasts longer and isn't hydroscopic.

It's NOT heptane...it's hexane which is more volatile. For the solution "L", you can use "petroleum ether/ligroin" which is good enough. I use ACS reagent or analytical grade and pure 2-propanol to avoid the contaminants of consumer components.

 

[Krosavcheg]

For lenses, "regular" 100% isopropanol is fine. But, for sensors, where the surface is more delicate, I would want 100% that is heavily purfied and filtered for contaminants. Actually, I'd never use isopropanol on any glass...it just creates a mess.

So far, the most consistently reliable solution is the Sensor Clean (VisibleDust). Mostly because it's not alcohol based and therefore lasts longer and isn't hydroscopic.

It's NOT heptane...it's hexane which is more volatile. For the solution "L", you can use "petroleum ether/ligroin" which is good enough. I use ACS reagent or analytical grade and pure 2-propanol to avoid the contaminants of consumer components.

Thank you for clarifying.
Will try to get around Sensor Clean then!
Just had a look at VDust Plus - says contains isopropanol..
Afaik, you use Sensor Clean first, then VDust after...

 

[docholliday]

Thank you for clarifying.
Will try to get around Sensor Clean then!
Just had a look at VDust Plus - says contains isopropanol..
Afaik, you use Sensor Clean first, then VDust after...

You don't the need the VDust. Sensor Clean is used by itself (or if you use Smear Away, then you must follow up with Sensor Clean). The VDust is absolute garbage and just causes more problems. I've used Smear Away once (literally after finding oil spots on a sensor) and it's great at removing oil deposits, but leaves it's own special hell of streaking.

 

[Krosavcheg]

You don't the need the VDust. Sensor Clean is used by itself (or if you use Smear Away, then you must follow up with Sensor Clean). The VDust is absolute garbage and just causes more problems. I've used Smear Away once (literally after finding oil spots on a sensor) and it's great at removing oil deposits, but leaves it's own special hell of streaking.

Great tip. Though last time I used Sensor clean it left massive visible streaks on the sensor that just wouldn't clear until I did a pass with VDust.
Maybe I should have left it a bit longer before applying to sensor?

 

[docholliday]

Great tip. Though last time I used Sensor clean it left massive visible streaks on the sensor that just wouldn't clear until I did a pass with VDust.
Maybe I should have left it a bit longer before applying to sensor?

I found that the reason for streaking in most cleaning applications is that the instructions are very poor and don't give a good understanding of how to clean! The trick is to use as *little* solution as necessary to moisten the swab. In the case of the medium format 33 or 44mm swabs, 3 drops is enough - sometimes 2. I'll put one drop 1/4 of the way in from the edges and one dead center (think of it as 25%, 50%, and 75% of the width). For small format swabs, I'll only do two drop, about 1/3 of the way in from each edge (33% and 66% of the width).

Then, (and here's what the instructions lack) let the solution soak in and evaporate for about 15-20 seconds. I'll actually "fling" the swab like an old mercury thermometer. The centrifugal force will help distribute the solution a bit and it allows for the oversaturated points to dry up a bit. Then, do the cleaning wipes. If you see solution on the sensor, it's actually too wet. It should barely show any solution appearing on the IR filter itself. Most people are afraid of the swab drying out, so they drop a soaking wet swab onto the sensor and it looks like somebody mopping a floor. There's a fine line between too dry (and the fibers possibly scratching) and too wet, where the solution can't evaporate fast enough.

I get almost no streaking and the dry is very consistent and fast. Then, I'll let the sensor sit for about another 60-90 seconds to fully evaporate before "flushing" it with a puff of air, either from a filtered hand bulb or (usually) a shot of Duster All-Way.

If there is still dirt, I'll swipe again with the same swab or I'll get out the Sanyo swabs and use them to erase the streak/stain. If I need to spot clean, I'll put one drop of solution on a Sanyo swab and use it as a precision cleaning point. A dry Sanyo swab can be used to mop up any streak or solution left behind.

I have no problems using the same VD swab over and over during the cleaning process. If you blew off the sensor sufficiently before wet cleaning, there should be no grains of anything rough on the sensor - or it would've scratched on the first pass anyways. You may need to add a drop of solution strategically onto the swab if it's dried up though. Just be sure that the swab doesn't touch any surface between the last use and the next use, i.e., don't set it down on the table, no matter how clean you think it is. I'll put it back into the clear pouch the swab came in or keep it held like a cigarette in my hand until I'm sure all is well. That's where the long-arm stereo microscope comes in handy - I can hold the swab in one hand and use the other to move the scope around for inspection.

One more thing - I've seen people recommend cleaning in a slightly humid room, like a bathroom where the shower has been run. They say it's to help drop dust out of the air. Don't do it. The excessive moisture in the atmosphere will start to condense onto the sensor (or if it's a hydroscopic solution, infiltrate the remaining solution) and streak or leave blobs.

 

[Krosavcheg]

I found that the reason for streaking in most cleaning applications is that the instructions are very poor and don't give a good understanding of how to clean! The trick is to use as *little* solution as necessary to moisten the swab. In the case of the medium format 33 or 44mm swabs, 3 drops is enough - sometimes 2. I'll put one drop 1/4 of the way in from the edges and one dead center. For small format swabs, I'll only do two drop, about 1/3 of the way in from each edge.

Then, (and here's what the instructions lack) let the solution soak in and evaporate for about 15-20 seconds. I'll actually "fling" the swab like an old mercury thermometer. The centrifical force will help distribute the solution a bit and it allows for the oversaturated points to dry up a bit. Then, do the cleaning wipes. If you see solution on the sensor, it's actually too wet. It should barely show any solution appearing on the IR filter itself.

I get almost no streaking and the dry is very consistent and fast. Then, I'll let the sensor sit for about another 60-90 seconds to fully evaporate before "flushing" it with a puff of air, either from a filtered hand bulb or (usually) a shot of Duster All-Way.

If there is still dirt, I'll swipe again with the same swab or I'll get out the Sanyo swabs and use them to erase the streak/stain. If I need to spot clean, I'll put one drop of solution on a Sanyo swab and use it as a precision cleaning point. The dry Sanyo swab can be used to mop up any streak or solution left behind.

I have no problems using the same VD swab over and over during the cleaning process. If you blew off the sensor sufficiently before wet cleaning, there should be no grains of anything rough on the sensor - or it would scratch on the first pass. You may need to add a drop of solution strategically onto the swab if it's dried up though.

Excellent points! Thank you very much for this! Will try today actually.

 

[KC_2020]

I found that ......

Product formulations change, availability can be an issue and so I'm curios docholliday if this is still your method and advice for sensor cleaning ?

 

[docholliday]

Product formulations change, availability can be an issue and so I'm curios docholliday if this is still your method and advice for sensor cleaning ?

Yes it is. I've consistently cleaned hundreds of sensors (and others hundreds of times) this way over the years. Products do change, but not much nowadays. In the early days, everybody was experimenting with the "best" method. It's been fairly stable and consistent since about 2016. The only thing that's changed is I now have a nicer Zeiss scope with a longer working distance so I can have more mobility in the mirrorbox (if it's on one of my 1Dx bodies) or more swab angle if I'm doing a DB!

 

[KC_2020]

Yes it is. I've consistently cleaned hundreds of sensors (and others hundreds of times) this way over the years. Products do change, but not much nowadays. In the early days, everybody was experimenting with the "best" method. It's been fairly stable and consistent since about 2016. The only thing that's changed is I now have a nicer Zeiss scope with a longer working distance so I can have more mobility in the mirrorbox (if it's on one of my 1Dx bodies) or more swab angle if I'm doing a DB!

Thanks for the reply !

I copied and pasted your method into a text file and referenced it until it became habit back in 2021. I'm glad to hear it's still the way to go !