Sony A7R IV: Horrible Sensor Dust Problems (updated 2019-11-17)
Related: digital sensor, Glenn K, James K, James P, Jason W, sensor cleanliness, Sony, Sony A7R, Sony A7R II, Sony A7R III, Sony A7R IV, Sony mirrorless, Walter B
See also Sensor Cleanliness: how to change a lens so as to minimize dust on the sensor.
I am going nuts with dust on the sensor of the Sony A7R IV.
Just as I reported back in 2015 with the Sony A7R II, the Sony A7R IV has an anti-dust feature that is worthless. I mean worthless. As in I have never seen it remove a single particle of dust of any size.
Today, I used Sony’s clean sensor feature, which vibrates like heck for a second or so, then requires you to turn off the camera. Of all the visible dust, not a single speck budged in spite of doing so 10 times. This is not stuck-on dust as dabbing with a micro fiber cloth proves.
Other camera vendors use ultrasonic cleaning, and that works pretty well. Not Sony—it’s a total joke, a placebo checklist useless feature. Shame on Sony for such worthless tech.
I ended up using a clean micro fiber cloth to gently dab at dust on the sensor. Often that has worked for me to eliminate a large speck or two that were problems, and that was true today too, but today in exchange I got more and tinier particles, a WTF nightmare.
The Sony sensor seems to have a static charge that gloms onto dust and won’t let it go.
Anyone know how to clean a Sony sensor with dust, out in the field?
James P writes:
This seems like a bit of a trite answer but it wasn't listed on your post so I'll mention it anyway - have you tried just using a dust blower? I've rarely had to do much to maintain my Sony sensors other than use a blower on them. It seems to solve the issue for me 90% of the time. The one I use is a filtered silicone blower which I find more effective and portable than the usual Giottos rocket blower you see everywhere:
VSGO Hurricane Air Blower at Amazon @AMAZON
Perhaps where you are you get more sticky dust than me, but it seems to do the job, and I occasionally inspect my sensors under magnification too. I've owned all the A7R range including the latest IV and don't seem to have had the problems I've heard about from others like yourself. Then again, I avoid changing lenses in bad outdoor conditions when I can, and when I do change lenses it's always with the camera pointed down, quickest possible change, etc.
DIGLLOYD: silly me, I had a blower along with me and I forgot about it! Though I don’t carry it when out and about shooting, I did try the blower and it helps a great deal. Even so, one larger spec would not respond, but touching it lightly with a micro fiber cloth dislodged it, and the blower then blew it away. The blower seems like the best possible solution, since it is portable, free of risk of freezing, there is no contact with the sensor, etc. Just keep it clean in a ziploc bag.
The blower I have is the Giottos dust blower (larger is better for stronger air movement), but it sounds like it is worth trying the Hurrican Dust Blower that James mentions.
Jason W writes:
I identify with your struggle with Sony mirrorless sensor dust. It was a huge pain in the ass for the short time I owned my A7R, a problem never encountered with Nikon, Canon or Fuji.
While I didn't try all options for mirrorless sensor cleaning (such as some of the gel based dust wands) I can state that the VisibleDust swabs combined with a couple drops of their cleaning fluid eliminated dust on the sensor. You can easily take these with you in the field and is a far safer operation than a microfiber cloth and fingernails. The downside is they are very pricey and if you're maniacal about eliminating dust you will go through a lot of them.
DIGLOYD: every solvent I've used leaves streaks.
Dan Llewellyn of maxmax.com writes:
Cleaning glass is something I am quite familiar with! And it is really hard to do past a certain point. I think there is a book just on cleaning glass.
The Sony 'dust cleaning' really won't do much because they just use the image stabilization method to shake the sensor. How exactly is that supposed to remove dust? When the weather is dry and static filed, the problem will be worse.
The dust cleaning mechanism in most sensors works via a piezo strip bonded to the front piece of glass over the sensor. Usually, the glass is part of the ICF/AA and the glass has an IR cut coating. Sometimes the glass has a hydrophobic coating that helps keep the dust from sticking. The piezo strip vibrates the glass directly at a high frequency. The hope is that the vibration is enough to loosen the dust but that will only work on the larger particles.
With image sensors, the glass is very close to the sensor pixels so it doesn't take much dust to become noticeable in a picture versus a spec of dust on an external camera lens filter. While you can remove large particles of dust pretty easily, the small one get bound by electrostatic forces and get 'glued' to the glass. When I have a piece of glass on my workbench, I can see the tiny particles and using a clean room swab usually just moves the dust slightly, and tiny particles come off the swap in the process.
When I am cleaning glass, I use a sequence of procedures which end up with the glass in a custom built vacuum plasma chamber (not for publication) that burns off any organic contaminants, a thin layer of the glass and modifies the glass surface. BTW, when I put in a Class 100 clean bench in 2005, the next week Lifepixel said they had one too. I have seen people get into trouble trying to rid their cameras of dust.
If you search hard enough, even with a brand new unused camera, you can find specs of dust if you start shooting a F22 and pixel peeping. I have seen ruined ICF/AA glass where someone became obsessive and ended up scratching the glass.
That's why I recommend:
- If the camera has some specs of dust that can't be seen at the sharpest aperture (typically F8-F10), don't bother cleaning.
- If the specs aren't really noticeable, don't bother.
- Next step would be to use canned air but you have to be careful not to tip the can too much lest you spray liquid on the sensor which will leave a mark. I use something called a Sno-Gun which is mostly used in the semiconductor industry for cleaning wafers. The Sno-Gun is a simple device that hooks to a CO2 tank with a siphon tube. The Sno-Gun sprays CO2 ice crystals which are pretty good at knocking off particles and then turning in a gas. You have to use the Sno-Gun sparingly else you start freezing the substrate and getting moisture condensation. Once in a while, a used Sno-Gun turns up on Ebay. If you see one, it is probably worth getting.
- After that, I would use clean room swabs brushing in one direction. But, once you touch the sensor, you will *always* be leaving something behind. That's why I recommend to not touch the sensor unless you really have to.
- Lastly, wet cleaning. Even with lab grade 99.999% pure solvents, I will see an evaporation line. And you might wash off oils from other parts onto the sensor. I really don't like wet cleaning.
DIGLLOYD: for years I have not touched my sensors and done well enough, but when chunkies 20 pixels across start ruining all my skies, with diffuse 50-pixel faint ones too, it becomes severe headache. with skies and such. And sometimes such things overlay important detail and are very difficult to retouch or fix.
I'm not happy with wet cleaning as I’ve never been able to do so without leaving streaks.
Walter B writes:
I have found only one solution and it seems to work for all sensors, Eyelead Sensor Cleaner kit made in Germany. I have seen Leica videos showing how they manufacture their digital cameras. In the last phase prior to packing the camera, the technicians use this tool to make one final sensor cleaning. It appears to be available through Amazon, at roughly $50.00.
The sensor cleaner kit consists of a plastic stick with a slightly sticky gel pad attached at one end and a small pad of stickier disposable papers. You simply gently press the gel on the sensor and it grabs the dirt. Then transfer the dirt onto the sticky paper by pressing the gel on the paper as you would the sensor. Presto, the gel is clean again and then you repeat the process.
There are two types of sensor cleaning kits, one with a blue gel and another with an orange gel. The Orange gel is for the Sony and Leica cameras. The Blue is for all others as I read the instructions. I have thrown out all other sensor cleaners, period.
DIGLLOYD: sounds interesting.
James K writes:
Try Zerostat 3 Milty @AMAZON on Amazon.
DIGLLOYD: will look at.
Gerry K writes:
I don't have a Sony A7RIV so can't speak to that particular model's dust issues. I can say that sensor dust is universal headache and there doesn't seem to be any really useful cleaning technologies that involve liquids, brushes, and cloths.
I switched to Sony mirrorless a few years back. As an architectural photographer shooting on location - often at dirty construction sites - I found dust to be a nightmare problem that I was finally able to resolve by buying three A7RII bodies: One is permanently dedicated to a Canon 17mm TS/E, one permanently is dedicated to a Canon 24mm TS/E, and the third is mainly dedicated to a great sample of the Sony/Zeiss f4 24-70mm zoom. (I have dedicated Metabones adapters and Rogeti frames [https://www.northlight-images.co.uk/rogeti-tse-frame-update/] for each of the Canon lenses.) With this setup I do 90% of my work and only rarely have to change lenses, usually for longer teles for detail shots. The camera that is open, even briefly, inevitably accumulates dust. I find that a couple of judicious puffs of filtered compressed air (from an actual air compressor, not canned air) does trick. I haven't 'cleaned' a sensor any other way for several years now.
DIGLLOYD: in my experience canned air is partially effective. It must be oil-free and it carries risks to the sensor (freezing it). A very high grade air compressor might be a lot better if oil-free.
Glenn K writes:
I am intrigued by the disparity of reports on Sony sensor dust. Some like you and Tony Northrup have seen major problems, others report nothing out of the ordinary.
I was worried about this on my trip to Colorado earlier this fall and I did end up changing lenses in some pretty dusty situations by roads with traffic.
I did three things to mitigate the issue.
1) First, I made sure the camera was turned off for at least a while (~1 minute) before removing the lens.
2) Second, during that time I wiped off the rear of the new lens with a microfiber cloth and blew the rear element with a rocket blower.
3) Finally I did a quick blow with the rocket blower to the sensor box with the camera facing down before attaching the new lens. I ended up with no significant dust issues.
Of course, I could have just been lucky!
DIGLLOYD: care and luck can help, I suppose it is a matter of how many repetitions. Also, some lenses breath in this physical sense: focusing them forces air movement (quite pronounced with some DSLR lenses).
My approach has been to (1) turn off the camera, (2) partly loosen the lens so I can remove it quickly, (3) take the lens cap off the next lens, (3) swap quickly, (4) replace lens cap on len asap. All while shielding the camera from wind as well as I can.
I would guess that turning off the camera for a minute might dissipate a static surface charge. But when working at dusk, a minute is an eternity, where the light changes by the second—non-starter under those conditions. Plus... I regularly turn off the camera when changing lenses anyway—maybe not a minute, but there is a meaningful delay of 10-20 seconds.
Where I work in the field, this multi-step thing is just not very feasible. Frequent wind means taking the time to blow off the lens just means more crap inside the camera and the dust just returns instantly anyway. Also, physically precarious conditions and hand-numbing cold just don't make it feasible. Also, with an overstuff pack carrying a blower is itself a nuisance, plus gloves on/off, rapidly changing light, etc.
Dust varies—quite a lot of it in the bone-dry Sierra even over granite areas. At night hiking back, I can see tons of dust via my headlamp that one does not see in the day.
Humidity in the Eastern Sierra typically ranges from about 8% to 20% this time of year. I'd bet that in Colorado, it's much higher humidity in many locations. Low humidity means outrageous static electricity*. In the White Mountains, humidity might have been as low as 5%—I woke up with blood on my lips from split/cracking due to it.
* I killed another Cineo Matchbox light with static shock this trip, in spite of having grounding wire taped to it its brigtness control knob. Previously, I killed five of them, and Cineo replaced them, this one they probably won’t—fundamentally defective design and I hold them accountable for not owning up to it, though I give them credit for replacing the first batch of dead lights which I finally had to post-mortem for them—poor engineering and fault diagnosis.
