Digital sensor as aneroid barometer?

[adan]

Advertisement (gone after registration)

I just thought I'd pass along an interesting experience, with speculation as to the cause.

 

I took a trip from mile-high (5,280 feet, 1600 meters elevation) Denver to sea level last weekend. Mostly shot R lenses on a Canon 5D2, but at the end of the trip pulled out the M9 with 90 Summicron.

 

To my dismay, the 90 suddenly seemed badly broken. Shots @ f/2 were grossly out of (front) focus. Set to Infinity via the lens scale, the actual focus point was at about 60 feet. All images had far more CA corner smearing than the last time I used it - think "color Holga" = and also showed generally excess corner fuzziness.

 

35 and 21 lenses seemed - emphasize "seemed" - mostly normal on quick review.

 

All in all, I was sure a lens element must have come unstuck. It was that bad. Caught my flight home, figuring I'd have to send the lens in Monday a.m. for servicing.

 

Monday am, back in Denver, gave the lens one last check - and - the problem had vanished! Focus good, corner CA and focus back to normal. Set to infinity, the lens was focused at infinity.

 

I started wondering what could have affected the lens temporarily during my trip, and came up with this theory:

 

Assume that there is a small gap or airspace between the silicon sensor and the glass cover plate/IR filter. There is a black mask around the edge of the sensor, and adhesive(?) to hold the glass in place. There are also microlenses, but I think they are integral to the silicon surface.

 

Assume that that gap is very well sealed (after all, dust getting in behind the cover glass would be a really BAD thing, requiring complete sensor disassembly for cleaning).

 

Assume the sealing is, however, somewhat permeable, so that the internal air pressure can equalize with external pressure over a long period (weeks, months).

 

Then - a change in elevation from Denver to sea level would leave the sensor "compressed" - with "Denver" pressure inside, and sea level pressure on the front of the cover glass. (see diagram) The differential would be about 2.6 pounds per square inch. Sensor area is 1.5 sq. inches, so total pressure on the 0.7mm-thick-glass of 3.9 lbs. (1.77 Kg).

 

(Of note, Denver's average air pressure is a full inch of mercury below the lowest ever recorded at sea level - 24.63 vs. 25.63 in a cyclone/hurricane - Typhoon Tip, 1979. So normal atmospheric highs and lows are much smaller changes than this elevation change)

 

My theory is that that was enough to bend the cover glass and turn it into effectively a cheapo rear-mounted close-up lens, fouling the focus of the 90 (but with less effect on my wide lenses, which is normal for CU lens attachments) and generally degrading the optical performance.

 

On my return to Denver the pressure equalized, the cover glass flattened out, and the lens returned to normal performance. Had I stayed at sea level more than 2-3 days, the pressure might also have equalized due to slow leakage.

 

If I'm right, that has implications for cover glass breaks (stress on the glass) and the occasional "odd" change in focus performance.

 

Questions:

 

Why isn't this obvious and prevalent amoung DSLRs? Because they have thicker glass and more layers (AA filter, heavier IR filter) and thus are stiffer?

 

Why haven't I or other M9 users noted a link between elevation changes and focus (if any) before? It was just taken to be "normal variation" in M focus performance? My M9 and I have made 3-4 trips to near sea level in the past year. In all cases, I have noted some focusing oddities, but just put them down to different lenses, or mistakes, and so on.

 

Why has no one noticed this with M8s, which have an even thinner cover glass (0.5mm)? Same as previous - plus, the M8 cover glass spans a shorter distance?

 

Anyway - discussion time.....

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[Shootist]

Well I think the M8 cover glass was not sealed to well. On one M8 I had it got some type of brass, shiny brass colored, specks underneath the cover glass.

[farnz]

Andy,

 

An interesting theory you propose: a digital sensor as a balloon.

 

My problem is the degree of distention that would be required by the glass to produce the effects that you'd experienced. As we know, glass is a ceramic material and therefore a solid and brittle structure that tends to crack and shatter rather than bend and deform. A polycarbonate might be able to behave in the way you've described but I have to admit to being a little dubious that glass would.

 

You got me thinking though and in 2009 I took a trip to Canada where I flew from Vancouver at sea level to Calgary at 1,083 m (3,553 ft) and then climbed higher in the Rockies, and after that I flew onto Toronto, which is 80 m (262 ft) above sea level. I had my M8 with me and didn't experience any misbehaviour from it either going up or down. Admittedly Denver is higher again but I'd have thought that Calgary would be high enough to reveal similar behaviour.

 

Pete.

[jaapv]

I'm not sure the cover glass is stuck along the edges. Isn't it a film type of adhesive? Experts?

[mjh]

As we know, glass is a ceramic material and therefore a solid and brittle structure that tends to crack and shatter rather than bend and deform.

Its elasticity depends on its thickness but glass does bend to some degree. The elasticity of glass can be used to good effect when grinding a Schmidt corrector plate to the desired aspherical shape. Quoting from Wikipedia:

 

Schmidt himself worked out a second more elegant scheme for producing the complex figure needed for the correcting plate. A thin glass disk with a perfectly polished accurate flat form was placed on a heavy metal pan. The upper edge of the pan was ground at a precise angle or bevel based on the coefficient of elasticity of the particular type of glass plate that was being used. The glass plate was sealed to the ground edge of the pan, then a vacuum pump was used to exhaust the air until a particular negative pressure had been achieved. This caused the glass plate to warp slightly. The exposed side was then ground and polished to a perfect flat. When the vacuum was released, the plate sprang back until its bottom surface was again plane, while the upper surface had the correct figure.

[pico]

I have often wondered if M9 problems were caused by air shipping, and I've expressed it here.

 

You might be onto something very important to Leica and the rest of us. Thanks for the good work.

[SJP]

Advertisement (gone after registration)

It could be the lens itself if it has cemented doublets/triplets with air in between. I would be surprised if curvature of the sensor cover plate would mess up the image to the extent you describe: (i) it is very close to the sensor, (ii) it is a flat (planparallel) sheet of glass, so curvature should not change the light path too much, it would give very slight geometric distortion but focus should remain OK.

 

We need some volonteers with a vacuum pump

[LeicaBraz]

I thought pressure in air plane cabins was compensated.

[bill]

I thought pressure in air plane cabins was compensated.

 

But not to the same extent in holds...

 

This is really interesting.

 

Regards,

 

Bill

[ho_co]

I thought pressure in air plane cabins was compensated.

 

As I understand it, Andy isn't referring to the on-board pressure, but to the difference in pressure between sea level and Denver.

 

In other words, if the idea is correct, travelers to Lhasa would find that the M9 didn't initially focus accurately; and on returning home, would find that the camera again gave fuzzy images until the pressure within the sensor again equalized.

 

 

Easy to check: Calling all M9 users!! Travel, shoot and keep records!

[bill]

travel, shoot and keep records!

 

Isn't that the Leica users' mission statement

 

Seriously, my watch has a pressure release valve; is this a variation on the same theme?

 

Regards,

 

Bill

[ho_co]

Yes.

 

Seems that Leica was so concerned with the crazies who felt the camera wouldn't withstand rain that they forgot to let it breathe!

 

 

 

Don't misunderstand. I think it's a fascinating idea that should be tested. I doubt it's accurate, but as Andy points out, the M9 is on the cutting edge of the possible.

 

 

Maybe we'll all have to learn to give our cameras CPR during travels.

 

Andreas could set up pressurization centers in major airports around the world for LUF Premium Members only.

(I mean, even if there's nothing to it, it's a great rumor to start, and a great opportunity to meet while our cameras recuperate, right? )

[LeicaBraz]

Then you should probably refer to the altitude in Solms.

Doesn't make sense to me.

[ho_co]

No.

 

If the idea is accurate, then leakage would likely allow the sensor to adjust to wherever it is over a short period of time. (Andy didn't say how long he was at sea level when he discovered the problem, apparently just a couple days.)

 

(If the idea were accurate and there were no sensor leakage, then M9s would work best only at the same elevation above sea level as the Kodak facility where the sensor is produced. That's an effect we definitely haven't seen.)

 

 

The above speculation about lens elements shifting with changes in air pressure seems off-base to me. First, the elements are much thicker in (almost?) all cases than the sensor's IR filter. Second, the volume of air in an "air lens" would be much greater than in the sensor, and less susceptible to pressure differences. Third, consumer-grade lenses (i.e., non-technical-grade) aren't designed to be air-tight. (All speculation on my part, just FWIW.)

 

 

First question: Has anyone else noticed the air-pressure-related phenomenon Andy mentions?

• For example, did someone get unexpected amounts of chromatic aberration with an M9 and 90/2 after flying from Houston to Denver?
  

 

Second question: Does anyone know whether the sensor cover plate is attached in such a way as to be more or less air-tight?

 

 

My guess is that it's more likely a lens problem; but the idea is definitely worth exploring. (You didn't have anything to drink in Florida, did you, Andy? )

[farnz]

... Seriously, my watch has a pressure release valve; is this a variation on the same theme?

No, Bill, that's for looming deadlines and slows down time.

 

Pete.

[Nicoleica]

I thought pressure in air plane cabins was compensated.

 

Not to anywhere near sea-level pressures. And these days the pressure tends to be lower than it was some years ago. Good clean pressurised air costs the airlines money, and so both the quality and the pressure drop as the airlines try to save on fuel. Another of the reasons as to why I prefer not to fly these days.

[farnz]

Not to anywhere near sea-level pressures. And these days the pressure tends to be lower than it was some years ago. Good clean pressurised air costs the airlines money, and so both the quality and the pressure drop as the airlines try to save on fuel. ...

I'll second that! The last 3 times I've flown Easyjet I've found myself gasping for air in mid-flight but I couldn't tell whether it was low pressure, low oxygen or both. I complained once to the trolley do- ... stewar ... oops, flight attendant:rolleyes: and asked her to ask the driver to increase the oxygen content but she feigned ignorance and said there was nothing she could do. I left it at that because I didn't want them to decide that I was unwell and charge me for unneeded medical assistance.

 

I feel sure that Ryanair would have turned up the oxygen content for me. For £5,000.

 

Next time perhaps I'll open a window.

 

Pete.

[SJP]

I see Howards point about lenses not bending (much) but even a few micron change in curvature or slight shifting of elements (against rubber rings) could be noticable I guess.

 

For some further info an aircraft cabin pressure is kept at between 5000 and 7000 feet equivalent altitude (pressure at 8000 feet altitude is the lowest pressure allowed by Boeing) i.e. typically somewhat lower than Denver air pressure. This would imply that pictures taken from an aircraft during flight would be off focus. This might be hard to check quantitatively but I would expect it should have been noticed.

 

The forces involved for Denver vs. sea level would be from ca. 200 hPa pressure difference, a 90 mm summicron lens has a front element of 45mm diameter so a surface of 45^2.pi/4 = 1590 mm^2, which gives a force of 31 N or 3 kg, 6 lbs roughly.

 

For a full frame sensor the force would be 17 N, 1.7 Kg, 3 lbs. Quite a lot for a thin cover slide. In fact sensor cracking territory.

 

Hope I got all the units right but this looks roughly OK at a first glance.

 

I am still brooding on whether this is getting us anywhere useful, more experimental evidence makes more sense.

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[Nicoleica]

I'll second that! The last 3 times I've flown Easyjet I've found myself gasping for air in mid-flight but I couldn't tell whether it was low pressure, low oxygen or both. I complained once to the trolley do- ... stewar ... oops, flight attendant:rolleyes: and asked her to ask the driver to increase the oxygen content but she feigned ignorance and said there was nothing she could do. I left it at that because I didn't want them to decide that I was unwell and charge me for unneeded medical assistance.

 

I feel sure that Ryanair would have turned up the oxygen content for me. For £5,000.

 

Next time perhaps I'll open a window.

 

Pete.

 

Pete. You might find these links interesting.International Travel and Health - chapter01_02 and Research examines flight cabin air quality | abc7news.com or even Airplane Cabin Air: It's Toxic

 

It's also interesting to note that air quality used to be better when smoking was allowed on aircraft.

[adan]

Just for the record, jet aircraft pressurization is excess pressure bled off the engine turbines - the same pressure that is already pushing you along at 600 mph. There is not much "cost" to it any more than there is a "cost" to redirecting excess car engine heat to warm your toes while driving in winter. The excess will be there in any case - if not used, it just gets wasted into space. "Use it or lose it," so to speak. (The "bleed" takes place before the air reaches the combustion chamber, so you aren't breathing exhaust fumes - unless something goes really haywire!)

 

FAA regulations require that "cabin altitudes" never exceed 8,000 ft (i.e. lower than Mexico City) for commercial jets certified for use in the US (which means pretty much all of them, since nobody wants to be left out of the US market).

 

There is no "extra" oxygen involved unless the masks drop. It is just regular outdoor air (78% nitrogen, 21% oxygen, 1% trace elements) compressed to breathable densities.

 

EDIT: But yes, contamination of airliner air is becoming an issue.