Canon PowerShots not fully manual?

I just discovered that my Canon PowerShot A75 is not *fully* manual
even when I operate in manual (M) mode.

When pointed at a very bright light source that overwhelms the frame,
the camera will stop down to f/8.0, even though I am trying to take
the picture at f/2.8, it won't let me; it insists on trying to expose
the picture more normally. But that's not what I'm after. I WANT the
light source to overwhelm the frame (not that it's important, but I'm
shooting a window and I want the light from it to bloom as much as
possible for effect). When I put a camera in manual mode, I really
mean it. I don't want a babysitter.

You can recreate this effect by putting the camera in manual mode,
zoomed all the way out, point it very close to a light source and try
to shoot it without having your aperture forced shut -- you can't.

The same think happens in Aperture Mode, in which I am supposed to
have full control of the aperture.

Does anyone know how I can force this camera to do what I want it to
do?

DB.

 

Send it to obedience training classes at the camera store. If that doesn't
work, use a small pipe wrench on the lens and crank it open to the 2.8 you
want.

 

Perhaps the bright light soure is too bright, i.e. even the faster
shutter (1/2000 s or so?) would still overexpose, and you may damage the
sensor. Perhaps the camera is trying to protect itself? Try a newutral
density filter if this is the case. let us know.

 

Don't quit your day job.

DB.

 

Well, I want it overexposed, that's the whole point. But what you say
is interesting. I didn't know it was possible actually to damage the
sensor with light. Can anyone confirm this? It seems doubtful to me.
My interim soolution has been to half-press the shutter on a dark
background, and then point at the light I want overexposed before
taking the pic -- this seems to sidestep the forced-iris closing, but
if this is going to damage my CCD, I won't do it anymore.

Can anyone shed any light? (no pun intended)

DB.

 

[snip]

Well, anything can be damaged by too much light. Think of a "landscape"
of the Sun (focus at infinity) where the shutter is open for a few
seconds and the lens is open all the way. the Sun will be focused on the
sensor and it will burn the sensor. You can light a fire using the
focusing the Sun onto a match or paper. Sure the sensor will be damaged
this way.

What you should do is take a picture and then analyze the shutter speed.
if the shutter is 1/2000s (or whatever the maximum shutter speed is on
your camera), then it seems this is what is going on. if on the other
hand the shutter is much slower, then I don't know what is going on. are
you sure you are in manual mode??? is there some sort of custom setting
on your camera? maybe that does something strange???

 

One more thing: check the EXIF data, make sure you are in manual mode.
perhaps the dial on your camera is broken, it appears to be in Manual
mode but the camera is actually in other mode?

 

The lens is not focused to a single point like you do when trying to
start a fire with a magnifying glasss. If there is a shutter that only
opens when you take the picture, what is sensing the image that is
displayed on the LCD before you take the picture. I don't think the sun
will hurt the sensor.

 

The shutter speed is 1/60, manually set, and does not vary from
whatever choice I make, nor does the white balance or ISO or anything
else. Only the f/stop is forced shut in response to an overblown frame
-- even in manual mode (and also in Aperture Av mode, by the way, in
which I should have total control of the iris but apparently don't in
extreme exposures).

I have checked the EXIF data. Everything is as expected. My dial isn't
broken, it switches modes perfectly as indicated on the LCD; it isn't
even loose. This A75 is less than 3 weeks old and works perfectly. I
believe this behaviour is built in to the firmware of the camera.
Anyone else with an A75 (and probably other A-series) could probably
duplicate the behaviour by setting the cam to f/2.8 in manual mode,
and pointing it directly at a light bulb in close-up: watch the f/stop
jump. As I have said, there is a work-around by half-pressing the
shutter before pointing at what I want overblown, but this is not a
'fully' manual camera. But since most people will not ever try to
massively overexpose a major part of the frame for effect, they will
never notice.

I find the idea of the Sun damaging the sensor difficult to believe
but I am still open to more evidence/argumentation along this line.

DB.

 

I've been doing more tests, and I'm finding this overexposure nanny to
be more and more intrusive. It's not just overblown frames; if I'm
shooting outdoors and trying to keep the iris open to maintain a
narrow depth of field (for interesting macro work), the camera forces
a middle exposure; it won't let me push the exposure even a little bit
for a high-key look; again unless I point at something dark or cover
the lens with my hand while I half-press the shutter (and again in
full manual 'M' mode as indicated on the LCD).

I would be interested as to whether other A-series owners can
replicate this issue; I'm willing to bet that they can.

DB.

 

[snip]

It's up to you what you believe or not. I know that my film SLR manual
says not to point the lens to the Sun and lock up the mirror since this
way I might burn the shutter curtains. I do believe in some cases (not
always) you can damage the sensor or something...

 

Too late.

 

I just tried my A80. It does give me absolute control in manual mode.
I tried shooting a desk lamp on my desk. In Av, with aperture set to
f/2.8 and at ISO 50, the evaluative metering chose an exposure of 1/400
sec. This left the fluorescent tube in the lamp itself somewhat
overexposed, but details of the surrounding reflector visible. The
rest of the frame was near black. Then I switched to manual and tried
exposures of 1/8 second and then 1 second. The camera shot both images
as instructed, even the one that's 9 stops overexposed.

Well, so far we've heard only of your camera doing this. Does Canon
mention this in the manual anywhere? (Have you read the manual
cover-to-cover? Canon is usually pretty thorough about mentioning
camera eccentricities somewhere in the manual.)

The A75 definitely has different firmware than the A80. For example,
the A75 has the ability to superimpose the shooting date or time *in
the image area*, a feature of no other Canon camera I've looked at the
manual for.

Dave

 

The sun is a danger to sensors. Focused sunlight from a f/2.8 lens (the
focal length doesn't matter; it's the f/number that determines light
intensity in the image) can burn paper or wood. Applied to a Bayer
sensor, it could fade the colour filters if they are organic dyes. It
could overheat the silicon itself to the point of damaging it, or
melting something in contact with it like the filter layer or lenslet
array. This could also be an issue with extremely bright artificial
light sources, such as a carbon arc or Xenon arc lamp (e.g. in a video
or movie projector). On the other hand, an ordinary desk lamp isn't
going to be a risk.

Trivia: semiconductors like silicon and germanium get *very* hot when
exposed to sunlight, hotter than surrounding objects. Truly black
objects absorb energy at all wavelengths, but they also radiate
longwave IR well, so incoming energy is radiated back to the world.
Fully white and transparent objects are poor radiators, but they don't
absorb much energy in the first place, so they stay relatively cool.

On the other hand, silicon and germanium look close to black at visible
light wavelengths, so they absorb visible light and near-IR efficiently
and convert it to heat. But they are pretty much transparent in the
far-IR, which makes them very poor radiators of heat. So these
materials absorb energy, get hot, can't re-radiate the heat, and thus
get very hot. Don't leave a piece of raw silicon sitting on your car
dashboard in the summer.

A Google search for "sun damage to CCD" also gives some
interesting-looking links.

Dave

 

You make a good argument for why 'prolonged' exposure to sunlight
might damage the camera, but you haven't said anything that shows
there is any danger of heat damage at the fraction-of-a-second shutter
speeds we're talking about here.

I did do some Google searching, but -- forgive me for this -- your
search terms are biased. Why? Because you're much less likely to find
the phrase 'sun damage' on a page where is an argument is made that
brief sun exposure is safe for the CCD. It's like searching for the
conclusion that I want to read. Instead I chose to search for more
neutral terms like CCD and "direct sunlight" (including quotes) and
came up with some interesting results.

First of all, there are varying opinions out there, like on anything.
But whenever an actual camera manual is directly quoted (and it they
are quoted very often on this point on USENET), it ALWAYS indicates to
avoid *prolonged* exposure to sun or *leaving* the camera pointing or
sitting in the sun. I could find no quotation by anyone that
recommends against pointing into the sun long enough to snap a single
shot. From my own manual:

"Do not aim the camera directly into the sun or at other intense light
sources which could damage your eyesight."

Nothing there about damaging the camera at all. They're just warning
you against looking into the sun with your eye.

And a little later:

"Avoid using, placing or storing the equipment in places subject to
strong sunlight or high temperatures."

Obviously they mean for a prolonged period here, since it's ludicrous
not to even *use* the camera in direct sunlight. They don't write very
precisely do they? And funny how they're always vague on the side of
overwarning you ridiculously. Lawyers.

Don't get me wrong. I am still open to this argument. But I would like
to see even one somewhat authoritative reference (not just a user's
opinion pulled out of nowhere), just one, from some manufacturer or
lab, that *specifically* states that the CCD or colour filters (not my
eye! that isn't the issue!) can be damaged by being pointed at the sun
for the short time necessary to compose and take a single shot. I have
been searching, through pages and pages of google results, and I can't
find a single such reference. And indeed, common sense dictates that
nothing could heat up significantly in such a short time, even with a
magnifying glass focusing sunlight to a point on it. Have you ever
burned anything with a magnifying glass? It takes quite a while even
focused to a point MUCH smaller than my CCD.

DB.

 

Quick follow-up (I have posted another far more detailed message
simultaneously with this one, but USENET being odd that way who knows
which will show up in the thread first)...

Here is the link to the *only* source about brief (not prolonged) CCD
exposure to direct sunlight that I have found: every other reference I
have found is just a user's opinion posted somewhere. This is an
actual PC magazine article, and it says there is no danger to the
camera in framing and shooting a shot into the sun. I don't consider
this very authoritative at all; but it's as authoritative as I can
possibly find. When I see a camera manufacturer or a professional
testing lab claim that the brief exposure from framing and snapping a
single shot of the sun can damage my camera, I will consider changing
my ways. Anyone got one? If this is really true, there *should* be
such a specific reference, not just vague near-miss camera manual
quotes about heating up the camera body or protecting your vision.
Until I see such a reference, I'll continue in my habits of keeping
the camera protected from heat and sunlight but allowing the sun to
hit the lens occasionally for the 5 to 20 seconds it takes to compose
a nice solar flare shot; these habits have served me well for quite a
while through several CCDs with no ill effects I could ever detect.

DB.

 

sigh< I have posted two other replies to this message. In the second
one, I said 'here's the link' but forgot to include it, so here it is
(I know, I'm a multi-posting doofus)...

http://www.pcworld.com/reviews/article/0,aid,115658,pg,3,00.asp

DB.

 

Do you use the LCD display for previewing the exposure? If not, then
the CCD is only exposed for a fraction of a second, and I agree there is
minimal chance of CCD damage.

But most people do use the LCD display, and that requires the shutter to
sit there open until just before the exposure.

Well, you're not going to take anything I say as an authoritative
answer, so I don't know why I bother, but: Have you ever tried burning
something using the focused sunlight from a f/2.8 camera lens? A
magnifying glass is often a higher f/number (just divide focal length by
diameter) and is a single-element uncorrected lens. A camera lens is
well-corrected and focuses the light into a nice round image of the sun.

And the image *will* be quite small. The sun covers an angle of about
0.5 degree. The wide-angle focal length of the A80's lens is about 7
mm, so the focused image of the sun is 7 * tan(0.5) = 0.06 mm or about
2.4 thousandths of an inch in diameter. The A70/A75 have a smaller CCD
and even shorter FL lens. If the camera is stationary (on a tripod) the
same pixels get that energy on a continuing basis while you compose the
shot.

Let me suggest an experiment: Take any f/2.8 camera lens, or a larger
aperture stopped down to f/2.8. A 50 mm FL lens collects a lot more
total energy at f/2.8 than a 7 mm lens, but both produce an image with
the same number of watts per square mm - the longer FL lens just covers
more area with the image. Using the lens, focus an image of the sun on
your finger. How long can you do this for? Now reflect that your
finger isn't as efficient at absorbing light as the CCD, so it won't get
as hot.

Dave

 

Good point about the LCD preview ... I hadn't thought of that.

Well, I'll certainly listen to an argument or chain of reasoning based
upon mutually agreeable facts (see above), but I usually won't just
accept untested a plain assertion unsupported by reference to some
creditable source that something is true or isn't true, from a
stranger. No offence, and I certainly don't think your opinion is
worthless. It's just the nature of essentially anonymous conversation.

I see where you're going with this reasoning, and maybe I am wrong and
a camera lens does produce a stronger effect than a magnifying glass
.... but your logic is a little bit odd. The angle of the sky covered
by the sun can't really be a factor. If I zoomed in on the sun making
it take up the whole frame, by your logic it would be less dangerous.
I find that odd. Surely the issue is not the size of the sun but the
amount of convergence (and therefore concentration) that is produced
in its rays on any particular substance; that is what will produce
rapid heating. It's a plain uncontestable fact that a camera will not
produce as much convergence on the film plane or CCD as a magnifying
glass focused to a point on something, because if it did it would be
useless in capturing an image (every image would consist of a single
point!)

Well ... this experiment can be misleading ... again, if I am focusing
all of the light-gathering power of the lens upon a single point, then
I am concentrating a lot more heat than a film lens does on a the film
plane. I don't have a separate lens to try it with, but I'm not sure
of the value of this experiment for the reasons I've stated. I'll
stipulate that a camera lens may transfer more the sun's heat than a
magnifying glass, but keep in mind that the fastest lens is no lens at
all. Remove the glass or the lens and you have the widest f-stop
possible, with all of the sunlight hitting the material. The real
issue is how tightly the rays are converging. Maybe for a 1/3 of an
inch CCD they are converging tightly enough to be an issue, especially
along with your excellent point about the LCD preview being on meaning
the light is in fact getting to the sensor.

But I'd still like to see some manufacturer just come right out and
say it, instead of vaguely inapplicable manual warnings about my eyes.

DB.

 

No, my point was that the light intensity at the surface is determined
by the f/number. If you have a constant-aperture zoom, the brightness
of the image is the same at any focal length. On the other hand, with
the more typical variable-aperture zoom, brightness goes down as you
zoom in. (This is all assuming the camera's on a tripod. If it's
handheld, the wideangle lens is probably less dangerous simply because
it produces a small image that you can't keep aimed at the same spot all
the time.)

I'll contest that! You seem to be under the impression that a
magnifying glass somehow magically focuses the light from the sun to a
point. It does not. No optical device can do that, since the sun isn't
a point source. A magnifying glass focuses the sun's light to a circle
that is an image of the sun, with the diameter of the circle
proportional to the focal length of the magnifying glass. Most
magnifiers have focal length somewhere between 50 and 250 mm. A 50 mm
camera lens will produce exactly the same size lens as a 50 mm hand
magnifier (which would be labelled 5X).

The very best a lens can do is focus all the light from one point on the
sun to one point on the image of the sun, and a well-corrected camera
lens comes very close to that idea. A magnifying glass has poorer
optics, gives a fuzzier image, and spreads the same energy over a
slightly larger area - even when the focal length is the same. The only
time a magnifying glass, or a lens, focuses light to anything
approaching a point is when the light comes from a point source such as
a distant star (not the sun), or the coherent light from a laser.

No lens can focus light from the sun on a single point.

But with no convergence at all. This corresponds to an infinite
focal length and infinite f/number. Any positive lens will do better.

Which is determined entirely by the f/number of the lens!

CCD size doesn't matter. DSLRs are less vulnerable simply because they
don't have LCD preview.

Did you see this movie, which someone else posted recently?
http://ireid.home.cern.ch/ireid/images/JuraSunset.wmv
Note the effect after the sun passes a point on the CCD.

Dave