15-sec exposures of night sky yields too many stars?

Hello, all

I mounted my Canon A40 on a tripod and aimed it at the South Florida night
sky (west Kendall, edge of Everglades). I took many 15 sec exposures at
different angles and orientations (ISO 50, 2 sec self-timer to minimize
shake). I was amazed at the results until I realized that fences and walls
also had "stars" on them. Any ideas as to what happened? Did the camera pick
up actual light, or is this what they call "noise"? Thanks for your help.

Juan

 

Definitely noise. Try shorter exposures and you'll see fewer "stars." I'm
not sure if 15 seconds is long enough to see star movement but if it is you
will be able to see which are stars and which are noise.

 

I've noticed similar "dots" when doing long exposures at night with my Kodak
DX6490. The dark side of the moon had a star in front of it!

--
--
Ben Thomas - Software Engineer - UNICO Computer Systems
Melbourne, Australia

Opinions, conclusions, and other information in this message that do not
relate to the official business of my employer, UNICO Computer Systems,
shall be understood as neither given nor endorsed by it.

 

That's "noise" and it's why astronomers use special cooled CCDs. A CCD has
to be cooled below freezing to get the noise down to levels acceptable for
astronomy. However, you may be able to get around the problem, partly, by
subtracting a "dark frame" (an identical exposure with the lenscap on) from
the image.

 

Yep, noise.

 

Possibly noise, but seems less likely at ISO 50. If they're all in the same
place from frame to frame, they might be stuck pixels.

 

Ben,
check your thread about "landscape mode" where I responded to that
question.

As for your question about possible noise with your Canon A40 set to ISO
50 & 15 sec. exposures at the night sky, without seeing your pictures I
can't say what it might be but is does not sound like noise to me.

The A40 automatically applies "dark frame subtraction" noise reduction
to any shutter length of 1.3 sec. or longer. I have taken hundreds of 5-15
sec. long exposures @ ISO 50 both indoors & outdoors & have yet to see
anything close to objectionable or even noticeable noise.

If you want to see what CCD noise on your A40 looks like, set the ISO to
400 & take a 1 sec. exposure of the night sky or anything that won't result
in an over exposure. The noise will be there because of the high ISO
setting & the fact that the noise reduction will not be applied at this
shutter speed. It's possible that you may not have manually set your flash
to "off", if the flash fired during this long exposure it very likely
illuminated dust, lint or even pollen that may have been floating by.

I know people are probably thinking the night air is clean but you need
only record video with an near infrared sensitive camera & a near infrared
LED cluster illuminator & you will be amazed at how much "tiny little stuff"
floats by & is illuminated so clearly. If you see very little, stand in
from of the camera & tap on your shirt or jacket & lots of bright tiny
specks of dust & lint will be released into the air & float by the camera.
It sure amazed me the first time I noticed it, I though it looked like
random snow flakes floating by but I knew this could not be since it was an
80 degree august night.

It's a strange world out there & our human eyes only see a tiny spectrum
of what's out there. Insects & nocturnal animals see higher & lower
frequencies than we do so things look different to them & that's also why
they can see at night without a problem. Digital CCD & CMOS sensors can
also see more than we can with our eyes especially in the near infrared so
that's why they all have infrared block filters built into them or it would
mess-up the color rendition from what we see & consider "normal".

Best of luck with your A40 it's a great little camera. If you really
want to see what it's capable of in existing light situations, take it with
you to a restaurant. Set the "ISO to 50", white balance to "tungsten",
photo effect to "vivid", "flash off" & use the 2 sec. self timer. Sit it
near the end of the table or across the top of a glass, aim it in the
general direction you want with it on "full wide angle" & press the shutter.
Start with a 5 sec. exposure & work up from there or down if needed.
Anything that moves will be strongly blurred but the ambiance of the place
will be clearly captured. To be less distracting to others I turn off the
camera sounds & even place a peace of black electrical tape over the flash
so that the self timer flashing is not seen.

Enjoy, it's a great little camera as is the A60 which replaced it. My
A40 has over 3,600 pictures on it & it's going to continue to be used until
1 of us stops working!

Respectfully, DHB

 

Thank you for your thoughtful reply, I posted one of the shots in question
at http://www.photosig.com/go/photos/view?id=1091875

Unfortunately I can only post once every 72 hours, I would have liked to
post a couple more.

I tend to go with the noise theory because any dust specks would have to
stand still for a couple of seconds at least to create an impression. The
other possibility is stray headlight reflections from the walls, but that
doesn't explain the specks on the inside side of a 6 foot fence.

Juan

 

It is noise..

This is normal with digital cameras. It happens with digital SLR's as well..
except it takes about 90 seconds before the dots start showing up.

 

Now that I have seen 1 of your pictures, I must admit that it does look
like random CCD generated noise. Am a bit surprised because mine produces
such clean results but then again I am shooting in low light situations &
not "almost no light" situations as your night sky shot basically was.

Now you have my curiosity up & I am going to have to try to reproduce
your picture with my A40 just to see if mine exhibits the same or similar
noise artifacts. Like it or not I am just 1 of those people that don't like
unsolved mysteries, so now I will have to try to recreate similar conditions
& try a few shots with my A40 in order to satisfy my curiosity! Yes I'll
let you know what mine did & may do likewise with my A70 just for any added
insight that it might yield.

The floating dust or etc. was only a possibility if you had forgotten to
turn off the flash because if it fired it would have illuminated anything
floating within about 15 ft or less of the camera even at ISO 50. But if
the flash did not fire than that suggested possibility has been ruled out.

Do hope you try what I suggested in a restaurant or even a bar/lounge
because the effects are often quite interesting & surprisingly clear with
deeply rich colors.

Those lead weights are tugging at my eye lids so I had best let them
have there way & get some sleep.

Respectfully, DHB

 

Not so--you can earn more posting slots by making critiques that other
people consider helpful.

________________________________________________________________________
Louise Bremner (log at gol dot com)
If you want a reply by e-mail, don't write to my Yahoo address!

 

Digital noise is *very* temperature-dependent and you will notice a
difference even between 70 F and 40 F.


--
Clear skies,

Michael Covington -- www.covingtoninnovations.com
Author, Astrophotography for the Amateur
and (new) How to Use a Computerized Telescope

 

On my DSLR (10D) with the lens cap on, I got 12 "dots" in 164 seconds at ISO
1600. 12 out 6,291,456!

I did this to do a blackframe subtraction since I had taken a few shots of
the Northern Lights at 1600 before I realized it and set it back to 100.

 

I mounted my Canon A40 on a tripod and aimed it at the South Florida

On my Canon A80 at its max of 15 seconds at ISO 400 I get no "dots."

 

You'll find that you get more white pixels with the lens cap off. These aren't
stuck pixels, but more accurately, they're saturated pixels.

The 'saturated' pixels in long exposures are sensor locations that can't bleed
off the charge created by light falling on them.

With no light, (lens cap on), the charge doesn't occur. With the cap off, even
in low light, there's enough to saturate some of the sensors and cause white
dots.

Try a 5 minute shot with lens cap on, then another 5 minute shot under low
light using ISO100 at around f/22. You'll see many more 'hot' pixels on that
image.

 

Maybe we should reverse the order to eliminate the CCD temperature factor. A
cooler CCD will have less noise.

Juan

 

In message <[email protected]_s03>

A quick guide to 'star trails', or rather avoiding them, assuming the
camera is fixed -

With a 135mm lens use a maximum of approximately 5 seconds exposure.

Use proportionately longer times for shorter focal lengths, shorter
for longer focal lengths.

This is something you can do in your head.

A 50mm lens is approx 1/3rd of 135mm, so you can use 15 seconds.
A 500mm lens is approx 3x a 135mm, so you can use 2 seconds.

etc.

Barry

 

It's now 3:30am where I am in MA USA & 17.3 deg. F. Just came back in
from taking 10, 15sec. pictures of the sky & I froze my A** off because my
body has not yet adjusted to the cold.

Anyway the 1st 4 were taken with the camera set to ISO 50, 15 sec., f8,
WB=tungsten, Picture effect=vivid, self timer set to 2 sec. Took 2 shots in
a row without moving the camera in order to see if what looked like stars
might have been noise or hot pixels. Now I am a bit perplexed because the
only dots I noticed in all for pictures was 1 bright star & nothing that
looked like your picture.

The next 6 pictures I did exactly the same thing but changed the f stop
to 2.8 (wide angel wide open) & repositioned the camera lower in the sky in
order to have part of my neighbor's house in the picture too. Got more
stars, a little area of white clouds & may neighbor's house was dimly
visible against the sky. However not a single picture showed a single
unexplained or expected dot of light.

Either I got real lucky & got a better than average CCD or I am doing or
setting something different than you are. As for the "tungsten" WB & the
vivid modes, I simply forgot to change them from the settings I usually use
when taking existing light pictures both indoors & outdoors. Probably
should try resetting the WB to Sunny or cloudy & drop the vivid photo effect
mode to see it that makes any difference.

The very cold night may have reduced noise so much as to be undetectable
but if that was true than I should have seen it at least on the 1st 2 shoots
as I took them as soon as I got outside before the camera/CCD could have
cooled down much. So now it all seems like more of a mystery so I'll have
to do it again at a later date & if I still see nothing I'll try it with my
other digital cameras also.

Sorry if this did not help much but I try again another night.

Right now it's well past my bedtime, so before I fall asleep with my
head on the keyboard, I best call it a night.

Respectfully, DHB

 

Big difference from the mid 70's temperature when my pictures were taken!
I'm also out on the edge of town where the skies are a bit clearer than
downtown,

Can't you see any stars with the naked eye? Down here you can, what I do is
point the camera to a cluster. The sky was clear when I took my shots.

I'm not sure if the exif data carried on to the picture I posted on
Photosig, but everything was set to normal except for the long exposure. I
did need to re-size the picture fpr Photosig, but no other adjustments were
made.

It looks like I need to wait for a cold night myself. Thanks for the
experiment.

Juan