Astrophotography w/ Digital SLR Sensor Sensitivity
Billy
and I'm amazed - when I take the picture I can many more stars in the
image than I can with the naked eye. Does anyone know why this would
happen, could the image sensor be stronger than my eyesight?????
Marvin
The sensors in a digicam integrate the light that is
received over the length of an exposure. Your eye-brain
sensor isn't as good about integration.
The eye has a large dynamic intensity range in part because
it changes the lens opening and has other mechanisms for
adjusting sensitivity over time. Are your eyes fully
dark-adapted when you look up at the sky?
Don Stauffer
Unfortunately, you cannot "adjust" the exposure time with the human eye.
It turns out the quantum efficiency of the two are pretty comparable,
but you can make a long exposure with the camera, so it is collecting
more photons than the eye.
Billy
amazing, didnt know this!
Pat
I didn't know you collected photons. I have some -- including some
rare ones -- that I was thinking about selling. I boxed up some
lovely purple ones in 1972, if you are interesting. Plus I have the
complete set of Kennedy Commemorative ones.
Jeff R.
This isn't just a "digital" thing - exactly the same thing happens
(fortunately!) with film.
I have only ever seen (visually - not photographically) the Coal Sack nebula
a couple of times - these when I've been way out bush with carefully
dark-adapted eyes - yet I have managed to photograph it (film) from my
heavily light-polluted Sydney suburban backyard.
It takes exposures of only a couple of seconds to make naked-eye-invisible
stars appear on film.
One of the reasons I enjoy astrophotography so much.
Hint: in order to increase your astro fun, google: "barn door mounts"
--
Jeff R.
Ray Fischer
You can't set your eyes to take a 30-second exposure.
And, as a related bit of trivia: one of the longest exposures ever
taken is the 1,000,000 second Hubble deep-field photo.
http://hubblesite.org/newscenter/archive/releases/2004/07/
--
Ray Fischer
rfis...@sonic.net
Billy
> Billy wrote:
> > I recently began taking images of stars from my Nikon D300 digial SLR
> > and I'm amazed - when I take the picture I can many more stars in the
> > image than I can with the naked eye. Does anyone know why this would
> > happen, could the image sensor be stronger than my eyesight?????
>
> (fortunately!) with film.
>
> I have only ever seen (visually - not photographically) the Coal Sack nebula
> a couple of times - these when I've been way out bush with carefully
> dark-adapted eyes - yet I have managed to photograph it (film) from my
> heavily light-polluted Sydney suburban backyard.
>
> It takes exposures of only a couple of seconds to make naked-eye-invisible
> stars appear on film.
>
> One of the reasons I enjoy astrophotography so much.
>
> Hint: in order to increase your astro fun, google: "barn door mounts"
>
> --
> Jeff R.
Yea, I'm do a one-one thousand, two one-thousand ........up to four
one-thousand using "bulb" and the shutter depressed only 20 to 30
miles outside of the NYC light pollution. I just never look up NE,
which is where the lights from NYC destroy the sky. The amazing images
I took were last night during a chilly brisk clear winter night
pointing up to the S-SW in a dark room with the window open. I could
see maybe 20 stars with my eye, whereas I must have about 100 or more
in the image. I wonder if the other dots were dust on the window
screen. :)
T.Baxter
Most digital cameras can easily resolve stars as dim as magnitude 9 at ISO 200
or 400 at widest aperture when given a 30-60 second exposure, and when skies are
dark and clear. P&S or DSLR, most all of them can do this. Your eye can
generally only see stars as dim as magnitude 5.0 to 6.5 on a good night, the
higher the number the dimmer the star. A star that is 1 magnitude higher in
number is 2.5 times dimmer. Though on a good night of "seeing" (thin clear cold
atmosphere), from a dark-sky site I have managed to see stars as dim as
magnitude 7.6 with my naked-eye. Trying to find the constellations on a night
like that it then becomes almost impossible to pick out the main constellation
stars from so many. What a grand sight though if you are ever in a place and
time where skies are so clear, dark, and steady to get to enjoy it. I was able
to read the print on some papers by starlight alone that night. The "Milky Way"
directly above providing most of the light that time. It appearing to my
naked-eye as you often see it in long-exposure photographs. Staring into the
"Milky Way" I felt as though I was going to fall into the heart of it, the only
thing holding me back from the fall was the tenuous gravity of earth fighting
with the greater gravity of our galaxy.
To give you an idea of how many more stars are visible to your camera compared
to your eye, there are 9,096 stars visible from earth down to magnitude 6.5 and
there are 299,485 stars visible from earth down to magnitude 9.0. (~19 million
stars when you go as dim as magnitude 15.0, and ~526 million stars down to
magnitude 20.0.)
If you are in the northern hemisphere a common urbanite's dim-star eyesight and
atmosphere quality test is trying to find or count the seven main stars in "The
Little Dipper" (Ursa Minor). On a fairly decent night away from most
light-pollution (or strong moonlight) you will easily see all seven. The dimmest
of the seven stars being magnitude 4.95. The two brightest stars are near
magnitude 2.0, one of them is near magnitude 3.0, three of them near magnitude
4.3, and one near magnitude 5. This ability to see dimmer stars also depends on
how dark-adapted your eyes are and how dark your skies are. If in an urban area
the sky-glow from light-pollution will quickly overwhelm the fainter stars, to
your eyes, as well as your camera no matter how long you set the exposure. The
sky-glow being brighter than the fainter stars.
You would have been able to resolve even more faint stars if you had not pointed
your camera through an open window with that much temperature difference between
inside and outside. The large exchange of cold outside air with warm inside air
creates a lot of atmospheric turbulence between your camera and the stars (or
the converse in summer, cold dry inside air with warm humid outside air). This
destroys the "seeing" (steadiness) of those small points of light. Smearing the
dimmer stars' light into larger areas during a longer exposure, the camera
unable to resolve them. Small telescopes should never be used through an open
window unless the inside and outside air are equal in temperature and humidity
and there is no air-flow through the window.
You should have also not shot through a window-screen. This acts as a diffuser
to those points of light. Spreading out the stars' images into 4-point
diffraction spikes, just as a "Star-Filter" does on the front of your lens to
give radiating spikes to point-sources of light. This causes further loss of dim
starlight being spread into the darkness around them.
whisky-dave
"T.Baxter" <tba...@spamfree.net> wrote in message
news:2erqn41ssiptod16i...@4ax.com...
>
> Most digital cameras can easily resolve stars as dim as magnitude 9 at ISO
> 200
> or 400 at widest aperture when given a 30-60 second exposure, and when
> skies are
> dark and clear. P&S or DSLR, most all of them can do this.
I used to take pictures of the nights sky, using HP3 & HP4 film....
A problem I first had was is that dust on the neg, print etc. or a star .
If you know your stars it's easier, but how do you tell noise from stars ?
especailly if you set the speed to 400/800 ISO which is the sort of film
speeds I
was setting when using film.
I'd like to do astrophotography again some day......
>
T.Baxter
You can do frame-stacking with many shorter exposures to eliminate all
noise. Or frame-stacking with many longer exposures to eliminate all noise
and also increase the number of stars captured to greatly dimmer
magnitudes. The number of frames needed to eliminate all noise will be
dependent on what ISO and noise level is in your particular camera. A P&S
camera can easily surpass a few frames from a DSLR sensor when using lower
ISOs with longer exposures and many more frames. Noise-free exposures of
many hours can be obtained this way. "Registax" is very capable freeware
for this purpose. http://www.astronomie.be/registax/ There are others.
In lieu of that, with a single noisy frame, then you use a histogram
adjustment to cut off all dim pixels at a known dark limit. You will lose
some valid dim stars with this method but at least you won't be wrongly
displaying noise as stars. Do this while having displayed alongside a same
area sky-chart, using some excellent star-charting software like "Cartes du
Ciel v3". Probably one of the best freeware programs you'll ever find.
http://www.ap-i.net/skychart/index.php When using the free "SKY2000 Master
Star Catalog" data you will display stars to magnitude 9, which should be
enough for most digital cameras when using single-frame images. If you use
frame-stacking methods then get the additional "Tycho 2" star catalogs of
2.5 million stars of magnitude 9 to 12, and the "HST Guide-star Catalog" of
19 million stars to magnitude 15. You can obtain these from the older (and
in some ways, better) Windows-only version of "Cartes du Ciel v2.76" at
http://www.stargazing.net/astropc/download.html There are many free star
catalogs available for this software. You can also get the free CD of the
"USNO-SA 2.0 Astrometric Reference Catalog" with stars from magnitude 16 to
19, another 15 million stars,
http://tdc-www.harvard.edu/software/catalogs/usnosa2.html (Back-patting
note: I compiled the first Tycho star-catalogs for this program.)
[Note that the "HST Guide-star Catalog" is not 100% accurate, as some of
the guide-stars listed in it for The Hubble Space Telescope are dust and
artifacts from the original photographic plates that were used at the time
of it being compiled. Even top-level science can suffer from the same
problems that you'll face with your cameras. They've since revised that
star catalog to remove most blatant errors but some might still remain.
Don't feel too bad if your image has a dim star that might be noise, even
scientists won't blame you too much. They occasionally do it themselves on
far more important projects, until someone finds the error and bothers to
correct it.]
View the "Cartes du Ciel" star-chart with stars, nebulae, and galaxies
displayed to a magnitude limit dimmer than your suspected dimmest stars in
your photograph. Adjust your single-frame photograph's histogram to cut off
all warm or noisy pixels that don't show known objects in the "Cartes du
Ciel" star-chart. If the dimmest pixels in your image do not match the
positions of a known star or galaxy in the chart then raise the
photograph's histogram adjustment dark limit another notch until all that's
left in your photo are known objects shown in the sky-chart. You can then
use brightness & contrast or curves tool (better to retain relative
brightnesses) to bring up those left-over valid star pixels to an
acceptable display brightness. Much too will depend on your intended use
for the photograph. If you want a denser star-field for artistic effect
then what's the harm of leaving in some faint noise as stars? Unless one of
your friends is an astronomer with a photographic memory nobody will be the
wiser nor care. It might make for a better photograph for your intended
purpose. That's up to you.
Printing night-sky photographs is a whole other matter. The ink-bleed from
most inkjet printers will inevitably fill in some of the dimmest stars in
the field making them disappear when printed. You might have to increase
the star sizes by using blurring methods and/or increase their brightness
greatly to try to retain them in prints. There's no cut and dried method to
accomplish this. You'll have to experiment with your particular printer,
inks, and papers; as well as the amount of ink being laid down and your
printer's dpi settings. You might even find dim stars disappear over a
period of days in your prints as the ink pigments slowly migrate to fill in
the lighter areas. Quite annoying to find this happen after you've spent
hours adjusting and printing things, thinking you finally got it right.
You might want to also get the "PGC/LEDA 2008 Galaxy Catalog", of another
1.7 million galaxies from http://x.astrogeek.org/software/cdc/catalog.php
for "Cartes du Ciel" so you aren't cutting off noise which might be valid
galaxies. Throw in a few of the nebulae catalogs too. Star and object
catalogs for "Cartes du Ciel" are scattered all over the net. You might
have to hunt some down. There's some more here http://www.schoenball.de/
and here http://www.astroclub.biz/index.php?par1=11&par2=1&lang=rus for
starters.
I just found this "Cartes du Ciel v3" page listing the principle (but some
being older) catalogs
http://www.ap-i.net/skychart/en/documentation/installation_of_extra_catalogs
They are now hosting them here
http://sourceforge.net/project/showfiles.php?group_id=64092&package_id=208104&release_id=455946
whisky-dave
cheers for all that, will be most useful when I get around ito it. :)
So Stored and archived.
Now if I just didn't live in London
Twibil
> I have managed to see stars as dim as
> magnitude 7.6 with my naked-eye. Trying to find the constellations on a night
> like that it then becomes almost impossible to pick out the main constellation
> stars from so many. What a grand sight though if you are ever in a place and
> time where skies are so clear, dark, and steady to get to enjoy it.
Yes. I had an epiphany of that sort one night while camping out on the
11,500' summit of Mt San Gorgonio here in southern California. Woke up
at circa 3:00 AM on a fall night needing to tap a kidney, and happened
to look upwards as I was doing so.
The Andromeda Galaxy -normally a faint smear of blue to even dark-
adapted human eyes- was an obviously oval swath of light twice the
size of a full moon! I had to look twice to be sure it *was*
Andromeda, as there were so many stars it was difficult to pick out
the constellations.
I've never seen it that way before or since, but the sight burned
itself permanantly into my memory as few other things have.
~Pete