Click here if you are having trouble logging into the forums
Privacy Policy |
Please read our Terms
of Service | Signup and
Troubleshooting FAQ | Problems? PM a Red or a Green Gu.... uh, User
rg55
member
Reged: 08/02/08
Posts: 94
Loc: western US
|
|
Pardon my ignorance, but does anyone have a clear description of how and why to do these things to improve imaging? I've seen references to them, but am still in novice mode.
Thanks!
Richard
--------------------
C8 on AS-CG5 mount
Orion 127 mak
C80ED
Canon XSi
Sony HD videocam
Celestron Neximage
|
AlanT
sage
Reged: 08/20/07
Posts: 488
Loc: 122º36' W, 47º37' N, WA USA
|
|
I gave a beginning AP presentation to our local club a while back. I had a few slides in it that gives an overview of what flats and darks are, and how they're taken & used.
I captured a few slides out of the presentation and put it where you can download it from my website. If you rock the last three slides back and forth you can quickly see the effect of applying flats and darks to "calibrate" your frames.
The snippet can be had here: www.alberts-astro.com/misc/PartialPres_ImageCal.pps
Note that it's just a rough overview, and there are a lot of variations in methods. But hopefully, the slides will help you get the concept of what they do.
al
--------------------
al
Meade 80mm APO
Celestron 6" SCT
CGE, GM-8, & CG-5 GT
ST-2000xcm, DSI Pro II, DSI Pro
( www.alberts-astro.com/astro )
|
Alex McConahay
super member
Reged: 08/11/08
Posts: 130
|
|
Alan's powerpoint is a good demonstration, but here is my take on it.
When you take a picture you get "signal"--which is real photons hitting your sensor and causing them to brighten. YOu also get "noise"--which is heat, gamma rays, camera sensor defects, and other things which cause the sensor to brighten, but are not really in the scene you are imaging.
Flats and darks allow you to correct for the noise. Basically, if you know what the errors are, you can adjust the signal to correct for them. (If you had a thermomoter that consistently reads five degrees too low, you could use it to tell the temperature accurately by just adding five degrees to every measurement it takes. That is what image calibration (using flats and darks) does.)
No imaging system is perfect. There are non-random things that affect the system and change what the camera records so that it is not longer real information. Flats and darks allow you to correct for these non-random things.
Darks correct for electronic noise that shows in your pictures. On each exposure, you will get hot pixels and dead pixels(They are always on or always off, respectively, whether a photon hits or not). You will also get warmer areas of the chip. These will tend to glow brighter, even though there is no light coming from your object in these areas. Darks take a picture of these errors, and corrects for them in your light frames. To make a long story short, you take a picture with no light coming in (no signal), and the only thing the camera records is noise.
Flats correct for some defects in the optical system. These include uneven illumination in the field, dust in front of the sensor, and other such things that cause what would normally be a perfectly even flat area to have spots, shaded corners, and such. You take an image of a perfectly flatly illuminated area. And what you get will have darker corners, maybe a bright spot in the middle (or off center), and little dark round spots from dust.
When you combine these darks and flats (noise pictures) with the lights (signal plus noise) you can get calibrated frames that have corrected for the non-random errors.
(And then you combine a number of these light frames to get rid of non-random errors like gamma ray strikes, which show up as light that is not really there.)
Hope this helps.
Alex
|
Doubleglaze
sage
Reged: 11/01/07
Posts: 201
Loc: Pacific NW
|
|
Good description Alex, with one additional point I'd add or change -
Capturing and transferring an image all the way from the source in the night sky to viewing on your PC for viewing contains both "signal" and "noise". From a visual perspective, signal is good, it represents the image of the object you;d like to capture and enjoy. Noise is bad, it steals from the visual enjoyment of the image in that it doesn't look "right".
You've labeled artifacts from the electronics in the camera as noise, where I'd label them as signal + noise. Some think of it as a matter of semantics, but I'd call any constant offset or scale factor in the system a "signal", and any time varying component of the signals as "noise".
Imagine you have a camera with lots of dark current present in every shot, in every pixel. Every time you take a picture you get the same dark current from each pixel. Each pixel can have a different value (as in thermal hot spots or places where light leaks in). This is a signal that you can subtract from every image you take, if it always stays constant. It could be a terrible camera with lots of dark current, but if it were constant you could calibrate it out.
Problem is, each image is unique on a pixel by pixel basis, because the pixel intensity changes on a moment by moment basis (both the received intensity from the sky and the electronic "intensity" reported back to the camera output). Noise comes into play when you look at each of the signals that you are capturing or are built into the camera - every signal in the system also has a noise component, which really boils down to an uncertainty in what the value is for each pixel's intensity at any given moment or frame you capture.
This is why you want to take lots of dark shots to get a MasterDark of the pixel intensity signal with no sky signal. Taking one dark frame gives you a snapshot of the signal present on the camera at that moment in time, but its not constant - the pixel intensity is "value" +/- "uncertainty" every time you take a shot. What you're doing in taking and combining lots of dark frames is computing an expected value for the dark current signal for each pixel, and reducing the error, which is the "uncertainty" or the "noise". The more frames you combine, the better the estimate can be, improving by the square root of how many you combine.
That's my take - thinking about what are signals and what is noise in your camera can go a long way towards a better understanding of imaging. Took me a while to wrap my head around it, and the CN folks have gone a long way towards helping me figure it out.
Mark
--------------------
Vixen VMC260L / Sphinx SXD
Pentax 75 SDHF
Canon 40D / 50mm f1.4 / 100mm f2.8 macro / 28-135 f/3.5-5.6 / 200mm f2.8L
http://www.astrophotogallery.org/showgallery.php/cat/500/ppuser/166
|
AlanT
sage
Reged: 08/20/07
Posts: 488
Loc: 122º36' W, 47º37' N, WA USA
|
|
Alex and Mark have added good info here. I think you might see that that the subjects involved with calibration of light frames can get deep real quick, (which is why I skim over the top of it in my beginning AP presentation).
If you want to get deeper into the subject. Here's a few other resources:
- There are several sites on the web that have discussions on the subject. The Starizona site has pretty descriptive section here: http://starizona.com/acb/ccd/advimcal.aspx
- When I started out, I picked the book The Handbook of Astronomical Image Processing by Berry & Burnell. It contains several sections on Signal vs Noise covering: the various causes (static & random); the related math & statistics; relation to exposure times (re: bkgrd sky adu, & number of darks needed); processing methods; etc..... Lots of math. It's enough to make your head swim.
- I haven't read it, but I'm sure Ron Wodowski's book, The New CCD Astronomy, has a thorough treatment of the subject.
However as a practical matter, taking and applying Flats and Darks is not particularly difficult, give it a try, I think you'll find it's worth the effort. Start with shooting and applying Darks. Then learn how to take and apply Flats (they're a bit tricky, but not hard once you get your method down).
al
--------------------
al
Meade 80mm APO
Celestron 6" SCT
CGE, GM-8, & CG-5 GT
ST-2000xcm, DSI Pro II, DSI Pro
( www.alberts-astro.com/astro )
|
Doubleglaze
sage
Reged: 11/01/07
Posts: 201
Loc: Pacific NW
|
|
Good reminder Al!
When I started out imaging last year I wanted to get some of the background math and found "The Handbook of Astronomical Image Processing" by Berry & Burnell. Over the Christmas holiday I read through it and it filled me up more than the holiday turkey did. The API4WIN software is useful as well.
My biggest hurdle was understanding what is the signal part and what is the noise part, once I had that the rest fell into place. Well, that and reading the "focal ratio myth" threads which were entertaining and informative ).
Mark
--------------------
Vixen VMC260L / Sphinx SXD
Pentax 75 SDHF
Canon 40D / 50mm f1.4 / 100mm f2.8 macro / 28-135 f/3.5-5.6 / 200mm f2.8L
http://www.astrophotogallery.org/showgallery.php/cat/500/ppuser/166
|
AlanT
sage
Reged: 08/20/07
Posts: 488
Loc: 122º36' W, 47º37' N, WA USA
|
|
Quote:
...... I read through it and it filled me up more than the holiday turkey did. The API4WIN software is useful as well.
My biggest hurdle was understanding what is the signal part and what is the noise part, once I had that the rest fell into place. Well, that and reading the "focal ratio myth" threads which were entertaining and informative ).
Mark
Yeah, I found the book a tough (but good) read. It's been a good reference for me, I keep referring back to it whenever I get myself confused. I still use some of the tools in AIP4Win. I haven't been using it much lately though as I've been using an OSC camera and DSS does the calibration and stacking so easily.
I still get confused when I try to think in terms of f-ratio. It's really easy for me to get my logic turned around in it... feels like I'm in a mirror maze 
al
--------------------
al
Meade 80mm APO
Celestron 6" SCT
CGE, GM-8, & CG-5 GT
ST-2000xcm, DSI Pro II, DSI Pro
( www.alberts-astro.com/astro )
|
bebert
sage
Reged: 07/21/08
Posts: 247
|
|
These are all good summations. But for a more in depth treatment you should take a look at Graig Stark's series from earlier this year. It covers much more than just calibration frames.
http://www.cloudynights.com/item.php?item_id=1966
--------------------
Celestron C8
Orion 80 ED
Vixen 70s for guiding
CG5-ASGT
SSAG
PHD Guiding
Canon 300D(a)
Celestron 0.63 focal reducer
WO 0.8x FF/FR version II
|
|
6 registered and 4 anonymous users are browsing this forum.
Moderator: Charlie Hein, knuklhdastnmr
Print Thread
|
Forum Permissions
You cannot start new topics
You cannot reply to topics
HTML is disabled
UBBCode is enabled
|
Thread views: 214
|
|
|
|
|
|
|
Previous
Index
Next
Threaded
Edit
Reply
Quote
