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
DaemonGPF
Post Laureate
Reged: 03/22/08
Posts: 3563
Loc: New Mexico
|
|
Keep going guys.. I have money on this thread (that it will become the longest thread ever on CN) LOL.
--------------------
-Josh
http://cleardarksky.com/c/AlbuqNMkey.html
My AP Gallery
|
Jared
Carpal Tunnel
Reged: 10/11/05
Posts: 2534
Loc: Piedmont, California, U.S.
|
|
Quote:
Keep going guys.. I have money on this thread (that it will become the longest thread ever on CN) LOL.
We're not even close, Josh, unless you add up all the other threads on the same topic...
--------------------
- Jared Willson
- TMB 152 f/8 Apochromat
- Fluorostar FLT-110 w/ TEC optics
- Stellarvue SV80S
- Astro-Physics Mach1 GTO
- Takahashi Teegul SP Mount
- STL-11000
|
freestar8n
sage
Reged: 10/12/07
Posts: 299
|
|
OK - I'll say a little more. I think a lot of f/ratio stuff is simple and derives from the fact that the irradiance seen at a detector, for nebulosity, just depends on f/ratio independent of aperture. But there are some subtleties.
It would be wrong to think that I can image m57 at f/6 and blow it up to match f/10 and magically get a big win with my c11. The real win is when you keep the pixel scale the same, and you go deep and then stretch the faint stuff to bring out details. Even if the f/6 image makes the object a bit smaller, it is often fine to make it smaller while providing more context by showing stuff around it.
Now - if both images are exposed a long time and not stretched much, they should look comparable but maybe with a loss of resolution at f/6. But if I do Ha imaging and try to get extended nebulosity to show well, and stretch it, and then blow it up to match the f/10 image - the f/6 may look better. This would be because the faint stuff may be limited by read noise and shot noise.
On the other hand, you could bin the f/10 and be somewhat equivalent to f/5 - but you would not have the same wide field.
So - this would be an interesting test. Compare a medium deep Ha image of a small nebula at f/10 and f/6 with and without binning, for the same exposure, and process each to bring out nebulosity as well as possible, and show them at the same angular scale. The stretching should be limited by noise in the faint regions, as it often is, and see which image looks best and appears to go deepest.
If the images are exposed a long time and not stretched much, they may not appear different at all, but if they are more noise limited - f/6 may still win even when its image is scaled up to match f/10.
It's worth a try. I will aim to do it one day, but not sure when. Ha is good because it would be less affected by sky glow and changing conditions would be less of a factor.
Frank
|
Arkalius
scholastic sledgehammer
Reged: 08/03/06
Posts: 878
Loc: Orange County, CA, USA
|
|
A lot has happened in this thread... but it looks like you guys are coming to an understanding and I think you're on the right track.
I want to interject a little something in here though. Signal to noise ratio is really an evaluation of the precision of a measurement. Photons don't arrive in a perfectly regular pattern, so the more you collect, the less deviation from the expected amount you have. It comes down to that. The more photons a pixel collects, the higher the SNR will be for that pixel.
SNR can't really be applied to an object by itself without considering a tool of measurement. So, because of this, we cannot say that f/ratio is the only thing that determines the SNR you'll get for an object. One has to consider how it's being measured, and since we generally use digital sensors for that, the size of the pixels being used are the important thing to consider.
There was an interesting example someone posted comparing 3 different scope SNRs... It demonstrated very well how you can't just rely on f/ratio to know how your noise performance is going to go. We had one f/10 scope give us a better SNR than another f/10 scope, because the first one used a camera with larger pixels. We also saw an f/5 scope give us the same SNR as an f/10 scope, because the faster scope had a camera with smaller pixels.
Once you've got the camera settled, then and only then does f/ratio become an authoritative determinant (ignoring things like binning).
--------------------
-Arkalius
11" Celestron SCT on Orion Atlas EQ-G
Celestron 100ED Refractor
8" Zhumell Dobsonian Reflector
|
freestar8n
sage
Reged: 10/12/07
Posts: 299
|
|
For a patch of sky representing a uniform region of an extended object, and for a single pixel of a given detector, f/ratio alone determines the SNR you get for that pixel. That is what I have been saying, and it's all pretty well defined. Not only does f/ratio determine it - but the key point is - aperture is irrelevant for this well-defined situation.
Now - that is pixel SNR for a given detector. The seperate question of how an object looks after stretching it and blowing it up to the same scale, for a given aperture, goes beyond SNR. It directly trades off pixel noise for loss of resolution - and the comparison isn't a well defined thing like SNR in the presence of various noise sources - but how it appears to a human visual system. That is more subtle and may still show a win for faster f/ratio when the background is stretched heavily to show faint stuff.
Both situations go counter to the f/ratio myth. The first is pretty solid, while the second is more subtle and would need good examples.
Frank
|
Galaxyhunter
Pooh-Bah
Reged: 01/02/06
Posts: 1251
|
|
Quote:
For a patch of sky representing a uniform region of an extended object, and for a single pixel of a given detector, f/ratio alone determines the SNR you get for that pixel. That is what I have been saying, and it's all pretty well defined. Not only does f/ratio determine it - but the key point is - aperture is irrelevant for this well-defined situation.
So bottom line is (if I understand what your saying) that if I used the same camera on two different apertures with the same f/ratio, the SNR will be EXACTLY the same. 
--------------------
Carl
My lousy skies at Hawkeye Observatory
|
freestar8n
sage
Reged: 10/12/07
Posts: 299
|
|
The SNR at each pixel is the same, and as long as you don't blow up or reduce the image size, and present all images at the same pixel scale, that is the SNR value that will matter most for depth and overall quality of the view.
In other words- at the same f/ratio but smaller aperture, the image will be smaller but just as bright. It will lose some resolution by being smaller.
Frank
|
Galaxyhunter
Pooh-Bah
Reged: 01/02/06
Posts: 1251
|
|
Quote:
In other words- at the same f/ratio but smaller aperture, the image will be smaller but just as bright.
OK, I'm not going to argue this any further. It does not seem logical that if you imaged something with a 24" f4.5 & imaged the same object through a straw @ f4.5 that it would be the same SNR. Somebody is going to have to prove that to me.
I guess that the photons for visual use is different for imaging. When I first started this hobby in '91, the talk was that it required an 8" scope to see Pluto. So for imaging, a 80mm ED is JUST as capable to record some obscure 19th mag nebula as a 16" scope can.
--------------------
Carl
My lousy skies at Hawkeye Observatory
|
freestar8n
sage
Reged: 10/12/07
Posts: 299
|
|
Pluto is effectively a point source, so aperture is critical, and f/ratio plays less of a role.
On the other hand, things like the Orion nebula and the Andromeda galaxy are visible to the naked eye as gray patches. In a telescope they get bigger and show more detail - but they are just as "bright" in terms of how many photons strike each rod or cone - now talking about visual work. For extended objects and visual work - telescopes don't make things "brighter". Whether it's the moon or a nebula - the best an extended object can do is get bigger while maintaining the same brightness. That is very surprising when people first learn this - but it is described on many web sites and in books.
All the aperture in the world won't make extended objects have greater irradiance at the retina because the f/ratio has a minimum value set by the iris.
Frank
|
Arkalius
scholastic sledgehammer
Reged: 08/03/06
Posts: 878
Loc: Orange County, CA, USA
|
|
Quote:
So bottom line is (if I understand what your saying) that if I used the same camera on two different apertures with the same f/ratio, the SNR will be EXACTLY the same.
Well... in the strictest sense, no. If the f/ratio is the same but the aperture is different, that also means the focal length is different, which means that the image will be arranged differently on the pixels of the sensor. You will take a different image, which is differently resolved, so the SNR's won't be *exactly* the same. But, for all practical purposes, they would be. The extended sources would all be the same general brightness in the image, but stars would be dimmer due to being point sources.
Why do you find it so hard to fathom? When dealing with camera lenses, the f/ratio and exposure time are what determine exposure level. You never see any mention of the aperture of a camera lense... in fact when someone talks about the aperture of a camera lense, they are actually talking about f/ratio! The point is that the size of the aperture doesn't tell you anything about how well exposed an image will be (unless you are talking about point sources).
A DSLR with a zoom lense actually can demonstrate the very situation with which you are having trouble very well actually. Take a picture with the aperture as wide as it goes with the lense at it's max focal length and take a picture. Then, zoom all the way back to the minimum focal length and take a shot at the same f/ratio. In the second case, the aperture iris will close down the actual aperture of the lense to maintain the same f/ratio. So you can compare big and small aperture with same f/ratio this way, and see that the images will have similar noise characteristics. Of course, with such an extreme shift in focal length it can be very hard to compare the two images since the image scales will be very different and scene composition will make it tough to do a useful comparison.
--------------------
-Arkalius
11" Celestron SCT on Orion Atlas EQ-G
Celestron 100ED Refractor
8" Zhumell Dobsonian Reflector
|
Jimmy2K63
Pooh-Bah
Reged: 04/26/09
Posts: 1188
Loc: Kentucky
|
|
Post deleted
--------------------
http://astronomyguy63.blogspot.com/
LXD75 SN6-UHTC
Cave Astrola 10" f/5
Garrett 15x70/FarSight
Canon XS (1000D)
Edited by Jimmy2K63 (09/24/09 08:42 PM)
|
saadabbasi
member
Reged: 08/23/09
Posts: 46
Loc: Pakistan
|
|
This thread made my head spin.
|
russ_watters
Pooh-Bah
Reged: 11/24/04
Posts: 1277
Loc: Trappe, PA
|
|
Well, I do like how it turned out, though. There can be so much bickering about minutae that people lose the forest for the trees. Most people have some specific constraints on what they are going to get, so arguing about what you can do with two cameras isn't that useful and detracts from a more real-world point (discussed just above):
I own a DSI II and a C11. Due to the DSI II being utterly unique at its price point (when I got it....I need a DSI III now), there was no other reasonable option. But I was considering a C8, a C9.25 and a C11 to go with it. The bottom line is that the DSOs I image would have the same snr (be as bright/dim, noisy/clear) in each (all are F10), just bigger in the bigger scopes. I got the C11 knowing that it wouldn't yield me brighter images than a C8 and would be harder to guide, but it would get me clearer planets 
--------------------
Equipment: Orion Atlas 11, ED80, DSI-C, DSI II Pro, Dell Inspiron Laptop.
www.russsscope.net
|
|
2 registered and 1 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: 3017
|
|
|
|
|
|
|
Previous
Index
Next
Threaded
[Re: 
Edit
Reply
Quote
