Return to the Cloudy Nights Telescope Reviews home page


Astrophotography and Sketching >> CCD Imaging & Processing

Pages: 1 | 2 | 3 | (show all)
Orion64
professor emeritus


Reged: 11/29/09

Loc: Western Cape, South Africa
Why do I want to image at a faster f ? new
      #5475465 - 10/17/12 05:02 PM

Okay I know that I am probably going to stir up some old threads, but is there anyone who can just tell me in one word (yes or no) whether I would actually get the same s/n for a shorter exposure on a shorter f ratio telescope than for a longer f ratio on the same scope and same ccd camera.
I have read so many discussions on this and still cannot get to the verdict.
PS: Ignore all point sources (stars) in this thread.


Post Extras: Print Post   Remind Me!   Notify Moderator  
jmasin
Carpal Tunnel


Reged: 12/22/08

Loc: Murphy, TX (DFW)
Re: Why do I want to image at a faster f ? new [Re: Orion64]
      #5475696 - 10/17/12 07:05 PM

My understanding is same S/N, but with less time. It's just a more absorbent sponge, you'll soak up more of what you want, but also more of the sky noise (shot, LP etc.) also.

Post Extras: Print Post   Remind Me!   Notify Moderator  
jerryyyyy
scholastic sledgehammer


Reged: 10/06/11

Loc: Stanford, California
Re: Why do I want to image at a faster f ? new [Re: jmasin]
      #5475835 - 10/17/12 08:40 PM

My dumb answer is the longer you take the more "dark current" or noise in the system, so faster the better....

Post Extras: Print Post   Remind Me!   Notify Moderator  
Rick J
Post Laureate
*****

Reged: 03/01/08

Loc: Mantrap Lake, MN
Re: Why do I want to image at a faster f ? new [Re: jerryyyyy]
      #5475995 - 10/17/12 10:24 PM

No I can't It is far too complicated for a yes/no answer.

You can't talk about f ratio alone. It works in conjunction with aperture, pixel size, chip size, focal length, camera electrical characteristics etc. Change focal length alone and you get one result. Change aperture alone you get another result, change pixel size alone and yet another result etc. Change two or more and things get complicated fast!

Keep in mind pixel size is a very important part of this. A 6" f/8 scope with an 8 micron camera will be identical in all respects to a 6" f/4 scope with a similar camera that has the same number of 4 micron pixels. No difference. Sub exposure time will be the same. Total time will be the same etc.

If you keep everything but focal length the same then the shorter focal length will need shorter subs. At f/4 they would be one forth the time needed at f/8 to be photon (sky) limited. But for the same signal to noise ratio of the image the total time governs not focal length. Though the shorter focal length will hide some of the noise due to lower resolution of both noise and image detail. This loss of resolution is at the gain of field of view. This may or may not be a worthwhile trade depending on what you are trying to accomplish. Faster subs helps with poorer polar alignment and lesser mounts. This too may be worth the trade of lower resolution.You can't talk about f ratio alone. It works in conjunction with aperture, pixel size, chip size, focal length, camera electrical characteristics etc. Change focal length alone and you get one result. Change aperture alone you get another result, change pixel size alone and yet another result etc. So with a shorter focal length you need more but shorter subs to get the same total time. But a low read noise camera can similarly reduce the sub exposure time needed to be sky limited gaining both shorter subs and no loss of resolution.

Oh yes, since stars cover several pixels they react the same as extended objects as f ratio changes may or may not change the number of pixels their image is spread over, same as extended objects.

In many cases, not all, fast systems cover a wider field of view at lower resolution than slow systems. You gain field and lose resolution. It won't necessarily speed up total image time, stars or extended objects for the same signal to noise ratio. Aperture is usually the main controlling factor for exposure time, this is why observatories are building larger and larger mirrors, not necessarily faster ones. Large pro scopes tend to be faster because of mechanical issues as these monsters get longer.

No matter the f ratio or most other factors the more total time used to create an image the better the signal to noise ratio. So more time is always of benefit but you do reach a point of little gain. 4 minutes of total time has half the noise of 1 minute, 16 minutes halve it again, as does 64 minutes, then 4 hours, then 16 hours then 64 hours but going to 65 hours has no significant change, for that matter going from 16 to 17 had a lot bigger change but it too would likely not be visible. Only an imager can say when enough is enough.

Rick


Post Extras: Print Post   Remind Me!   Notify Moderator  
jmasin
Carpal Tunnel


Reged: 12/22/08

Loc: Murphy, TX (DFW)
Re: Why do I want to image at a faster f ? new [Re: Rick J]
      #5476146 - 10/18/12 12:29 AM

Quote:

If you keep everything but focal length the same then the shorter focal length will need shorter subs. At f/4 they would be one forth the time needed at f/8 to be photon (sky) limited. But for the same signal to noise ratio of the image the total time governs not focal length.




This is what I was trying to say but Rick said it far better.

For these "what's different about A and B" type threads, I assume we are talking "all else equal".


Post Extras: Print Post   Remind Me!   Notify Moderator  
Orion64
professor emeritus


Reged: 11/29/09

Loc: Western Cape, South Africa
Re: Why do I want to image at a faster f ? [Re: jmasin]
      #5476300 - 10/18/12 03:18 AM

Okay, so if I understand correctly. By keeping everything constant except the focal length, I will get twice better S/N on the f/2.8 than on the f/4?
So my question is actually: Do focal reducers give us less total imaging time needed or not? Otherwise it is only going to let us lose resolution and no gain in any other area ( except a wider field, which can be done with a mosaic anyway)


Post Extras: Print Post   Remind Me!   Notify Moderator  
orion69
professor emeritus


Reged: 05/09/10

Loc: Croatia
Re: Why do I want to image at a faster f ? new [Re: Rick J]
      #5476414 - 10/18/12 06:23 AM

Quote:

Aperture is usually the main controlling factor for exposure time, this is why observatories are building larger and larger mirrors, not necessarily faster ones.





Are you sure about that?
Because exposure time for 100/1000mm scope and 200/2000mm will be exactly the same if everything else is the same.
F-number controls exposure time, aperture controls "resolution".

Knez


Post Extras: Print Post   Remind Me!   Notify Moderator  
freestar8n
Carpal Tunnel
*****

Reged: 10/12/07

Re: Why do I want to image at a faster f ? new [Re: orion69]
      #5476449 - 10/18/12 07:14 AM

I think the thing to note about professional observatories, including the HST, is that they state the f/number of the imaging modes prominently - because it is now and always has been an important and fundamental metric of the expected irradiance at the image plane.

These discussions are problematic because they don't state up front what the goal is for the imager. You can't make concrete statements about SNR unless you say exactly what you mean by "signal" in the first place.

If instead you provide a single detector and two imaging systems - and a single object - and you just compare an 8x10 image of the full detector - with no scaling to match one image scale to the other - then things are more concrete. This is also, in practice, how people tend to image.

So compare 8" f/2 hyperstar with 14" f/11 on the Orion nebula. The hyperstar in a few 10s exposures would make a nice 8x10 image of the extended nebulosity and it would take little effort to capture - but it would not be high res. The 14" would give more close up detail if it is well guided, but you would need to expose longer to get the fainter nebulosity to rise above read noise if read noise is a factor.

I think it's crazy in these discussions and comparisons to force the 8" image to be blown up - and even CROPPED - to match the 14". The goal is to make a nice image to look at - and as long as the nebula is big you benefit from a wider field. If it is small, like the ring nebula, you get less benefit - but even then many images of galaxies and small objects look good in a somewhat wide view.

Frank


Post Extras: Print Post   Remind Me!   Notify Moderator  
Orion64
professor emeritus


Reged: 11/29/09

Loc: Western Cape, South Africa
Re: Why do I want to image at a faster f ? new [Re: freestar8n]
      #5476531 - 10/18/12 08:39 AM

Yes I think the reason for building larger and larger scopes is mainly to get better and better resolution on objects. When looking at complex star systems and where the aim is to find planets aroound stars, a fast focal lenght helps nothing. This is where you want the highest resolution possible. Thats what I think but please feel free to correct me.

Post Extras: Print Post   Remind Me!   Notify Moderator  
kfir Simon
Vendor-PowerNewt


Reged: 09/21/10

Re: Why do I want to image at a faster f ? new [Re: orion69]
      #5476628 - 10/18/12 10:00 AM

Please see this image -it is a SINGLE 15 minutes image with a 10" F2.8 telescope...
This shouls give you the answer....:)))

Kfir

http://www.pbase.com/tango33/image/146785280


Post Extras: Print Post   Remind Me!   Notify Moderator  
morten
super member


Reged: 07/22/08

Loc: Denmark
Re: Why do I want to image at a faster f ? new [Re: Orion64]
      #5476637 - 10/18/12 10:10 AM

I’ll try a slightly different angle, with the risk that it all been said in the excellent posts in this thread.
It is true that for a given aperture the single pixel SNR will improve by lowering the f/ratio. This is simply because more photons will be captured by that pixel when the f/ratio is lowered. More photons means less shot noise, which is (should be) the major source of noise in your images.
When many pixels are combined to form an image you still get less noise because the individual pixels are less noisy, however you pay a price in resolution because at a given aperture lower f/ratio means shorter focal lenght and lower pixel resolution.
When considering a reducer, or a scope with a lower f/ ratio a practical stance on this is:
1) Look at your FWHM on one of your best nights. If it is 2.5 pixels you have the best resolution you can probably get.
3) If FWHM is larger than 2.5 then you can reduce the f/number without loss of resolution, but with a free gain in field and SNR. A reducer will give you a different picture, but one with the same amount of detail in your DSO, and better SNR and wider field
4) If your pixel resolution is smaller than 2.5 pixels then a lower f/ number will make you pay a price in resolution, but you will still have a gain in a wider field, and a better SNR. A reducer will again give you a wider field, a better SNR, but you will loose detail in your DSO.
One can argue if the number should be 2.5. Craig Stark has written excellent, and highly recommended, texts on this on his webpage. He would probably argue for 2.0 (?). Other imagers would argue for a figure closer to 3.0 or perhaps even above 3.0, especially if they want to use deconvoluting processing algorithms.
If you are into narrow band, I’d say that you need to get as many photons into your pixels as possible. Unless you are imaging a bright object, have a lot of time under clear skies and, are very ambitious I’ll recommend a pixel resolution somewhat lower than i.e FWHM<2 .5.
Finally note, that binning for practical purposes work as halving your f/ratio, but without the reward of a wider field. For some chips, notably the 8300, there is a price for binning, i.e. the blooming in the lateral registers. A major drawback of this chip. You can soft-bin, but that is another story.

Kfirs post came while I was writing this. I have the same ASA X0.73 reducer that is the heart in the powernewt. I have seen for myself how the use of this reducer improves the SNR. I operate it at f/3.6, which makes life a little easier than working at f/2.8. Working at very short f/ ratios makes the system progressively sensitive (hysterical..) to tilt, collimation and focus. It is thus more demanding on the imager.
Kudus for the good single frame shown by Kfir, it brings the point on SNR afn f/ratios home.


Edited by morten (10/18/12 12:26 PM)


Post Extras: Print Post   Remind Me!   Notify Moderator  
korborh
professor emeritus
*****

Reged: 01/29/11

Re: Why do I want to image at a faster f ? new [Re: Orion64]
      #5476739 - 10/18/12 11:08 AM

Quote:

Do focal reducers give us less total imaging time needed or not?




Yes, they do.


Post Extras: Print Post   Remind Me!   Notify Moderator  
jrcrillyAdministrator
Refractor wienie no more
*****

Reged: 04/30/03

Loc: NE Ohio
Re: Why do I want to image at a faster f ? new [Re: korborh]
      #5476810 - 10/18/12 11:59 AM

Quote:

Quote:

Do focal reducers give us less total imaging time needed or not?




Yes, they do.




Sort of. They permit you to take a different image with less exposure time. To achieve the SAME image with less exposure time you'd need to increase aperture while holding the focal length constant.


Post Extras: Print Post   Remind Me!   Notify Moderator  
orion69
professor emeritus


Reged: 05/09/10

Loc: Croatia
Re: Why do I want to image at a faster f ? new [Re: jrcrilly]
      #5476815 - 10/18/12 12:02 PM

Quote:


Sort of. They permit you to take a different image with less exposure time. To achieve the SAME image with less exposure time you'd need to increase aperture while holding the focal length constant.




And if you do that, you actually have a faster scope...

Knez

Edited by orion69 (10/18/12 12:08 PM)


Post Extras: Print Post   Remind Me!   Notify Moderator  
morten
super member


Reged: 07/22/08

Loc: Denmark
Re: Why do I want to image at a faster f ? new [Re: jrcrilly]
      #5476864 - 10/18/12 12:27 PM Attachment (3 downloads)

Quote:



Sort of. They permit you to take a different image with less exposure time. To achieve the SAME image with less exposure time you'd need to increase aperture while holding the focal length constant.




If you are oversampling, then you can take the same image (meaning the same amount of detail) with better SNR and field by reducing.

I have a single 10 min frame of the same object as above af f/3.6. The SNR is clearly lower with the higher f ratio. I think the resolution is higher in my frame.

https://dl.dropbox.com/u/27494488/Autosave%20Image%20-017HA600_1X1a.jpg


Edited by morten (10/18/12 01:06 PM)


Post Extras: Print Post   Remind Me!   Notify Moderator  
jgraham
Postmaster
*****

Reged: 12/02/04

Loc: Dayton, Ohio
Re: Why do I want to image at a faster f ? new [Re: morten]
      #5476959 - 10/18/12 01:22 PM

This is always a fun discussion. I thought there was a fairly good analytical analysis published a while back in Sky & Telescope, though I don't recall the conclusion. I get the position that exposure time is set exclusively by aperture, but experience teaches me otherwise. Anecdotally, I do a lot of imaging with a 6" f/5 and an 8" f/10 and if this were true then exposures in the 8" should be shorter than those in the 6", but by far the opposite is true, I gotta go a lot longer with the 8". Also, I've seen some amazing images taken with an 8" Hyperstar with rediculously short exposures. I gotta look up the article and see if it grinds through the math, which is the right way to do it. Plan-B, it would be interesting to see results taken with the same scope and camera operating at different focal lenghts, say an 8" SCT at f/10, 6.3, and Hyperstar (f/3?).

Anywho, the bottom line for me has been fast f/ratio good but small, slow f/ratio less good but bigger. Not very scientific, but dem's da breaks.


Post Extras: Print Post   Remind Me!   Notify Moderator  
garret
super member


Reged: 07/26/09

Loc: Netherlands
Re: Why do I want to image at a faster f ? new [Re: morten]
      #5476969 - 10/18/12 01:28 PM Attachment (13 downloads)

I have a ASA F#3.6 Newton only because I want to have a large field for imaging, not for any 'noisy' reason...

I do have very good results for extended objects with only 14x60 seconds exposure time with a dslr...like m17:

Garret van der Veen

Edited by garret (10/18/12 01:29 PM)


Post Extras: Print Post   Remind Me!   Notify Moderator  
jrcrillyAdministrator
Refractor wienie no more
*****

Reged: 04/30/03

Loc: NE Ohio
Re: Why do I want to image at a faster f ? new [Re: orion69]
      #5476976 - 10/18/12 01:30 PM

Quote:

Quote:


Sort of. They permit you to take a different image with less exposure time. To achieve the SAME image with less exposure time you'd need to increase aperture while holding the focal length constant.




And if you do that, you actually have a faster scope...

Knez




Yes, but you didn't get there by adding a focal reducer (which was the question to which I responded). You got there by increasing aperture.


Post Extras: Print Post   Remind Me!   Notify Moderator  
morten
super member


Reged: 07/22/08

Loc: Denmark
Re: Why do I want to image at a faster f ? new [Re: jrcrilly]
      #5477012 - 10/18/12 01:46 PM

Quote:



Yes, but you didn't get there by adding a focal reducer (which was the question to which I responded). You got there by increasing aperture.




Yes you actually do get there by adding a reducer, but you may have to pay a price in resolution, if you are not sampling above FWHM=2.5pix.
Really, it isn't rocketscience but sophomore math.


Post Extras: Print Post   Remind Me!   Notify Moderator  
Orion64
professor emeritus


Reged: 11/29/09

Loc: Western Cape, South Africa
Re: Why do I want to image at a faster f ? [Re: jrcrilly]
      #5477040 - 10/18/12 02:01 PM

Morten, what is the aperture of your scope and what is the pixel size on your ccd?

Post Extras: Print Post   Remind Me!   Notify Moderator  
Pages: 1 | 2 | 3 | (show all)


Extra information
16 registered and 20 anonymous users are browsing this forum.

Moderator:  Oldfield, bilgebay 

Print Thread

Forum Permissions
      You cannot start new topics
      You cannot reply to topics
      HTML is disabled
      UBBCode is enabled


Thread views: 1704

Jump to

CN Forums Home




Cloudy Nights Sponsor: Astronomics