Jump to content


Photo

A Comparison of Exposure Lengths... XMas Tree

  • Please log in to reply
63 replies to this topic

#1 JWalk

JWalk

    Apollo

  • -----
  • Posts: 1365
  • Joined: 06 Jun 2010
  • Loc: San Antonio, TX

Posted 07 February 2013 - 07:56 PM

I read a very long thread the other day on here and I wanted to see if I could make some type of headway on the subject. I had some great skies again last night and I re-shot the Xmas tree area. I had previously shot it at 93X300sec. I shot it last night with 20min exposures. 9X1200sec. I tried to process them as good as I could.

Here are the results....

9X1200sec -40C

20min exposures

93X300sec

300sec exposures

I hope this might help a bit. I sure enjoy being able to shoot the 20min exposures on certain stuff.

Jimmy

#2 rainycityastro

rainycityastro

    Viking 1

  • *****
  • Posts: 718
  • Joined: 29 Mar 2010
  • Loc: Seattle, WA

Posted 07 February 2013 - 08:08 PM

The difference is quite astonishing. I would have thought they would be very close if not indistinguishable. With 93 subs I would have given the edge to 300 sec exposures.

Are you sure you processed them similarly? :-)

Regards,
--Ram

#3 JWalk

JWalk

    Apollo

  • -----
  • Posts: 1365
  • Joined: 06 Jun 2010
  • Loc: San Antonio, TX

Posted 07 February 2013 - 08:15 PM

I gave the 93X300 a lot of tries before I posted it. I went round and round with it. The 20min one from this morning was very easy to process. I worked it with my morning cup of Joe before I played.

#4 neptun2

neptun2

    Viking 1

  • -----
  • Posts: 845
  • Joined: 04 Mar 2007
  • Loc: Bulgaria

Posted 07 February 2013 - 09:12 PM

Well both examples are really good but the longer exposures really show more detail. You should have really dark skies to be able to make such long exposures. I can rarely go over 5 minutes without the background to become too bright.

#5 korborh

korborh

    Viking 1

  • *****
  • Posts: 801
  • Joined: 29 Jan 2011
  • Loc: Arizona

Posted 07 February 2013 - 09:29 PM

The 20min one is definitely better (deeper). Thannks for doing this. Do you know the read noise of your camera and the sky flux ?

#6 mcarroll

mcarroll

    Apollo

  • -----
  • Posts: 1029
  • Joined: 28 Aug 2010
  • Loc: Greensboro, NC

Posted 07 February 2013 - 09:34 PM

Nice comparison. Personally, I'd be thrilled with either image. I think it makes sense that the 1200s image would be deeper. My simple understanding is you would need at least 16x more subs to get similar signal to noise with 4x shorter exposure time per sub.

#7 neptun2

neptun2

    Viking 1

  • -----
  • Posts: 845
  • Joined: 04 Mar 2007
  • Loc: Bulgaria

Posted 07 February 2013 - 09:40 PM

I think that the problem is not only the SNR. There is faint detail which is simply missing in the shorter exposures and no matter how many of them you stack it just can't be pulled out.

#8 Rick J

Rick J

    Fly Me to the Moon

  • *****
  • Posts: 5750
  • Joined: 01 Mar 2008
  • Loc: Mantrap Lake, MN

Posted 07 February 2013 - 10:34 PM

Just what I expected. Reading your post before opening the images I thought "This won't be even close" and it wasn't. Your skies are very dark so you need a lot of time to drown read noise in the low sky noise. OSC is considerably worse this way than mono cameras of the same read noise level. The 300 second exposures wouldn't be close to photo(sky)limited in your dark skies so noise would cause weak data. It's always far better to stack images close to being photon limited than those a long way from it when going for something as weak as this area.

Have you measured what times it would take to reach sky limit? It likely will vary depending on which color you measure. You'd want to expose for the one needing the longest time as long as saturation wasn't a major issue.

Rick

#9 neutronman

neutronman

    Surveyor 1

  • -----
  • Posts: 1918
  • Joined: 10 Jan 2008
  • Loc: Dallas, Texas

Posted 07 February 2013 - 11:32 PM

Thanks for posting this, Jimmy. Yup, about what I expected. I ain't no technical genius, but I have observed that even given the same total exposure, the longer subs will always net more signal and fainter detail. If your skies be good (and yours are), go for as long as your tracking (and saturation) allows.

Now, one might say "gee, then why not do 3 x 60 min exposures instead of 9 x 20 min exposures"? Well, you need a decent number of subs to allow data rejection to work and to get cleaner combined subs.

BTW -- I'm coveting your ASA :waytogo:

#10 Warhen

Warhen

    Vendor (IP4AP)

  • *****
  • Vendors
  • Posts: 1074
  • Joined: 12 Nov 2005
  • Loc: WV, USA

Posted 08 February 2013 - 12:03 AM

Well done JWalk. As I said to you here in December '...I'm going to make a suggestion- longer exposures, fewer subs. Most who know me well, know a serious math guy I ain't, but I believe the evil 'law of diminishing return' rears its ugly head at ~16 frames, and really flatlines after ~24. With OSC, even at your very fast f/ratio, I'd suggest 600s... I'd have no problem w/ you going to 900s and I think 600s is a good move back. It's more about the lack of significant improvement with a huge number of frames, so we achieve greater impact gathering more signal in the individual exposures.'

Right on RickJ, Neutronman, etc. We simply cannot gain dimmest detail with a gazillion short exposures. And as John Smith pointed out at AIC 2012, which has been reinforced by VPcirc and RickJ on the forum, this is especially true in dark sky where overwhelming read noise is an added necessity.

#11 JWalk

JWalk

    Apollo

  • -----
  • Posts: 1365
  • Joined: 06 Jun 2010
  • Loc: San Antonio, TX

Posted 08 February 2013 - 12:12 AM

Trial and Error. I'm still new to this and it is all about getting to know your equipment. I have experimented with exposures and I had the data from some experimentation with exposures and just wanted to give some raw data on the current subject that has been hot on the forum. I have learned that longer seemed to be better.

#12 freestar8n

freestar8n

    Soyuz

  • *****
  • Posts: 3921
  • Joined: 12 Oct 2007

Posted 08 February 2013 - 02:41 AM

I read a very long thread the other day on here and I wanted to see if I could make some type of headway on the subject.



One of the main points I tried to make in that thread was that the benefit of longer exposures depends on the parameters of the imaging session. I see your camera listed so I can guess the gain and read noise - but I don't know the background sky flux in e/s. If that is small due to dark skies and the use of OSC Bayer filters, there will be greater win from longer exposures.

I expect the benefit you got from longer exposures will be consistent with the simple noise model once all the parameters are known - but I doubt you really needed to go 20m - though it wouldn't hurt as long as you don't lose subexposures. Longer is always better - to some extent - as long as you don't saturate, lose exposures, reveal flexure or field rotation, have too few subs for a good combine - etc.

You also went from 5m to 20m exposures, which is a factor of 4. A more common comparison would be going from 10m to 20m - i.e. doubling. People tend to expose an amount they are pretty comfortable with and then ask - should I double it. If they can comfortably quadruple it - then there is no real reason not to do 10m exposures in the first place.

Frank

#13 bilgebay

bilgebay

    Skylab

  • *****
  • Moderators
  • Posts: 4192
  • Joined: 06 Nov 2008
  • Loc: Turkiye - Istanbul and Marmaris

Posted 08 February 2013 - 02:46 AM

Beautiful Jimmy.

I guess well depth is also important with the longer subs. Your stars are not washed out and they have very nice colors.

#14 Alph

Alph

    Surveyor 1

  • -----
  • Posts: 1758
  • Joined: 23 Nov 2006
  • Loc: Melmac

Posted 08 February 2013 - 04:01 AM

I have experimented with exposures and I had the data from some experimentation with exposures and just wanted to give some raw data on the current subject that has been hot on the forum. I have learned that longer seemed to be better.


Stacking is a very powerful tool. You just have to know when to use it. When your subs don’t meet the minimum exposure time requirement, then stacking will not work very well. Your images have demonstrated that simple fact very well. It is all about S/N and it has nothing to do with drowning ‘other’ noise in the sky background or another noise. Do not think for a minute that a dark sky makes your images noisy, quite opposite.

#15 vpcirc

vpcirc

    Soyuz

  • *----
  • Posts: 3997
  • Joined: 09 Dec 2009
  • Loc: Merced CA

Posted 08 February 2013 - 04:13 AM

Please see above as Warren said. We heard it from several sources. Those experts in CCD's, John Smith, Kevin Nelson, are not sea lawyers giving bad information, one manufactures CCD Cameras and is an engineer, the other writes the software to run them and is all an expert in ccd's, not sure if he's an engineer. Please don't restart this disagreement on JWalks excellent post that is very useful to others.

John has a great set of rules for life

http://www.hiddenloft.com/rule_42.htm

#16 Dean

Dean

    Fly Me to the Moon

  • *****
  • Posts: 5446
  • Joined: 31 Dec 2004
  • Loc: Bailey Co Elev 8780 feet

Posted 08 February 2013 - 06:37 AM

Well both examples are really good but the longer exposures really show more detail. You should have really dark skies to be able to make such long exposures. I can rarely go over 5 minutes without the background to become too bright.

One thing not often mentioned is that shorter exposures don't have as much range as longer exposures making for less subtle detail that is harder to bring out. An interesting test would be to take the 300 sec subs and do sum combines with sets of 4subs and then median combining those. That way the sets will have the same effective exposure length as the 1200 sec subs, albeit with the extra read noise.

#17 Alph

Alph

    Surveyor 1

  • -----
  • Posts: 1758
  • Joined: 23 Nov 2006
  • Loc: Melmac

Posted 08 February 2013 - 12:27 PM

John has a great set of rules for life


Great set of rules. I would only add one more.
Even the greatest teachers fail. Some people never learn.

#18 blueman

blueman

    Photon Catcher

  • *****
  • Posts: 5365
  • Joined: 20 Jul 2007
  • Loc: California

Posted 08 February 2013 - 01:20 PM

I did not see what FLI camera you were using. But well depth certainly has something to do with exposure time. If you can do 20 minutes without completely saturating all the stars and cores of objects, then that is great. But you do have to make sure that you get the exposure and saturation levels worked out.

I find that with with a faster scope, I have f/5.4 that I use and a ML8300, 10 minutes is about all I can do before saturation becomes a problem. Of course a dimmer object might benefit a bit, but even 12 minutes can be too long for most bright objects.
Floyd

I read a very long thread the other day on here and I wanted to see if I could make some type of headway on the subject. I had some great skies again last night and I re-shot the Xmas tree area. I had previously shot it at 93X300sec. I shot it last night with 20min exposures. 9X1200sec. I tried to process them as good as I could.

Here are the results....

9X1200sec -40C

20min exposures

93X300sec

300sec exposures

I hope this might help a bit. I sure enjoy being able to shoot the 20min exposures on certain stuff.

Jimmy



#19 Peter in Reno

Peter in Reno

    Fly Me to the Moon

  • *****
  • Posts: 5894
  • Joined: 15 Jul 2008
  • Loc: Reno, NV

Posted 08 February 2013 - 01:39 PM

From JWalk web site, it appears he is using FLI Microline 11002 OSC.

http://www.flicamera...ets/ML11002.pdf

From OPT web site, the QE is a little low for color:

Mono: 50% - Color (RGB): 34%, 37%, 42%

Could low QE be one of the reasons for increasing sub-exposure times?

Thanks for posting comparison images. It was useful.

Peter

#20 cn register 5

cn register 5

    Viking 1

  • -----
  • Posts: 760
  • Joined: 26 Dec 2012

Posted 08 February 2013 - 04:22 PM

All I see are two images that were acquired differently, processed differently and, as a consequence, look different.

There is no scientific data here that differentiates between the exposure length. The difference in processing will mask any difference that exposure time could make.

Just to be clear. I'm not suggesting that, all other things being equal, shorter exposures will make a better image, I don't think they will. All I think is that the difference will be small. This is what the science tells us.

Would the difference be enough to notice? That's the question I originally asked and this doesn't answer it, especially for me under my skies. No amount of tests under pristine skies will answer that of course.

The impression I get is that people want to operate their astro photography on faith and arguments from authority. That seems to me like a religion, not a science.

That's not for me, I'll trust science and practicality.

Chris

#21 Mike Wiles

Mike Wiles

    Viking 1

  • -----
  • Posts: 950
  • Joined: 04 Feb 2009
  • Loc: Goodyear, AZ

Posted 08 February 2013 - 04:42 PM

Ugh.....

Jimmy, before this devolves into ridiculousness.......thank you for doing this. I thought about these two images all the way down the highway on a long drive today. I think it's very instructive and I appreciate that you took the time to do it.

A few months ago I did a comparison taking two images with exactly the same equipment/settings from a dark site and a light polluted site one day apart. I took advantage of PixInsight's object serialization to process the two images with exactly the same parameters. Two images that were exactly the same in every respect except that they were taken in two different locations one day apart. People complained that it wasn't a good demonstration of light pollution's effect on imaging because I didn't process each image differently to try and compensate for the differences in sky background. Clearly, you're doing something right and I look forward to each new image from you.

Both images are really good - but the 20 minute subs are awesome, especially considering it's just 3 hours of total integration. The color depth alone is an obvious reason to shoot as long as is reasonably possible under your given sky conditions.

Mike

#22 JWalk

JWalk

    Apollo

  • -----
  • Posts: 1365
  • Joined: 06 Jun 2010
  • Loc: San Antonio, TX

Posted 08 February 2013 - 07:33 PM

Thanks Mike and others, All I know is that the 9X20min was considerably easier to process than the 93X5min picture. I'll stick with 20min on objects like this.

#23 Mr_T

Mr_T

    Apollo

  • -----
  • Posts: 1267
  • Joined: 06 Apr 2008
  • Loc: NJ

Posted 08 February 2013 - 07:36 PM

Very Good result i would stick with 20 looks great

Les

#24 blueman

blueman

    Photon Catcher

  • *****
  • Posts: 5365
  • Joined: 20 Jul 2007
  • Loc: California

Posted 08 February 2013 - 08:08 PM

Because your camera has a fairly deep well, 60,000 and QE of 37% average color, it will definitely do better with long exposures.
Rule of thumb is take the longest exposure possible without causing undue saturation (hitting 65535 adu on all stars)and without the mount causing issues.
So, if you can do 20 minute exposures with these restrictions, go for it! :cool:
Blueman

#25 vpcirc

vpcirc

    Soyuz

  • *----
  • Posts: 3997
  • Joined: 09 Dec 2009
  • Loc: Merced CA

Posted 09 February 2013 - 03:48 AM

Mike, you can lead a horse to water.....






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics