Go Back   IceInSpace > Equipment > Astrophotography and Imaging Equipment and Discussions

Reply
 
Thread Tools Rating: Thread Rating: 8 votes, 5.00 average.
  #1  
Old 01-02-2014, 05:47 PM
Shiraz's Avatar
Shiraz (Ray)
Registered User

Shiraz is offline
 
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,918
the REALLY EASY way to work out sub exposure times

Was going through SNR theory for another purpose and suddenly realised that there is a very easy way to work out the best CCD sub exposure time while imaging.

Rather than go to online calculators and calibrate the system, all you need to do is make sure that the sub exposure length is long enough that the average background sky signal is above a target ADU value found from:

targetADU = Bias + 10*RN*RN/camera_gain

where the Bias is the normal bias level your camera produces, the camera_gain is the internal conversion factor (electrons/ADU) and RN is the camera read noise in electrons. This data should be fairly readily available for your camera.

When you have done the above calculation once for your camera, you thereafter use that targetADU value and (on the fly) choose exposures long enough to reach it, for whatever scope/filter/conditions etc that you have. It will provide you with an SNR that is within 5% of the best possible SNR, while preserving the dynamic range (in the same way as the online calculators). You don't need to know anything more than a single targetADU figure to ensure that your exposures are as effective as possible for whatever system you wish to put your camera on, or whatever the sky brightness is like.

As an example, my H694 has a bias of about 1100, a read noise of about 5.6 and a gain of about 0.3, so the targetADU is about 2150. On a clear night I will typically get this in 5+ minute subs on my current system, but hazy bright sky nights will only require about 2-3 minute subs. I can use the same targetADU figure for working out colour subs (which need to be longer than luminance subs), unless I choose shorter colour subs for extended dynamic range. When I get my new scope running, I will use the same targetADU value and choose exposures that yield the targetADU with that system.

You can use shorter subs (lower ADU) if you wish to get high dynamic range, but you will need to expose for a longer total time. Longer subs (higher background than the targetADU) will not change the overall SNR significantly, but will reduce the dynamic range. The calculation will work for narrowband as well, although you will need to include dark current in the average background signal estimate.

This process is so easy to use that I am embarrassed that I did not think of it before now. Hope it is useful and would be grateful for any feedback.

I don't think that this applies to dslr cameras, where the system noise varies with exposure time - it requires a camera with stable characteristics. Also, have no idea what complications the Bayer filters on an OSC introduce.

regards ray

Last edited by Shiraz; 03-02-2014 at 08:10 PM.
Reply With Quote
  #2  
Old 01-02-2014, 06:36 PM
Bassnut's Avatar
Bassnut (Fred)
Narrowfield rules!

Bassnut is offline
 
Join Date: Nov 2006
Location: Torquay
Posts: 5,064
WHAT!. A typical Ha 3nm 40min bin 1 sub on my SBIG gives 1100 odd ADUS back ground uncalibrated and 50 after calibration. Bias subs show about 1000 ADUs. Are you saying I need to increase subs to several hrs for best S/N?.

Have I missed something?, my 50 ADU background after calibration, is that what you mean by "above bias"?.

What about blooming on an NABG cam!. Several Hr subs would be unusable?.
Reply With Quote
  #3  
Old 01-02-2014, 06:43 PM
Shiraz's Avatar
Shiraz (Ray)
Registered User

Shiraz is offline
 
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,918
Quote:
Originally Posted by Bassnut View Post
. Are you saying I need to increase subs to several hrs for best S/N?.

Have I missed something?, my 50 ADU background after calibration, is that what you mean by "above bias"?.

What about blooming on an NABG cam!. Several Hr subs would be unusable?.
yes, yes and yes.

I am not saying this method is necessarily useful for all NB applications (it derives from LRGB considerations), but it is still correct - the basic rule for NB is "as long as possible".

If there are other competing considerations, then clearly they must be taken into account. However, the suggested method will give you the best possible SNR for any equipment. If you cannot use long enough subs for other reasons, you will be losing out on SNR due to the read noise. For example, your quoted sub background of 50 ADU will presumably correspond to something around 50 electrons. the shot noise will therefore be about 7 electrons, which will be a lot less than a single read event with a typical big pixel camera - your system is dominated by read noise. The only thing you can do to improve the SNR (apart from increasing the overall imaging time) is to decrease the number of reads by increasing the sub lengths.

Just for interest, have you ever tried really long NB subs.

regards Ray

Last edited by Shiraz; 01-02-2014 at 07:07 PM.
Reply With Quote
  #4  
Old 01-02-2014, 07:01 PM
Bassnut's Avatar
Bassnut (Fred)
Narrowfield rules!

Bassnut is offline
 
Join Date: Nov 2006
Location: Torquay
Posts: 5,064
OK, well thats interesting. Someone else explained this to me painfully before too (he takes 1 hr subs). I understand the other factors that can make longer subs difficult, but youve renewed my interest. I might actually try and take say 2hr subs when I can and see if it makes a difference (in my stupid need-to-try-to-believe mind). I know that in dark skies, read noise is always the dominating factor.
Reply With Quote
  #5  
Old 01-02-2014, 07:26 PM
RickS's Avatar
RickS (Rick)
PI cult recruiter

RickS is offline
 
Join Date: Apr 2010
Location: Brisbane
Posts: 10,584
Ray,

Your calculation gives 1695 ADU for my camera. I have got good results for dim targets (the jets of NGC 1097 et al) with 20 and 30 minute luminance subs which have background signals in that vicinity (1600 to 1800 ADU just from a few quick measurements).

Cheers,
Rick.
Reply With Quote
  #6  
Old 01-02-2014, 08:16 PM
Shiraz's Avatar
Shiraz (Ray)
Registered User

Shiraz is offline
 
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,918
Quote:
Originally Posted by Bassnut View Post
OK, well thats interesting. Someone else explained this to me painfully before too (he takes 1 hr subs). I understand the other factors that can make longer subs difficult, but youve renewed my interest. I might actually try and take say 2hr subs when I can and see if it makes a difference (in my stupid need-to-try-to-believe mind). I know that in dark skies, read noise is always the dominating factor.
OK Fred. yep, I know I can be painful, but I mean well Increasing the sub length should make a noticeable SNR difference - got to be worth a try?

Quote:
Originally Posted by RickS View Post
Ray,

Your calculation gives 1695 ADU for my camera. I have got good results for dim targets (the jets of NGC 1097 et al) with 20 and 30 minute luminance subs which have background signals in that vicinity (1600 to 1800 ADU just from a few quick measurements).

Cheers,
Rick.
wooah, thanks Rick. I had some reservations about posting at all in case I had stuffed up - appreciate the positive feedback.

It really seems to be a very easy way to systematically set exposures based on first principles - do one simple calculation for your camera, remember the result and aim for it thereafter. regards Ray

Last edited by Shiraz; 04-02-2014 at 06:35 AM.
Reply With Quote
  #7  
Old 02-02-2014, 10:24 AM
Shiraz's Avatar
Shiraz (Ray)
Registered User

Shiraz is offline
 
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,918
Quote:
Originally Posted by Bassnut View Post
WHAT!. A typical Ha 3nm 40min bin 1 sub on my SBIG gives 1100 odd ADUS back ground uncalibrated and 50 after calibration. Bias subs show about 1000 ADUs. Are you saying I need to increase subs to several hrs for best S/N?.

Have I missed something?, my 50 ADU background after calibration, is that what you mean by "above bias"?.

What about blooming on an NABG cam!. Several Hr subs would be unusable?.
Fred
just realised that your calibration process is removing dark current and that is part of the background signal for NB - your "above bias" background signal is about 100 ADU rather than 50, so your background noise is about 10 electrons, not 7. Makes no difference to the earlier conclusion, your system is still read noise limited and would benefit from much longer subs. Regards ray

Last edited by Shiraz; 03-02-2014 at 07:43 AM.
Reply With Quote
  #8  
Old 03-02-2014, 08:51 AM
alistairsam's Avatar
alistairsam
Registered User

alistairsam is offline
 
Join Date: Nov 2009
Location: Box Hill North, Vic
Posts: 1,837
Hi Ray,

Can you help with data foe the kaf8300 that a lot of us use.
Once calculated, you mention we should aim at exposures that give us a background with this value. But that implies a few trial subs.
Isnt it easier to take on 3 min sub and use the online calculators.

I get 1800 in 3 minute L even with an LP filter, thats how bad my LP is and the online calculators always given me figures like 90 or 120secs.
However ive seen a marked difference with 5, 10 min subs.
So if I go by your calc or the online calc values, get lots of smaller subs, will that be better in terms of getting details

Thanks
Alistair
Reply With Quote
  #9  
Old 03-02-2014, 09:17 AM
multiweb's Avatar
multiweb (Marc)
ze frogginator

multiweb is offline
 
Join Date: Oct 2007
Location: Sydney
Posts: 22,060
+1. What's the ideal time with a QHY9 mono? Not that I can guide that long but here's wishful thinking.
Reply With Quote
  #10  
Old 03-02-2014, 11:50 AM
Shiraz's Avatar
Shiraz (Ray)
Registered User

Shiraz is offline
 
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,918
Quote:
Originally Posted by alistairsam View Post
Hi Ray,

Can you help with data foe the kaf8300 that a lot of us use.
Once calculated, you mention we should aim at exposures that give us a background with this value. But that implies a few trial subs.
Isnt it easier to take on 3 min sub and use the online calculators.

I get 1800 in 3 minute L even with an LP filter, thats how bad my LP is and the online calculators always given me figures like 90 or 120secs.
However ive seen a marked difference with 5, 10 min subs.
So if I go by your calc or the online calc values, get lots of smaller subs, will that be better in terms of getting details

Thanks
Alistair
Quote:
Originally Posted by multiweb View Post
+1. What's the ideal time with a QHY9 mono? Not that I can guide that long but here's wishful thinking.
The calculations depend on the parameters of the specific camera, not the chip.

You can easily measure bias by taking a short exposure dark - you need only do it once. Bias is likely set by a potentiometer or similar in the camera - yours will be different to mine.

Read noise and gain can be measured reasonably easily, but if you wish to use manufacturer's typical data, you should not be too far wrong. For example, SBIG quote 9.3 e read noise and 0.37 gain for their STF8300. QSI quote 8 electrons read and 0.5 or 1.1 gain (user can set) for their 583. If you have an SBIG, use (Bias + 2300). If you have a QSI583 use (Bias + 1300) at 0.5 gain.

If you have an 8300 camera for which you cannot find/measure gain or read noise data it would probably be reasonable to assume a read noise of 9 and gain of 0.4, so use (Bias + 2000).

Alistair, the online calculators have some possible pitfalls - for example you need to remove the bias from the test sub - it is not clear how this is done in some. With your camera, you will need longer subs if you are only getting 1800 for the sky (does this include bias or is it calibrated out?). As shown above, you need about (2300 + Bias), so longer subs might be more effective, if your tracking is OK,

Marc, the QHY9 is unique in that the user can set both the gain and bias. Suggest that you use a gain of 0, set the offset to get a bias around 1000ADU, measure the actual Bias and set your sub exposures so that you get somewhere near a background sky ADU of (Bias + 2000).

The real advantage of this approach is that you need do it once for your camera. Thereafter, just adjust your sub exposure times on the fly so that you get somewhere near the desired ADU and that's all you need to do. One sub should be enough to tell you if you need to adjust. It doesn't matter what scope you are using, which filter, what the sky is like or anything else - just adjust the sub exposure times so that you get in the ballpark of the calculated ADU on the sky background. Of course, getting it a bit wrong is not the end of the world; this process (and the online calculators) aim to optimise your overall imaging efficiency and you can still get good results if you get it a bit wrong - it will take a bit longer to get there if you underexpose or you will lose a bit of dynamic range if you expose for too long.

regards ray

Last edited by Shiraz; 03-02-2014 at 03:47 PM.
Reply With Quote
  #11  
Old 03-02-2014, 12:02 PM
multiweb's Avatar
multiweb (Marc)
ze frogginator

multiweb is offline
 
Join Date: Oct 2007
Location: Sydney
Posts: 22,060
Quote:
Originally Posted by Shiraz View Post
Marc, the QHY9 is unique in that the user can set both the gain and bias. Suggest that you use a gain of 0, set the offset to get a bias around 1000ADU, measure the Bias and aim for a background ADU of Bias + 2000.
Cool - will try that. I've always used gain 1. I think because at the time CCD Stack didn't like FITS files with zero gain in the header. Might have changed. Will check it out.
Reply With Quote
  #12  
Old 03-02-2014, 12:39 PM
Shiraz's Avatar
Shiraz (Ray)
Registered User

Shiraz is offline
 
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,918
I think gain 1 should be fine. Only suggested gain 0 because I think that is the unboosted state and I don't think gain 1 is too different (do you know what the difference is?).

Last edited by Shiraz; 03-02-2014 at 01:22 PM.
Reply With Quote
  #13  
Old 03-02-2014, 02:08 PM
alistairsam's Avatar
alistairsam
Registered User

alistairsam is offline
 
Join Date: Nov 2009
Location: Box Hill North, Vic
Posts: 1,837
Hi Ray,

Mine's an STF8300M and with the shortest I have which is a 3 min dark, below is what I got from ccdstack

background 928.78
maximum 1,053

and maxim
Minimum 730.000
Average 978.707

I do have bias frames and this is what I get

background 858.30
maximum 921

you've mentioned
SBIG quote 9.3 e read noise and 0.37 gain for their STF8300


ADU = Bias + 10*RN*RN/camera_gain

ADU = 970 + 865/0.37
which is 3307

Is this what I should aim for?

here are some values from my previous images

10min Ha - 1300
20min Ha - 1700
8min L - 4050
2sec to 60sec L - 1500 to 2000

Do we aim at a relatively empty patch for the measurements?
above was from different nights.

I might run a sequence and plot it to see the trend for the stf and my skies.

thanks
Alistair
Reply With Quote
  #14  
Old 03-02-2014, 03:08 PM
Shiraz's Avatar
Shiraz (Ray)
Registered User

Shiraz is offline
 
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,918
Quote:
Originally Posted by alistairsam View Post
Hi Ray,

Mine's an STF8300M

I do have bias frames and this is what I get

background 858.30
maximum 921

you've mentioned
SBIG quote 9.3 e read noise and 0.37 gain for their STF8300


ADU = Bias + 10*RN*RN/camera_gain

ADU = (970)860 + 865/0.37
which is 3307

Is this what I should aim for?

Your bias is ~860 from your measurements, combine that with the SBIG specs and about 3200 is what you aim for - no need for it to be exact, somewhere between 3000 and 3500 maybe.

Congratulations, you have now completed the calibration of your camera - from here on in, adjust sub exposures so that your lights have background values between about 3000 and 3500. That's all you need to do.

here are some values from my previous images


8min L - 4050

this was a reasonably good exposure for the conditions, but you could have shortened it a bit if you wanted - 5 minutes would have yielded a bit over 3000, so would have still been OK for noise, but with almost twice the dynamic range.

Do we aim at a relatively empty patch for the measurements?
above was from different nights.

Use the region that you are going to image (or somewhere close by). I presume that your capture software will allow you to put the cursor on a test image and read out the ADU - just test a few points in a blank sky region near the field centre to find a rough average value - that will do. Don't worry if it is a bit out of spec - the sky will vary in brightness through a sequence anyway, so just get exposures that are in the ballpark and err on the longer exposure side.

For NB, use "as long as possible", even your 20 minutes exposure did not get anywhere near 3000.



I might run a sequence and plot it to see the trend for the stf and my skies.

thanks
Alistair
Good luck. Ray

Last edited by Shiraz; 03-02-2014 at 08:17 PM.
Reply With Quote
  #15  
Old 03-02-2014, 03:43 PM
Shiraz's Avatar
Shiraz (Ray)
Registered User

Shiraz is offline
 
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,918
just had a thought. It would be dead easy for the camera makers to provide "suggested background ADU" figures for their cameras - how useful would that be!

regards Ray
Reply With Quote
  #16  
Old 03-02-2014, 06:07 PM
alistairsam's Avatar
alistairsam
Registered User

alistairsam is offline
 
Join Date: Nov 2009
Location: Box Hill North, Vic
Posts: 1,837
[QUOTE=Bassnut;1054084). I know that in dark skies, read noise is always the dominating factor.[/QUOTE]

Hi
I'm new to these concepts, pardon my ignorance, are you'll able to explain in simplistic terms what the above means?

We go to dark skies to image without light pollution where it would normally wash out the background.

So how are dark skies worse than brighter skies in terms of read noise or do you mean overall noise in several short subs as opposed to noise in fewer long subs as there are fewer reads?
Any links explaining thesr concepts would also help

When you say we can achieve the best snr with the calculated target adu, wont longer subs increase background values, so shouldn't we measure signal relative to noise and stop where signal doesnt increase significantly not just background values which could be LP? Are we really measuring noise?


Thanks
Reply With Quote
  #17  
Old 03-02-2014, 07:09 PM
Shiraz's Avatar
Shiraz (Ray)
Registered User

Shiraz is offline
 
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,918
hope you don't mind me jumping in Fred.

Alistair, the two sources of noise being considered here are read noise and shot noise from sky radiation.

The read noise is a burst of noise that you get on every read of data from a pixel - you can't do anything to reduce it, except by reducing the number of reads you perform (ie use very few long subs). The shot noise is determined by random arrival times of photons - it turns out that the signal to noise ratio due to shot noise is just the square root of the signal and the shot noise therefore increases as you increase the total detected signal. However, you are better off to have more signal, since the signal level increases more rapidly than the associated shot noise.

The read noise has a fixed average power, so if you can increase the signal, you eventually reach the point where the shot noise (and the signal) has gone up enough that the read noise becomes relatively insignificant in the total noise. At this point, all you are left with (practically) is signal and shot noise and that is as good as gets. The way to get more signal is to integrate longer - ie use longer subs. You do not want to use subs that are too long though, because you start filling up the wells on the chip and dynamic range suffers. To improve the signal to noise ratio once you have got to the point where shot noise is the limiting factor, you can add more subs - that way you get better signal to noise ratio without messing up the dynamic range and the shot noise will still be dominant over the read noise.

If the sky is bright, you will get more background signal in a given time, so you can use shorter subs and still have enough signal to overwhelm the read noise. the read noise does not change of course.

If the sky is really dark and you are using a narrow-band filter, there may not be much sky signal at all and you will need really long subs to get enough signal that the shot noise will overwhelm the read noise. The sky brightness makes no difference to the read noise, but it does determine how much effect the read noise has on the signal to noise ratio for a given sub exposure length.

have a read of this and look at the section on sky limited exposure and the references in that section.
http://starizona.com/acb/ccd/advtheoryexp.aspx

Re the question of what is signal and what is noise, the limiting noise in astro images is not in the bright bits - you are always trying to extract slight signals without making the sky background noise too objectionable - ie it is the noise in the background sky bits that determines how much of the dim stuff you can extract and background noise is what the analysis is all about. If you have a really bright object and don't care so much about background noise, use shorter subs.

Of course, the whole point of the suggested method is that it takes care of the details - you just have to remember one easily calculated ADU number and aim for it in your imaging. That's what I will be doing from now on.

regards Ray

Last edited by Shiraz; 04-02-2014 at 09:02 AM.
Reply With Quote
  #18  
Old 06-02-2014, 07:11 PM
stevous67 (Steve M)
Registered User

stevous67 is offline
 
Join Date: Jul 2008
Location: Melbourne, Australia
Posts: 633


Thanks Alistair for helping a lot of us with your data. Nice topic Ray, I'll just have to read it once or twenty times just to be sure. It's been very helpful and interesting.

Thanks,

Steve
Reply With Quote
  #19  
Old 06-02-2014, 09:37 PM
alistairsam's Avatar
alistairsam
Registered User

alistairsam is offline
 
Join Date: Nov 2009
Location: Box Hill North, Vic
Posts: 1,837
Thanks to Ray and Fred
I'm the same Steve, kept it for a "quiet read" atleast 20 times.

One more thing I wanted to ask was about well depth.
After this calc, we aim at reaching the target adu, but how do we deal with shallow well depth sensors like the 8300. 2x2 is not a good option for this and long exposures with rgb tend to saturate the stars rendering it almost colourless.
Any tips?

Rolf gets some amazing colours with his qsi which has the same sensor but his filters are the asronomiks I believe.
Any suggestions

Alistair
Reply With Quote
  #20  
Old 06-02-2014, 11:54 PM
Shiraz's Avatar
Shiraz (Ray)
Registered User

Shiraz is offline
 
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,918
Quote:
Originally Posted by stevous67 View Post


Thanks Alistair for helping a lot of us with your data. Nice topic Ray, I'll just have to read it once or twenty times just to be sure. It's been very helpful and interesting.

Thanks,

Steve
thanks Steve.

Rats, I was hoping that this would be easy to use - big problem is I have a tendency to rabbit on a bit and obscure what I am trying to say. The basic idea is very straightforward - do the calculation once and then aim for the targetADU value thereafter. Used it myself last night - really is dead easy. My camera has a targetADU of about 2150 and I got about a 2100 sky from 5 minute subs - close enough and gave good results.

Quote:
Originally Posted by alistairsam View Post
Thanks to Ray and Fred
I'm the same Steve, kept it for a "quiet read" atleast 20 times.

One more thing I wanted to ask was about well depth.
After this calc, we aim at reaching the target adu, but how do we deal with shallow well depth sensors like the 8300. 2x2 is not a good option for this and long exposures with rgb tend to saturate the stars rendering it almost colourless.
Any tips?

Rolf gets some amazing colours with his qsi which has the same sensor but his filters are the asronomiks I believe.
Any suggestions

Alistair
You can actually ignore well depth - it is taken care of automatically and you get an optimised combination of noise and dynamic range by aiming for the targetADU. If you really want more dynamic range, use shorter subs and image for a bit longer overall. Typical cameras will have targetADUs (for the background sky) of 1000-2000 + bias and they all have a top of ~65000 ADU, so the dynamic ranges will be similar - most cameras have dynamic ranges of about 70dB, regardless of the well depth, since those that have large wells also have higher read noise. The cameras with low read noise need shorter exposures than high read noise cameras, but the targetADU method takes care of all that for you. You really do not have to do anything more to get good sub exposures.

not sure what binning does - will give it some thought.

Last edited by Shiraz; 07-02-2014 at 12:17 AM.
Reply With Quote
Reply

Bookmarks

Thread Tools
Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +10. The time is now 08:20 PM.

Powered by vBulletin Version 3.8.7 | Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Advertisement
Testar
Advertisement
Bintel
Advertisement