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

Reply
 
Thread Tools Rate Thread
  #1  
Old 07-03-2020, 01:48 PM
HeavyT
Registered User

HeavyT is offline
 
Join Date: May 2019
Location: South Australia
Posts: 48
What is the relationship between dynamic range and imaging speed?

I'm imaging at f7.6, using an ASI1600mm cooled to -15C.

Taking 900s subs, I routinely clip at least the end of the histogram capturing Sii and Oiii. On Eta Carina I could almost properly expose Ha at 420s.

I like to try to capture faint nebulosity and I know I have the option of increasing gain from 139/15. But I haven't, because I don't want to lose dynamic range.

Perhaps it's desperation, but the thought occurred to me: dynamic range would be the exposure range between highest and lowest exposure. If these values are governed by the speed of your scope, then boosting all signal would increase both highest and lowest values - bringing my "lowest" value into the histogram properly.

So would lowest and highest be increased by the same amount in this situation, or do they get compressed closer together? Given that the scope is slower to begin with, so the photons have a harder time building up on the sensor.

Thinking about say increasing gain to 200 only. I know if I went to extremes with the gain, I'd lose range.

I know the best option would be to collect 3600s subs (I've seen tables with this recommendation for this imaging speed and gain setting), but gathering 75hrs on one target over shorter nights is not something I'm keen to do.

Anyone have thoughts or experiences to share on this?

Thanks and clear skies
Reply With Quote
  #2  
Old 07-03-2020, 01:54 PM
gregbradley's Avatar
gregbradley
Registered User

gregbradley is offline
 
Join Date: Feb 2006
Location: Sydney
Posts: 17,877
With digital cameras increasing ISO reduces the dynamic range of the camera.

It would be the same with these cameras as the gain is the "ISO" setting.

You would also get a gain in dynamic range by binning.

Greg.
Reply With Quote
  #3  
Old 07-03-2020, 03:13 PM
Camelopardalis's Avatar
Camelopardalis (Dunk)
Drifting from the pole

Camelopardalis is offline
 
Join Date: Feb 2013
Location: Brisbane
Posts: 5,425
Maximum possible dynamic range from the camera is attained at roughly gain 75, where ~12 stops can be achieved. This would likely double the length of the exposures you'd need.

In practice, the dynamic range is likely to be limited by your imaging location. Narrowband is not immune from city lights.

What's the background ADU of your images?
Reply With Quote
  #4  
Old 08-03-2020, 08:01 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 HeavyT View Post
I'm imaging at f7.6, using an ASI1600mm cooled to -15C.

Taking 900s subs, I routinely clip at least the end of the histogram capturing Sii and Oiii. On Eta Carina I could almost properly expose Ha at 420s.

I like to try to capture faint nebulosity and I know I have the option of increasing gain from 139/15. But I haven't, because I don't want to lose dynamic range.

Perhaps it's desperation, but the thought occurred to me: dynamic range would be the exposure range between highest and lowest exposure. If these values are governed by the speed of your scope, then boosting all signal would increase both highest and lowest values - bringing my "lowest" value into the histogram properly.

So would lowest and highest be increased by the same amount in this situation, or do they get compressed closer together? Given that the scope is slower to begin with, so the photons have a harder time building up on the sensor.

Thinking about say increasing gain to 200 only. I know if I went to extremes with the gain, I'd lose range.

I know the best option would be to collect 3600s subs (I've seen tables with this recommendation for this imaging speed and gain setting), but gathering 75hrs on one target over shorter nights is not something I'm keen to do.

Anyone have thoughts or experiences to share on this?

Thanks and clear skies
Suggest that offset 15 is too low and a little bit of the faint nebulosity is probably being thrown away. Maybe set it at 50 and leave it there for everything. If you change offset, you will need a complete new calibration set though - sorry.

Other than that, there is not much change in final dynamic range after stacking with this camera by going to higher gain - up to about gain 100. Over gain 100 you start to lose a bit of final dynamic range at higher settings, but in compensation, the final read noise drops, so you get better results on faint nebulosity. Gain 200 would seem to be a good compromise for faint narrowband. Use lower gain and shorter subs if you want to retain detail in the very brightest stuff (eg the homunculus), but you will reduce the signal-to-noise on the faint stuff by doing so (in your terms, maximum signal and noise get "compressed ..together" at high gain). There will always be this type of compromise to be made when trying to image something with wide variations in signal level and it depends on what you value in the final image. Modern stacking software allows high dynamic range stacking, where combinations of exposures taken at different sub lengths are combined to fill in lost detail in the brightest bits - never tried it, but might be useful if you value bright detail.

With a slow scope in narrowband dark sky conditions, longer subs will always be better. The final faint signal after stacking will be the same for a given total exposure time, regardless of sub length (eg signal(10x60) = signal(1x600). However, the total read noise will be higher if you use more/shorter subs (eg RN(10x60) = 3.3xRN(1x600). ie, in the example above, the (faint stuff) signal to noise ratio will be 3.3x better with 1x600 than with 10x60 (ignoring shot noise). Agree that it is not practical for numerous reasons to use extreme sub lengths - maybe settle on 10-20 minutes at gain 200 (depending on your mount).

cheers Ray

Last edited by Shiraz; 08-03-2020 at 08:36 AM.
Reply With Quote
  #5  
Old 08-03-2020, 09:10 AM
HeavyT
Registered User

HeavyT is offline
 
Join Date: May 2019
Location: South Australia
Posts: 48
Quote:
Originally Posted by Shiraz View Post
Suggest that offset 15 is too low and a little bit of the faint nebulosity is probably being thrown away. Maybe set it at 50 and leave it there for everything. If you change offset, you will need a complete new calibration set though - sorry.

Other than that, there is not much change in final dynamic range after stacking with this camera by going to higher gain - up to about gain 100. Over gain 100 you start to lose a bit of final dynamic range at higher settings, but in compensation, the final read noise drops, so you get better results on faint nebulosity. Gain 200 would seem to be a good compromise for faint narrowband. Use lower gain and shorter subs if you want to retain detail in the very brightest stuff (eg the homunculus), but you will reduce the signal-to-noise on the faint stuff by doing so (in your terms, maximum signal and noise get "compressed ..together" at high gain). There will always be this type of compromise to be made when trying to image something with wide variations in signal level and it depends on what you value in the final image. Modern stacking software allows high dynamic range stacking, where combinations of exposures taken at different sub lengths are combined to fill in lost detail in the brightest bits - never tried it, but might be useful if you value bright detail.

With a slow scope in narrowband dark sky conditions, longer subs will always be better. The final faint signal after stacking will be the same for a given total exposure time, regardless of sub length (eg signal(10x60) = signal(1x600). However, the total read noise will be higher if you use more/shorter subs (eg RN(10x60) = 3.3xRN(1x600). ie, in the example above, the (faint stuff) signal to noise ratio will be 3.3x better with 1x600 than with 10x60 (ignoring shot noise). Agree that it is not practical for numerous reasons to use extreme sub lengths - maybe settle on 10-20 minutes at gain 200 (depending on your mount).

cheers Ray

Thanks very much Ray, Dunk and Greg for your responses. I see I've misunderstood what was happening with my offset - I started at 50, then 21, then read something that must have made me think lower offset was better for dynamic range. Thanks for correcting my misunderstanding. All my darks and flats are at 21, but I'd be willing to re do them if it works better. I haven't sorted any bias frames, I've never used them in dslr or mono imaging, and again must have read something suggesting they're not needed with the 1600mm. Fully aware there's a whole bunch of "information" on the internet...

I'm still working out how to derive background ADU. I think the mean readout shown in pixinsight is 900, does that sound right?
Reply With Quote
  #6  
Old 08-03-2020, 10:43 AM
HeavyT
Registered User

HeavyT is offline
 
Join Date: May 2019
Location: South Australia
Posts: 48
Quote:
Originally Posted by Camelopardalis View Post
Maximum possible dynamic range from the camera is attained at roughly gain 75, where ~12 stops can be achieved. This would likely double the length of the exposures you'd need.

In practice, the dynamic range is likely to be limited by your imaging location. Narrowband is not immune from city lights.

What's the background ADU of your images?

Hi Dunk

If K is the ADU value when I left click on an image in Pixinsight, I am getting 416 on a single sulfur sub and 116 on a stack of 25 with darks and flats applied. I don't know what a good K value would be though?
Reply With Quote
  #7  
Old 08-03-2020, 10:51 AM
Shiraz's Avatar
Shiraz (Ray)
Registered User

Shiraz is offline
 
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,918
suggest that you choose a couple of standard gains (eg ~75 for broadband and ~200 for narrowband) and then fix them and the offset and temp forever. Use the standard settings for all imaging and calibration - changing anything needs new dark cal data (and bias if you use it).

agree that bias is not much use - it makes sense when scaling darks, but that is not very straightforward with this camera and is not worth doing unless absolutely necessary.
I find that it is best to calibrate with a set of darks that still have bias included (ie not calibrated) and a set of flats that have had flat darks subtracted (if the flat subs are short enough you can use bias in place of the flat darks). Then when calibrating lights, I only use the uncalibrated darks and the calibrated flats - bias is not selected. Also, works properly only if I turn off dark optimisation in PI.

If your flats have been dark calibrated they may still be ok, but you will need to redo darks if you change offset.

cheers Ray

Last edited by Shiraz; 08-03-2020 at 11:05 AM.
Reply With Quote
  #8  
Old 08-03-2020, 11:28 AM
Camelopardalis's Avatar
Camelopardalis (Dunk)
Drifting from the pole

Camelopardalis is offline
 
Join Date: Feb 2013
Location: Brisbane
Posts: 5,425
As Ray says, I only use 2 gain values (rarely, a 3rd) and have built dark libraries up over time. I always use offset 50 also.

Yeah the background ADU is the K value in PI, if you hover the cursor over background sky and not nebulosity or any other light emitting object. It might help to change the probe size to 9 or more pixels (click on the "play" button at the bottom where you see the K values). I often use a gain of 139 (and offset 50) and aim for a background ADU of around 1200 or better (looking at uncalibrated light subs).

Your (very) low offset will skew your figures relative to mine, but at gain 139 offset 50, a short dark frame of mine will result in a mean/median of around 800 ADU. I'd imagine that yours would come out at ~120ADU?
Reply With Quote
  #9  
Old 08-03-2020, 03:39 PM
HeavyT
Registered User

HeavyT is offline
 
Join Date: May 2019
Location: South Australia
Posts: 48
Quote:
Originally Posted by Camelopardalis View Post
As Ray says, I only use 2 gain values (rarely, a 3rd) and have built dark libraries up over time. I always use offset 50 also.

Yeah the background ADU is the K value in PI, if you hover the cursor over background sky and not nebulosity or any other light emitting object. It might help to change the probe size to 9 or more pixels (click on the "play" button at the bottom where you see the K values). I often use a gain of 139 (and offset 50) and aim for a background ADU of around 1200 or better (looking at uncalibrated light subs).

Your (very) low offset will skew your figures relative to mine, but at gain 139 offset 50, a short dark frame of mine will result in a mean/median of around 800 ADU. I'd imagine that yours would come out at ~120ADU?
Thanks again guys

Well, I compared linear subs on Eta Carina (offset 21) and NGC3576 (offset 15) and their K values are more or less the same as low as 428 on Sii and 780 Oiii. All at Gain 139.

I've also looked at two dark frames taken at 180s and 900s. These achieve minimum K values of 16 and 224 respectively. I think they are taken at offset 21. Their mean values are 324 and 970 respectively. So I guess that explains why I've been sitting at 900s subs!

It sounds like it is a problem that, for example, the mean readout on my Carina sulfur exposure is 989? The same as a dark frame taken at the same duration of 900s. The minimum value is 32.

I'm going to try to up the offset first while I collect Ha on the target I've been imaging lately, if that looks good I'll re do the Oiii and Sii, and then perhaps take the plunge of upping the gain to 200 and starting again with darks and flats. It'll be worth it if I can increase the number of subs to integrate, and reduce the reliance on noise reduction caused by underexposed images.

Thanks again for stepping me through this.
Reply With Quote
  #10  
Old 08-03-2020, 09:56 PM
Camelopardalis's Avatar
Camelopardalis (Dunk)
Drifting from the pole

Camelopardalis is offline
 
Join Date: Feb 2013
Location: Brisbane
Posts: 5,425
Something doesn’t sound right....

Admittedly, I’ve never run off a set of 900s darks, but the mean of my darks doesn’t increase that significantly with longer exposures. (And I can only speak from experience with my camera - also a 1600MM-Cool)

Maybe the 180s darks were taken at a lower offset?

Have you tried looking in the FITS header of the original dark files? In PI, open the file and then there is an option in the File menu.
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 05:14 AM.

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