Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Nikon d5300, exposure time recommendations?

  • Please log in to reply
19 replies to this topic

#1 betelgeuse91

betelgeuse91

    Viking 1

  • *****
  • topic starter
  • Posts: 518
  • Joined: 28 Mar 2018
  • Loc: Ontario, Canada

Posted 19 November 2019 - 01:44 AM

Hi, I have been shooting 5 min subs and been getting great results so far but I started thinking that I might not benefit much from long subs. 

 

I am thinking of shooting 3 min subs but what do you guys think? Any recommendations?

 

I shoot in Bortle 5~6 sky with ISO 200. I have no tracking issues. I got an f/4.8 80mm refractor.

 

Thanks!


Edited by betelgeuse91, 19 November 2019 - 01:57 AM.


#2 sharkmelley

sharkmelley

    Mercury-Atlas

  • *****
  • Posts: 2512
  • Joined: 19 Feb 2013

Posted 19 November 2019 - 01:59 AM

Shoot so the peak of the back-of-camera histogram is 1/4 from the left-hand-side.  This advice works for all ISOs, all f-ratios and all light pollution levels.

 

Mark


Edited by sharkmelley, 19 November 2019 - 02:01 AM.

  • bobzeq25 and Fervent Astronomy like this

#3 bobzeq25

bobzeq25

    Hubble

  • *****
  • Posts: 17326
  • Joined: 27 Oct 2014

Posted 19 November 2019 - 02:02 AM

Hi, I have been shooting 5 min subs and been getting great results so far but I started thinking that I might not benefit much from long subs. 

 

I am thinking of shooting 3 min subs but what do you guys think? Any recommendations?

 

I shoot in Bortle 5~6 sky with ISO 200. I have no tracking issues. I got an f/4.8 80mm refractor.

 

Thanks!

Depends strongly on your skies and your setup, copying someone else is unlikely to be useful.  sharkmelleys advice is good.

 

Once you get subexposure in the ballpark, the key to better images is more total imaging time.  Follow sharkmelley's recommendation, then shoot more subs.  <smile>



#4 betelgeuse91

betelgeuse91

    Viking 1

  • *****
  • topic starter
  • Posts: 518
  • Joined: 28 Mar 2018
  • Loc: Ontario, Canada

Posted 19 November 2019 - 02:19 AM

Shoot so the peak of the back-of-camera histogram is 1/4 from the left-hand-side.  This advice works for all ISOs, all f-ratios and all light pollution levels.

 

Mark

 

Depends strongly on your skies and your setup, copying someone else is unlikely to be useful.  sharkmelleys advice is good.

 

Once you get subexposure in the ballpark, the key to better images is more total imaging time.  Follow sharkmelley's recommendation, then shoot more subs.  <smile>

 

Thank you for the advice. I will pay attention to the histogram next time!

Hope 3 min will be enough... Since then I won't have to worry much about amp glow :) 



#5 sharkmelley

sharkmelley

    Mercury-Atlas

  • *****
  • Posts: 2512
  • Joined: 19 Feb 2013

Posted 19 November 2019 - 03:21 AM

Thank you for the advice. I will pay attention to the histogram next time!

Hope 3 min will be enough... Since then I won't have to worry much about amp glow smile.gif

I don't understand your comment about amp glow.  What difference are you expecting sub length to make?

 

Mark



#6 betelgeuse91

betelgeuse91

    Viking 1

  • *****
  • topic starter
  • Posts: 518
  • Joined: 28 Mar 2018
  • Loc: Ontario, Canada

Posted 19 November 2019 - 08:19 AM

I don't understand your comment about amp glow.  What difference are you expecting sub length to make?

 

Mark

Wouldn't there be more amp glow in 5 min subs than 3 min subs? 

I saw some people trying to minimize exposure times to avoid issues with amp glow... 



#7 terry59

terry59

    Cosmos

  • *****
  • Posts: 9203
  • Joined: 18 Jul 2011
  • Loc: Colorado, USA

Posted 19 November 2019 - 08:54 AM

I recommend experimenting with ADU values and finding what works. Far more precise and repeatable



#8 sharkmelley

sharkmelley

    Mercury-Atlas

  • *****
  • Posts: 2512
  • Joined: 19 Feb 2013

Posted 19 November 2019 - 09:40 AM

Wouldn't there be more amp glow in 5 min subs than 3 min subs? 

I saw some people trying to minimize exposure times to avoid issues with amp glow... 

Reducing exposure times to minimize amp glow is not something I've heard of.  I guess you're referring to thermal amp glow.  Sure there will be less amp glow in the shorter exposures but there will also be less signal. Amp glow as a proportion of signal will be identical for the 5min and 3min subs, will it not?

 

Mark



#9 betelgeuse91

betelgeuse91

    Viking 1

  • *****
  • topic starter
  • Posts: 518
  • Joined: 28 Mar 2018
  • Loc: Ontario, Canada

Posted 19 November 2019 - 11:47 AM

I recommend experimenting with ADU values and finding what works. Far more precise and repeatable

How do I do that....?? By any chance do you have a link for a tutorial?

 

 

Reducing exposure times to minimize amp glow is not something I've heard of.  I guess you're referring to thermal amp glow.  Sure there will be less amp glow in the shorter exposures but there will also be less signal. Amp glow as a proportion of signal will be identical for the 5min and 3min subs, will it not?

 

Mark

I actually don't know too much about amp glow but i heard that somewhere... Maybe it was not so correct... 



#10 brettkoz

brettkoz

    Vostok 1

  • -----
  • Posts: 131
  • Joined: 15 Dec 2015
  • Loc: Hesperia, Michigan

Posted 19 November 2019 - 12:55 PM

Following the ADU values is a good method, as is watching the histogram, but, for me, I only really care about not saturating too many stars. Of course that depends on what area of the sky you're looking at, but I find that for broadband I get pretty good star color at 120s and narrowband I can sometimes go up to 300s (with my 071MC Pro). If you're using Sequence Generator Pro, you can find out the star levels and the overal mean ADU value in the Image Statistics panel.



#11 terry59

terry59

    Cosmos

  • *****
  • Posts: 9203
  • Joined: 18 Jul 2011
  • Loc: Colorado, USA

Posted 19 November 2019 - 01:45 PM

How do I do that....?? By any chance do you have a link for a tutorial?

 

 

I actually don't know too much about amp glow but i heard that somewhere... Maybe it was not so correct... 

It is quite simple. SGP reports ADU values and the values for each sub can be found using PixInsight.  Shoot subs of various length and decide how many clipped stars are acceptable to you. For WB, it should be in the 3k-5k range and with NB it will generally be in the 400-500 range



#12 pinzmann

pinzmann

    Mariner 2

  • -----
  • Posts: 229
  • Joined: 02 Sep 2007
  • Loc: Yucca valley, ca

Posted 19 November 2019 - 01:52 PM

Shoot so the peak of the back-of-camera histogram is 1/4 from the left-hand-side.  This advice works for all ISOs, all f-ratios and all light pollution levels.

 

Mark

I image with a D5300 at 200 iso. To get my histogram at 1/4 from the left, I would have to shoot at least 15min subs. Too much chance of something going wrong at 15 min subs. Also not sure my equipment is capable. Scopes are f/6 and f/7, Bortle 4 skies.



#13 sharkmelley

sharkmelley

    Mercury-Atlas

  • *****
  • Posts: 2512
  • Joined: 19 Feb 2013

Posted 20 November 2019 - 01:01 AM

I image with a D5300 at 200 iso. To get my histogram at 1/4 from the left, I would have to shoot at least 15min subs. Too much chance of something going wrong at 15 min subs. Also not sure my equipment is capable. Scopes are f/6 and f/7, Bortle 4 skies.

At ISO 200, the 1/4 back-of-camera histogram criterion is quite conservative.  In fact you can drop the sub length down to 1/8 BoC histogram at ISO 200 without incurring any noise penalty in your final stack. The total integration time must remain the same of course.

 

If you have the tools available then a much more accurate criterion for minimum exposure length is for the background (i.e. light pollution) noise to swamp the read noise by at least a factor of 3.   Measure the standard deviation of pixel values in a single bias frame to obtain the read noise then measure the standard deviation in a background area (free of stars and nebulosity) of a single light frame (at the same ISO of course).  Adjust exposure length, so they differ by at least a factor of 3. 

 

For a one-shot-colour or DSLR you will need to do this in each colour channel separately though usually it is the blue channel that tends to be the weakest when the background level is dominated by sodium streetlights..

 

Mark


Edited by sharkmelley, 20 November 2019 - 01:04 AM.

  • khobar likes this

#14 betelgeuse91

betelgeuse91

    Viking 1

  • *****
  • topic starter
  • Posts: 518
  • Joined: 28 Mar 2018
  • Loc: Ontario, Canada

Posted 20 November 2019 - 09:35 AM

It is quite simple. SGP reports ADU values and the values for each sub can be found using PixInsight.  Shoot subs of various length and decide how many clipped stars are acceptable to you. For WB, it should be in the 3k-5k range and with NB it will generally be in the 400-500 range

I see.. Thank you.

 

 

I image with a D5300 at 200 iso. To get my histogram at 1/4 from the left, I would have to shoot at least 15min subs. Too much chance of something going wrong at 15 min subs. Also not sure my equipment is capable. Scopes are f/6 and f/7, Bortle 4 skies.

May I ask what your typical subexposure time is? 

 

 

At ISO 200, the 1/4 back-of-camera histogram criterion is quite conservative.  In fact you can drop the sub length down to 1/8 BoC histogram at ISO 200 without incurring any noise penalty in your final stack. The total integration time must remain the same of course.

 

If you have the tools available then a much more accurate criterion for minimum exposure length is for the background (i.e. light pollution) noise to swamp the read noise by at least a factor of 3.   Measure the standard deviation of pixel values in a single bias frame to obtain the read noise then measure the standard deviation in a background area (free of stars and nebulosity) of a single light frame (at the same ISO of course).  Adjust exposure length, so they differ by at least a factor of 3. 

 

For a one-shot-colour or DSLR you will need to do this in each colour channel separately though usually it is the blue channel that tends to be the weakest when the background level is dominated by sodium streetlights..

 

Mark

Interesting, but why do you measure the standard deviation instead of average or median? 



#15 sharkmelley

sharkmelley

    Mercury-Atlas

  • *****
  • Posts: 2512
  • Joined: 19 Feb 2013

Posted 20 November 2019 - 02:08 PM

Interesting, but why do you measure the standard deviation instead of average or median? 

Image noise caused by light pollution is typically the main constraint on imaging faint objects. The light pollution itself can be digitally subtracted but it leaves behind the image noise caused by the light pollution (technically known as photon shot noise).  As long as the exposure length is chosen so that this source of noise is the main source of noise (i.e swamping the sensor read noise) then it does not affect the ability to image those faint objects.  Standard deviation directly measures both the read noise (in the bias frame) and the combined read noise and photon shot noise (in the light frames) and that's why we use standard deviation rather than average or median.

 

Mark



#16 pinzmann

pinzmann

    Mariner 2

  • -----
  • Posts: 229
  • Joined: 02 Sep 2007
  • Loc: Yucca valley, ca

Posted 20 November 2019 - 02:54 PM

I see.. Thank you.

 

 

May I ask what your typical subexposure time is? 

 

 

Interesting, but why do you measure the standard deviation instead of average or median? 

My usual subs are 5 min



#17 betelgeuse91

betelgeuse91

    Viking 1

  • *****
  • topic starter
  • Posts: 518
  • Joined: 28 Mar 2018
  • Loc: Ontario, Canada

Posted 20 November 2019 - 04:22 PM

Image noise caused by light pollution is typically the main constraint on imaging faint objects. The light pollution itself can be digitally subtracted but it leaves behind the image noise caused by the light pollution (technically known as photon shot noise).  As long as the exposure length is chosen so that this source of noise is the main source of noise (i.e swamping the sensor read noise) then it does not affect the ability to image those faint objects.  Standard deviation directly measures both the read noise (in the bias frame) and the combined read noise and photon shot noise (in the light frames) and that's why we use standard deviation rather than average or median.

 

Mark

That is very interesting. Thank you for detailed explanation Mark!

 

 

My usual subs are 5 min

Thanks Pinzmann!



#18 erictheastrojunkie

erictheastrojunkie

    Messenger

  • *****
  • Posts: 497
  • Joined: 15 Jul 2016
  • Loc: Salt Lake City

Posted 20 November 2019 - 05:33 PM

From my bortle 5/~19.8 mag skies and with an Astrotech AT65EDQ (f6.5) I'm shooting 4-5 minute subs at ISO 400 with the Optolong L-pro filter, 10 minute subs at ISO 400 with the Optolong L-eNhance filter. Here's a recent Pleiades with the L-pro, 87 x 5 minutes:

 

49061425676_fed0f72456_h.jpgPleiades by Eric and Bry, on Flickr

 

A recent 2 panel mosaic of the North America and Pelican Nebula with the L-eNhance, each panel is 50 x 10 minutes:

 

49058116991_02e876180a_h.jpgNorth America and Pelican Nebula Mosaic by Eric and Bry, on Flickr



#19 Startex

Startex

    Vostok 1

  • -----
  • Posts: 147
  • Joined: 23 Dec 2018
  • Loc: Texas

Posted 21 November 2019 - 06:12 AM

Shoot so the peak of the back-of-camera histogram is 1/4 from the left-hand-side.  This advice works for all ISOs, all f-ratios and all light pollution levels.

 

Mark

So are all histograms similar? I use DigiCamControl with my 5300 and use that histogram.



#20 sharkmelley

sharkmelley

    Mercury-Atlas

  • *****
  • Posts: 2512
  • Joined: 19 Feb 2013

Posted 21 November 2019 - 07:01 AM

So are all histograms similar? I use DigiCamControl with my 5300 and use that histogram.

No all histograms are not similar.  The back of camera (BoC) histogram is the histogram of the JPG file that the camera calculates internally (whether or not JPG is written to the memory card).  It is therefore the histogram of white-balanced gamma-stretched non-linear data and is similar to the histogram you would see in Photoshop when you open the raw file which again is non-linear. 

 

But if you open the raw file in PixInsight then the histogram you see is the histogram of raw data linear values because the white-balancing and colour space gamma has not been applied - this is probably true of most astro-processing applications. The histogram will therefore look completely different to the BoC histogram.

 

I'm not familiar with DigiCamControl so I don't know which histogram it will display but most DSLR acquisition software will display the BoC histogram or something very similar.  If you compare it to the BoC histogram then you will soon know for certain.

 

It's potentially confusing but it's always important to know exactly what kind of histogram you are looking at.

 

Mark


Edited by sharkmelley, 21 November 2019 - 07:32 AM.

  • khobar likes this


CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics