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

Choosing Exposure Time

  • Please log in to reply
24 replies to this topic

#1 r3g

r3g

    Sputnik

  • -----
  • topic starter
  • Posts: 28
  • Joined: 20 Mar 2020
  • Loc: Bay Area, California

Posted 26 March 2020 - 11:44 AM

So my first AP setup is out for delivery and the forecast looks good for getting in some capturing today. I've been doing a lot of research and have landed on shooting M81 Bode's Galaxy because it is close to Polaris and will be starting pretty high in the sky giving me a good amount of time to image it. What I'm wondering is what is a good starting point in deciding my exposure times. Once I have my focus and framing good, snap a single frame and very I can see the object, where do I go from there? How do I decide if I'm going to do 120x60" or 24x300" etc? How do I decide what exposure scheme I'm going to use for a given period of time (say 2 hours).

 

 

 

 

Gear:

WO z73 APO

ASI183MC, 

EQ6R

WO 55mm guidscope with ASI290MM

 


  • tleroy1 likes this

#2 B 26354

B 26354

    Apollo

  • -----
  • Posts: 1,209
  • Joined: 05 Jan 2017
  • Loc: Southern California semi-desert (NELM mag 5.3)

Posted 26 March 2020 - 11:49 AM

https://www.amazon.c.../dp/0999470906/

 

https://www.amazon.c.../dp/1138055360/

 

https://www.amazon.c.../dp/1937538435/


  • tleroy1 likes this

#3 Madratter

Madratter

    Voyager 1

  • *****
  • Posts: 11,964
  • Joined: 14 Jan 2013

Posted 26 March 2020 - 11:50 AM

What software are you going to be using for your data acquisition?

 

If this is your first time out with your gear ever, I would suggest erring on the short side. 30" subs are a good starting place for many situations. You'll probably get at least something to show for your efforts.


  • tleroy1 likes this

#4 kathyastro

kathyastro

    Apollo

  • *****
  • Posts: 1,422
  • Joined: 23 Dec 2016
  • Loc: Nova Scotia

Posted 26 March 2020 - 12:05 PM

Start at 30 seconds until you get comfortable there.  Then move up to 60 seconds if everything looks good and you have no star trails.  Go longer if your sky conditions and tracking/guiding can handle it.


  • tleroy1 likes this

#5 r3g

r3g

    Sputnik

  • -----
  • topic starter
  • Posts: 28
  • Joined: 20 Mar 2020
  • Loc: Bay Area, California

Posted 26 March 2020 - 12:22 PM

What software are you going to be using for your data acquisition?

 

If this is your first time out with your gear ever, I would suggest erring on the short side. 30" subs are a good starting place for many situations. You'll probably get at least something to show for your efforts.

Astro Photography tool and PHD2. With exposures that should would it be worth while to still shoot for around 2 hours or will a shorter period do since I'll have a large number of frames?

 

 

 

Purchased. Won't see them for weeks though unfortunately given the current situation. Thank you!

 

 

 

 

Start at 30 seconds until you get comfortable there.  Then move up to 60 seconds if everything looks good and you have no star trails.  Go longer if your sky conditions and tracking/guiding can handle it.

Will do!


  • tleroy1 likes this

#6 Stelios

Stelios

    Cosmos

  • *****
  • Moderators
  • Posts: 8,459
  • Joined: 04 Oct 2003
  • Loc: West Hills, CA

Posted 26 March 2020 - 12:31 PM

The most important thing is how dark your skies are. If they are bright (Bortle 6 or worse) start with 30" as Kathy said. If they are dark (Bortle 4 or better) I would program a small sequence of 1x 90", 1x 120" and 1x180". Then use the longest exposure that doesn't show clipping (clipping is when--WITHOUT stretching--a lot of stars or any of the galaxy image are visible).

 

Total exposure time does NOT change because of length of exposure. You should aim for at LEAST 2 hours (much more from a light polluted site), and it doesn't matter if it is 120x60" or 60x120" or 40x180"


  • mtc, dghent and tleroy1 like this

#7 r3g

r3g

    Sputnik

  • -----
  • topic starter
  • Posts: 28
  • Joined: 20 Mar 2020
  • Loc: Bay Area, California

Posted 26 March 2020 - 12:39 PM

You should aim for at LEAST 2 hours (much more from a light polluted site), and it doesn't matter if it is 120x60" or 60x120" or 40x180"

This is a big thing I was wondering. I have Bortle 5 sky here and also have a quad band filter on the way to help further with light pollution. 

 

 

Just so that I'm understanding you correctly... You're saying I want the longest exposure that doesn't have a lot of visible objects before a stretch? I take it that is because if the objects are visible beforehand they'll be overexposed after stretching?


  • tleroy1 likes this

#8 bobzeq25

bobzeq25

    Hubble

  • *****
  • Posts: 18,973
  • Joined: 27 Oct 2014

Posted 26 March 2020 - 01:30 PM

Others suggestions of specific times mean absolutely _nothing_ to you.  Nothing at all.  Zip.  For proof, scan these tables.  See how much things vary by specific situation.  More than an order of magnitude.  You _must_ do this on a site specific basis.

 

https://www.cloudyni...olour-versions/

 

Here's a very decent method for the 183 (I have both of them).

 

Take a light and a bias.  Measure the ADU with something like PixInsight, IRIS, ...  Either the average or the obvious peak, doesn't matter.  Subtract, get the corrected ADU of the light.

 

Using the ZWO website change ADU to electrons.

 

Also using the ZWO website get your read noise and square it.

 

You want the first number to be 5-10 X the second.

 

Described in far more detail in the second book in post #2.

 

There are other methods.  Any method using site specific data beats the tar out of copying someone else's number.  You must do this on a site specific basis.

 

My image below is something like 650X10".  Has utterly no relevance to you.

 

This does.  My rule of thumb is one hour minimum total, 2 is better, 4 good.

 

Secondary point.  For right now, use just one exposure for M42.  Using multiple exposures is an advanced technique.  A common problem on CN is experienced imagers stating how they'd do something.   Their concerns are not yours.  For right now, keep this simple, you needn't concern yourself with high dynamic range processing.

 

Pleadies 2019 V3 small.jpg


Edited by bobzeq25, 26 March 2020 - 04:21 PM.

  • dghent, ks__observer and tleroy1 like this

#9 ks__observer

ks__observer

    Apollo

  • *****
  • Posts: 1,148
  • Joined: 28 Sep 2016
  • Loc: Long Island, New York

Posted 26 March 2020 - 04:02 PM

https://www.cloudyni...olour-versions/


  • tleroy1 likes this

#10 tleroy1

tleroy1

    Vostok 1

  • -----
  • Posts: 174
  • Joined: 04 May 2018
  • Loc: Georgetown, Ontario, Canada

Posted 26 March 2020 - 04:40 PM

I've ordered "The Deep-sky Imaging Primer, Second Edition" and "The Astrophotography Manual: A Practical and Scientific Approach to Deep Sky Imaging." Both shipped on 23 March and are scheduled for delivery sometime in the next two months.


  • bobzeq25 and B 26354 like this

#11 tleroy1

tleroy1

    Vostok 1

  • -----
  • Posts: 174
  • Joined: 04 May 2018
  • Loc: Georgetown, Ontario, Canada

Posted 26 March 2020 - 04:41 PM

Others suggestions of specific times mean absolutely _nothing_ to you.  Nothing at all.  Zip.  For proof, scan these tables.  See how much things vary by specific situation.  More than an order of magnitude.  You _must_ do this on a site specific basis.

 

https://www.cloudyni...olour-versions/

 

Here's a very decent method for the 183 (I have both of them).

 

Take a light and a bias.  Measure the ADU with something like PixInsight, IRIS, ...  Either the average or the obvious peak, doesn't matter.  Subtract, get the corrected ADU of the light.

 

Using the ZWO website change ADU to electrons.

 

Also using the ZWO website get your read noise and square it.

 

You want the first number to be 5-10 X the second.

 

Described in far more detail in the second book in post #2.

 

There are other methods.  Any method using site specific data beats the tar out of copying someone else's number.  You must do this on a site specific basis.

 

My image below is something like 650X10".  Has utterly no relevance to you.

 

This does.  My rule of thumb is one hour minimum total, 2 is better, 4 good.

 

Secondary point.  For right now, use just one exposure for M42.  Using multiple exposures is an advanced technique.  A common problem on CN is experienced imagers stating how they'd do something.   Their concerns are not yours.  For right now, keep this simple, you needn't concern yourself with high dynamic range processing.

 

Really informative bobzeq25! Great input from many others too.



#12 tleroy1

tleroy1

    Vostok 1

  • -----
  • Posts: 174
  • Joined: 04 May 2018
  • Loc: Georgetown, Ontario, Canada

Posted 26 March 2020 - 05:25 PM

madratter's site has great info for us beginners: https://astroimages.....com/about.html

 

Pixinsight tutorial and info about his hardware among other informative stuff.

 

I got it from your signature madratter.

 

Thanks for making a useful site and documenting what you've learned!

 

You may find it useful r3g.

 

Kind regards,

 

Ted


  • Madratter likes this

#13 17.5Dob

17.5Dob

    Fly Me to the Moon

  • ****-
  • Posts: 5,879
  • Joined: 21 Mar 2013
  • Loc: Colorado,USA

Posted 26 March 2020 - 06:51 PM

This is a big thing I was wondering. I have Bortle 5 sky here and also have a quad band filter on the way to help further with light pollution. 

 

Be aware that the quad band filter is only useful on emission nebula and will actually do a lot of harm if you try to use it on galaxies, reflection nebula, or star clusters.
 


Edited by 17.5Dob, 26 March 2020 - 06:52 PM.

  • bobzeq25 likes this

#14 r3g

r3g

    Sputnik

  • -----
  • topic starter
  • Posts: 28
  • Joined: 20 Mar 2020
  • Loc: Bay Area, California

Posted 26 March 2020 - 07:48 PM

Be aware that the quad band filter is only useful on emission nebula and will actually do a lot of harm if you try to use it on galaxies, reflection nebula, or star clusters.
 

I'm aware, thanks!

 

Wont be doing any imaging tonight everything arrived except the tripod and counter-balance weights :/



#15 bulrichl

bulrichl

    Mariner 2

  • -----
  • Posts: 233
  • Joined: 27 May 2018
  • Loc: La Palma (Canary Islands)

Posted 27 March 2020 - 05:47 AM

Here's a very decent method for the 183 (I have both of them).

 

Take a light and a bias.  Measure the ADU with something like PixInsight, IRIS, ...  Either the average or the obvious peak, doesn't matter.  Subtract, get the corrected ADU of the light.

 

Using the ZWO website change ADU to electrons.

 

Also using the ZWO website get your read noise and square it.

 

You want the first number to be 5-10 X the second.

 

Described in far more detail in the second book in post #2.

The approach that you recommend is very helpful. However, I think there is a little flaw: The IMX183 has a 12-bit ADC, i.e. the original intensity values are in the range of 0 to 2^12 = 4096. These values are scaled to 16 bit (multiplied by factor 16) before they are written to the FITS file. In the graph on the ZWO website displaying gain [e/ADU] vs. gain setting (see appended JPG), the values refer to the NOT scaled intensity values in ADU.

 

So after subtraction of the bias from the light, you will have to divide by factor 16 in order to get a correct result.
 

Bernd

Attached Thumbnails

  • 183-Gain-vs-gain.jpg


#16 dghent

dghent

    Ranger 4

  • *****
  • Vendors
  • Posts: 389
  • Joined: 10 Jun 2007

Posted 27 March 2020 - 11:57 AM

Plenty of good answers regarding your question already, but I'd like to add this: If this is your first setup and you think you're going to get under the skies with it immediately, think again - and I mean that in a concerned way.

 

There's nothing worse than getting under a clear sky with your gear for the first time, in the dark, and hurriedly trying to make a complex assemblage of gear do things in coordination and do those things correctly. Stuff will go wrong, there will be unforeseen needs or unaccounted for situations, and you will be making yourself deal with that under the stress of wanting to be productive. Don't do that to yourself.

 

Set everything up in your house or out in the yard during day time first. Go through the motions. Make sure you know how to do the basics, like initialize your mount and that your chosen software is talking to it, and it reacts in the expected ways. Make sure your camera settings are all there, that you have long enough cables, and all that stuff. Dry-run it. Even in the daytime. Get some confidence to step off with, because too many people have not done that and dove right in, got frustrated because they thought it would be easier, and ended up giving up because they just didn't take their time.


Edited by dghent, 27 March 2020 - 11:59 AM.

  • kathyastro and B 26354 like this

#17 r3g

r3g

    Sputnik

  • -----
  • topic starter
  • Posts: 28
  • Joined: 20 Mar 2020
  • Loc: Bay Area, California

Posted 27 March 2020 - 12:20 PM

Plenty of good answers regarding your question already, but I'd like to add this: If this is your first setup and you think you're going to get under the skies with it immediately, think again - and I mean that in a concerned way.

 

There's nothing worse than getting under a clear sky with your gear for the first time, in the dark, and hurriedly trying to make a complex assemblage of gear do things in coordination and do those things correctly. Stuff will go wrong, there will be unforeseen needs or unaccounted for situations, and you will be making yourself deal with that under the stress of wanting to be productive. Don't do that to yourself.

 

Set everything up in your house or out in the yard during day time first. Go through the motions. Make sure you know how to do the basics, like initialize your mount and that your chosen software is talking to it, and it reacts in the expected ways. Make sure your camera settings are all there, that you have long enough cables, and all that stuff. Dry-run it. Even in the daytime. Get some confidence to step off with, because too many people have not done that and dove right in, got frustrated because they thought it would be easier, and ended up giving up because they just didn't take their time.

 

My plan was always to set everything up inside and get to know it before heading outside. While I was by no means expecting to master anything I figured a 4 or so hours I had with the gear before nightfall would be a good enough starting point. Plan never went through because the tripod and counter-weights were shipped separately and wont arrive until later today and of course the next clear night isn't expected for a few days. I appreciate the concern.



#18 Noah4x4

Noah4x4

    Gemini

  • *****
  • Posts: 3,053
  • Joined: 07 Apr 2016
  • Loc: Colchester UK

Posted 27 March 2020 - 12:23 PM

Two answers suggest an hour or longer.  Whilst I don't disagree if you want to win Astrophotographer of the Year, I think it's the wrong place to start in "Beginning and intermediate imaging". I would start by using short stacked exposures and completely forget long exposure AP until one's skills develop. 

 

Folks using a camera for the first time will be wowed merely by embracing 20 x 20 second stacks given that a majority have, at best, only previously seem a 'faint fuzzies' through an eyepiece. This will overcome issues with field rotation and other mechanical challenges that can be tackled later.  The ease of getting those first few pictures is important. I nearly gave up with a DSLR and unforgiving camera software.

 

Another reason for starting with stacked short exposures. If I tried to capture one hour exposures under my UK urban skies, the chances are that clouds or aircraft trails would scupper me. 



#19 bobzeq25

bobzeq25

    Hubble

  • *****
  • Posts: 18,973
  • Joined: 27 Oct 2014

Posted 27 March 2020 - 12:30 PM

The approach that you recommend is very helpful. However, I think there is a little flaw: The IMX183 has a 12-bit ADC, i.e. the original intensity values are in the range of 0 to 2^12 = 4096. These values are scaled to 16 bit (multiplied by factor 16) before they are written to the FITS file. In the graph on the ZWO website displaying gain [e/ADU] vs. gain setting (see appended JPG), the values refer to the NOT scaled intensity values in ADU.

 

So after subtraction of the bias from the light, you will have to divide by factor 16 in order to get a correct result.
 

Bernd

That is incorrect.  It works _exactly_ as I said, there is _no_ flaw.  There is _no_division by 16.  I simply do everything in 12 bit.  When you convert to electrons, the bit depth is no longer relevant.

 

If someone did everything in a different bit depth, it would still work.  You just have to be consistent.  If you use 12 bit one place, and 16 bit another, yes, you'll have problems.


Edited by bobzeq25, 27 March 2020 - 02:07 PM.


#20 r3g

r3g

    Sputnik

  • -----
  • topic starter
  • Posts: 28
  • Joined: 20 Mar 2020
  • Loc: Bay Area, California

Posted 27 March 2020 - 12:33 PM

Two answers suggest an hour or longer.  Whilst I don't disagree if you want to win Astrophotographer of the Year, I think it's the wrong place to start in "Beginning and intermediate imaging". I would start by using short stacked exposures and completely forget long exposure AP until one's skills develop. 

 

Folks using a camera for the first time will be wowed merely by embracing 20 x 20 second stacks given that a majority have, at best, only previously seem a 'faint fuzzies' through an eyepiece. This will overcome issues with field rotation and other mechanical challenges that can be tackled later.  The ease of getting those first few pictures is important. I nearly gave up with a DSLR and unforgiving camera software.

 

Another reason for starting with stacked short exposures. If I tried to capture one hour exposures under my UK urban skies, the chances are that clouds or aircraft trails would scupper me. 

 

I definitely will start with shorter exposures. Since I basically have everything except the complete mount I have already put the scopes and cameras together and have been learning APT and PHD2. Currently making a library of darks frames at different exposures. Very convenient having cooling on the CCD.



#21 bobzeq25

bobzeq25

    Hubble

  • *****
  • Posts: 18,973
  • Joined: 27 Oct 2014

Posted 27 March 2020 - 12:40 PM

Two answers suggest an hour or longer.  Whilst I don't disagree if you want to win Astrophotographer of the Year, I think it's the wrong place to start in "Beginning and intermediate imaging". I would start by using short stacked exposures and completely forget long exposure AP until one's skills develop. 

 

Folks using a camera for the first time will be wowed merely by embracing 20 x 20 second stacks given that a majority have, at best, only previously seem a 'faint fuzzies' through an eyepiece. This will overcome issues with field rotation and other mechanical challenges that can be tackled later.  The ease of getting those first few pictures is important. I nearly gave up with a DSLR and unforgiving camera software.

 

Another reason for starting with stacked short exposures. If I tried to capture one hour exposures under my UK urban skies, the chances are that clouds or aircraft trails would scupper me. 

You're misunderstanding the difference between subexposure time, and total imaging time.

 

Yes, you take short exposures and stack them.   Everyone does.  The question is, for a given amount of total imaging time, do you take more shorter ones or fewer longer ones?  That's what this discussion is about.  Not about doing hourlong subs (which would almost never be correct).  The hour means doing an hour of _total_ imaging time.

 

So, once you have the subexposure about right _for your specific situation_, you shoot a lot of subs.  7 minutes (20X20") of data will get you something.  But an hour will get you something to work with in processing.

 

This isn't about going for IOTD.  My advice never is, I leave that to others.  It's about learning to do astrophotography.   Unless you have adequate data to process (including bias, flats, darks), you won't learn how to process properly.  It's that simple.


Edited by bobzeq25, 27 March 2020 - 12:46 PM.


#22 bulrichl

bulrichl

    Mariner 2

  • -----
  • Posts: 233
  • Joined: 27 May 2018
  • Loc: La Palma (Canary Islands)

Posted 27 March 2020 - 04:31 PM

Take a light and a bias.  Measure the ADU with something like PixInsight, IRIS, ...  Either the average or the obvious peak, doesn't matter.  Subtract, get the corrected ADU of the light.

I don't doubt that you do your calculations correctly. My concern is that a PixInsight beginner possibly will not get the correct intensity values (in ADU) following this description.
 

That is incorrect.  It works _exactly_ as I said, there is _no_ flaw.  There is _no_division by 16.  I simply do everything in 12 bit.  When you convert to electrons, the bit depth is no longer relevant.

In PixInsight, the setting in Statistics has to be set to '12-bit [0,4095]'. I guess that was meant by "I simply do everything in 12 bit.", and in this case you are completely correct. However this hint was not given in your first post (of 26 Mar).

 

My recommendation (divide the difference by 16) was meant for the setting '16-bit [0,65535]' in Statistics (and I also forgot to mention that).

 

Bernd



#23 r3g

r3g

    Sputnik

  • -----
  • topic starter
  • Posts: 28
  • Joined: 20 Mar 2020
  • Loc: Bay Area, California

Posted 27 March 2020 - 05:10 PM

The question is, for a given amount of total imaging time, do you take more shorter ones or fewer longer ones?


This is what I was trying to ask from the start.

Edited by r3g, 27 March 2020 - 05:10 PM.


#24 bobzeq25

bobzeq25

    Hubble

  • *****
  • Posts: 18,973
  • Joined: 27 Oct 2014

Posted 27 March 2020 - 05:22 PM

This is what I was trying to ask from the start.

I understood, from the start.  Did post #8 clear that up for you?   Note subsequent posts about being sure to use the same bit depth everywhere in the calculation.

 

Bottom line.  Subexposure is not absolutely critical, although it must be determined with site specific data.  Getting the skyfog to 5 - 10X the read noise squared (ie, within a factor of 2) is "good enough", although there are some refinements advanced imagers do, described in the second book of post #2.   I can't possibly summarize several pages in a short post here.  But they're just refinements.

 

At that point (subexposures close to optimal), subexposure time decreases in importance relative to total imaging time.  Total imaging time becomes the key thing for both image quality and capturing dim detail.  More is always better, although it's only a square root relationship.  To improve things by about a factor of 2, you need 4 times as much total imaging time.

 

Questions?


Edited by bobzeq25, 27 March 2020 - 05:27 PM.


#25 r3g

r3g

    Sputnik

  • -----
  • topic starter
  • Posts: 28
  • Joined: 20 Mar 2020
  • Loc: Bay Area, California

Posted 27 March 2020 - 05:31 PM

I understood, from the start.  Did post #8 clear that up for you?   Note subsequent posts about being sure to use the same bit depth everywhere in the calculation.

 

Bottom line.  Subexposure is not absolutely critical, although it must be determined with site specific data.  Getting the skyfog to 5 - 10X the read noise squared (ie, within a factor of 2) is "good enough", although there are some refinements advanced imagers do, described in the second book of post #2.   I can't possibly summarize several pages in a short post here.  But they're just refinements.

 

At that point (subexposures close to optimal), subexposure time decreases in importance relative to total imaging time.  Total imaging time becomes the key thing for both image quality and capturing dim detail.  More is always better, although it's only a square root relationship.  To improve things by about a factor of 2, you need 4 times as much total imaging time.

 

Questions?

 

 

I think I understand post #2, as soon as Imy tripod comes and I can actually run through the process it will probably click more then. I definitely understand there is no one size fits all approach I was just looking for a good place to start and I've got that now. 


  • bobzeq25 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