Exposure time question

[carl20171]

Hello, I see a lot of people taking 300 sec photos and stacking them. Maybe 10 or 20 of them.

1) Can I get the same data from 12 x 300 sec = 60 minutes exposure as from 20 x 180 sec = 60 minutes of  exposure or 60 x 60 seconds = 60 minutes of exposure?

2)Is there some reason not to take 30 minute exposures? Camera saturation or the possibility of a cloud our airplane? 

[nmoushon]

Short answer is no. But there is a lot that you have to take into effect when trying to calculate everything out. Long exposures are always better BUT only better if your local conditions allow it. Local conditions being your local LP, your local seeing and transparency, your local weather, ect. IF you have the perfect location and skies then longer exposures will always win out. This is for 95% of the objects out there though. Onjects like globular clusters and other very very bright objects will get burnt out after a certain exposure. There are several ways to get around this but thats for a different topic.

You are striving to get the best Signal-to-Noise (S/N) ratio you can. I see you have a CCD and not a DSLR so this also improve the S/N and max exposure length. Now I'm not 100% sure on the math so I will stay away from numbers and just give you a general notion of what I know. Someone else can come by and post a more detailed and in-depth response if they feel like.

Think of your exposure time as allowing you to get your images to go deeper and deeper. This will allow you to get the very faint stuff that is out there as well as pull more detail out of the brighter stuff. (Again this is all limited by your local conditions)

[carl20171]

Seeing conditions being equally good. Would it be that the pixel will not get saturated deep enough in one minute compared to 5 or 10 minutes? I am wondering what if I took 100 x 60 seconds would give me 100 minutes of data. Would I still not have the detailed data compared to lets say only 60 minutes of data that was 12 x 600 seconds? The reason I am asking is with the hyperstar and being around f2 I can get a lot of light in a short time and around 45 seconds it's really sharp. Why not just take 100 x 45 second photos and not use the auto guider etc. The mounts inaccuracy may even act as a form of dithering. Although my real question is still about the exposure time more than the autoguide comment. 

[Davey-T]

You should join the Fastar Yahoo group for expert advice.

Dave

[Stargazer33]

Using your Hyperstar at F2 you will be able to gather the same amount of detail in less time than, say an F10 scope. Although there is no difference in the overall time between doing 60 x 60 seconds or 6 x 600 seconds, your 60" exposures will not have time to acquire the faint parts of your target where as the 600" exposures will. Stacking 60 60 second images will not give you the depth of detail that the 600 second images will.

Think of it like dark adapting your eyes. Going outside from a brightly lit house you can only see the brightest stars for the first minute or so (short exposure). Then as your eyes are allowed to get used to the dark (longer exposure), you see fainter and fainter objects.

HTH

[nmoushon]

Seeing conditions being equally good. Would it be that the pixel will not get saturated deep enough in one minute compared to 5 or 10 minutes? I am wondering what if I took 100 x 60 seconds would give me 100 minutes of data. Would I still not have the detailed data compared to lets say only 60 minutes of data that was 12 x 600 seconds? The reason I am asking is with the hyperstar and being around f2 I can get a lot of light in a short time and around 45 seconds it's really sharp. Why not just take 100 x 45 second photos and not use the auto guider etc. The mounts inaccuracy may even act as a form of dithering. Although my real question is still about the exposure time more than the autoguide comment. 

Bryans analogy is a good one to help understand how longer exposures allow you to go deeper. It also increases your S/N ratio which will really help bring the noise in your image down as well. Also if it help think of what the professionals do. If stack lots of short exposures really did equal a much longer exposure everyone would be doing it. That would make AP WAAAAAAYYYYYY easier of a hobby! No one would be using guiding and everyone would have a 5000mm FL scope take 2 sec exposure all night long. I wish it would work that way but unfortunately physics wont let it happen.

If you are interested in the Fastar(Hyperstar) system I would suggest REALLY doing a lot of research on it. I've seen some really good images produced from it but I know it takes a lot of effort and time to get them to work right. There are a lot things to consider if you go down this route. For instance you cant use a filter wheel with your mono CCD as it covers up to much aperture and degrades the image (and by the the purpose of the fastar). So most people use a OSC CCD or DSLR with the Fastar system, which is a step down from mono imaging in terms of final image quality. Now you can image with a mono you just have to take the camera off and switch filters for every filter. Which means more than likely youll have to take new darks and lights for each filter.

[carl20171]

Thanks, I do already have the hyperstar though. I love it. It took a bit to get it collimated then it became very sharp. I do use a filter tray that fits right in the hyperstar so I just got 4 trays for the NB filters so I can keep the trays loaded and just slip them in and out. I also use the Atik 490 mono which is actually slimmer than the hyperstar. So far its been a great combo. I also got an adapter so I can use the filter trays on the back of the scope as well but I have not used it that way yet since I like the view of the hyperstar. It becomes an 11 inch with a great f stop. I didnt do the math but I think its around f2.2    I will get some picts tonight since we are both in Northern Virginia the sky is clear tonight finally. 

Thanks Carl

[sharkmelley]

Hi Carl,

12x300sec collects the same amount of data (i.e. photons) as 60x60sec.  The main difference between the two approaches is the read noise.  Doing 60x60 instead of 12x300 means you are taking 5x more exposures so you are acquiring 5x as much read noise.  However, the good news is the read noise is generally random so it aggregates quadratically and the actual effect in the final stacked image is that you will have 2.23x the noise (since 2.23 is the square root of 5).  Since you are using a cooled mono camera in narrowband then read noise is likely to be the biggest component of noise in your image, so an image with 5x as many exposures i.e. 2.23x as much read noise will be quite noticeable to the amount of faint detail you can extract.

There is no reason not to take 30min exposures (as long as you can guide sufficiently accurately) unless your target or background saturates.  But take a sufficient number that you can perform a decent sigma stack to remove cosmic ray hits, satellite trails etc. 

Mark

[Fordos Moon]

I use 300 seconds in the main for my modified Canon 1100D for 3 reasons.

1. At ISO800 it gives me a histogram about a third along the x-axis.

2. Olly Penrice said that the DSLR guys he works with go to a max of 8 minutes with no light pollution so I like to bring it down a bit from that in the UK

3. Because I enjoy sitting out in my observatory and looking at every subframe that comes in I find 5 minutes is a nice amount of time to wait between exposures ;-)

[Gina]

I do a few test subs to find the longest time I can image before I get saturation then take as many of those as the sky allows.  The exposures usually come out as much the same for Ha, OIII and SII using Astrodon 5nm Ha, 3nm OIII and 5nm SII.  It varies a bit from night to night and between different FOVs but as an example with f5 unbinned around 10m 600s and binned 2x2 (when seeing is poorer) half that.  That's using either the SW Esprit 80ED with 400mm FL or MN190 with 1000mm FL @ f5.3 and using the Atik 460EX mono CCD camers.  Bright objects like M42 only allow very short exposures for the centre part but I combine these with longer subs for the fainter stuff, then combine in Photoshop using layers and masking.  This object takes some careful processing.

[ollypenrice]

You'll get the best results by making subs as long as possible. What defines 'possible?'

-Skyglow taking the histogram peak past the 1/3 point.

-Guiding error. 

-Thermal noise in uncooled cameras.

As I often say on here, the need for short subs to control saturation is exaggerated. This is rarely a problem at a dark site. If you are chasing faint stuff, as we often are, long subs are certainly best but 'long' in a Hyperstar may not be all that long! Experimentation is everything.

Olly

[carl20171]

Ahh. The read noise. I figured there was a reason taking as long an exposure as possible for the target. Thanks for all the advice. I took 12x300 subs of NGC 7635  (bubble nebula) in Ha last night and the difference in detail from 12x60 is amazing. 

[nmoushon]

Thanks, I do already have the hyperstar though. I love it. It took a bit to get it collimated then it became very sharp. I do use a filter tray that fits right in the hyperstar so I just got 4 trays for the NB filters so I can keep the trays loaded and just slip them in and out. I also use the Atik 490 mono which is actually slimmer than the hyperstar. So far its been a great combo. I also got an adapter so I can use the filter trays on the back of the scope as well but I have not used it that way yet since I like the view of the hyperstar. It becomes an 11 inch with a great f stop. I didnt do the math but I think its around f2.2    I will get some picts tonight since we are both in Northern Virginia the sky is clear tonight finally. 

Thanks Carl

Well since you already have the hyperstar and have used it all you can do now is just experiement and test what works best. I'm in the Winchester area and even though we (in northern virginia) don't have the best skies around my LP will limit my exposure length before my guiding cant keep up. 10min subs is about as ideal as I can get from my back yard though I can easily do 15min subs. And that me shooting at F/4.75. Since you are shooting at F2.2 (depending on where you are exactly) you might only be able to do 5 min subs before your LP will start washing everything out. Even though at that FL you could probably guide from at least 15min subs. Thats where the experiementing comes in for you and you'll just have to spend a night or two playing with exposure length and watching your histagram.

[carl20171]

Sounds good, Thanks, I did take some 5 min subs last night but they dont look real good to me. I need more practice for sure. 

[dph1nm]

If stack lots of short exposures really did equal a much longer exposure everyone would be doing it. That would make AP WAAAAAAYYYYYY easier of a hobby! No one would be using guiding and everyone would have a 5000mm FL scope take 2 sec exposure all night long. I wish it would work that way but unfortunately physics wont let it happen.

Physics doesn't stop you  - there are chips out there which enable you to do this (essentially they have  zero read noise), although possibly not at a price the average amateur could afford! It might be the future though ...

NigelM

[nmoushon]

Physics doesn't stop you  - there are chips out there which enable you to do this (essentially they have  zero read noise), although possibly not at a price the average amateur could afford! It might be the future though ...

NigelM

The physics of money then.