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

Looking for Canon R5 ISO suggestions

  • Please log in to reply
4 replies to this topic

#1 stevejbo

stevejbo

    Lift Off

  • -----
  • topic starter
  • Posts: 17
  • Joined: 16 Sep 2020

Posted 02 December 2021 - 12:23 AM

Good evening

 

I have an old Celestron Celestar 8 that I would take out and do some viewing periodically. Yes I know I need a better scope - it is on my list of things to buy someday. I use it for small DSOs but mainly planet obs and some photography of the planets.  I recently used all my loose cash to buy a new Canon R5 (for nature and family photography) and decided to do some DSO astrophotography with it as well. I started out with the obvious - Andromeda, N. American Nebula, Pleidies, Orion, etc.  The Celestron is not going to work with its small field of view (it can just about fit the moon when I attach with a T-ring) but since it has a motor drive I polar align it - mount my camera on top with a 70-300 Canon Ultrasonic zoom lens (1:4 - 5.6) and point it at the DSO and set it to take photos for a couple hours. (I use a diffraction mask to focus) I can go 15-30 second exposures without startrails depending on the alignment.

I take a bunch of Dark, Dark Flat, and Flat frames - get rid of any bad ones - then use DSS to stack them.

According to clearoutside.com (an amazingly useful website for planning) I am under Class 7 skies (sigh) so there is a lot of light pollution. I deal with this in PS with a great plugin tool that I found related to the "Astronomy Tools Action Set". I go through the processing in PS according to the AstroBackyard Imaging Processing Guide.

And the results are.......meh

 

I include an example - - 280 minute exposure. 

 

Clearly I am doing something wrong - maybe a couple things. I chatted with someone I know who gets much better results and here are the culprits we have come up with:

 

1) Using a zoom lens means going through a lot of glass. I should be using a small refractor that is made to take DSOs. Of course I would need a mount to go with it so we are talking $2k - $3k at least. And since I just spent a tidy sum on a new Canon that is going to be a hard sell with my wife. This may be it but I have seen other folks on the internet that have taken DSOs with zoom lenses and they seem to do much better than me. So I want to make sure it is not a different issue. Do you think this is what is holding me back? 

 

2) I am really really bad at PS. Or at least there is something I am doing in the process that is overdoing it and I need a much gentler touch. Very possible. But I have fiddled around with the example files from AstroBackyard and gone through his process and they don't come out as good as his - but not too bad. So I don't think it is #2 - but I DO think I am screwing something else up - and in doing so I try to correct it in PS and I because the input is flawed the PS work goes wonky trying to fix it and make it look good.

 

3) My exposure of 30 seconds is too long. Maybe but I see pretty small star trails so I don't think this is it.

 

4) I am messing up DSS.  Maybe....

 

5) This is my best guess - I am using too high an ISO.  This picture uses an ISO of 6400.  My thinking here is - high ISO more signal. Yes I know it also means more noise and a hotter sensor. My thinking is that the Canon goes to a crazy ISO of 52,000 or something and visual inspection of 6400 doesn't seem too noisy. And the alignment is not perfect so the DSO and stars are moving slightly from one frame to another so when I stack it I subtract the flats and darks and dark flats so that gets rid of the noise on the same pixels and the slowly shifting DSO & stars are on new pixels - so when you stack them the hot pixels will disappear (I think this is called dithering).

This sounds like it should work - but I look at what others use and they seem to all be using ISOs below 1000. But I am guessing that by using an ISO of 6400 I have too much noise and also too much light pollution and subtracting that out is helping kill my image. 

 

So I am guessing - even though I have a Canon R5 with an amazing fast sensor I should bring my ISO way down. So my questions are - - which of the above do you think is the main problems? And if as I suspect it is #5 - what ISO should I be using? If I am really lucky and there is a Canon R5 person out there - - what ISO do you use as I figure it is camera specific what ISO is too high.

 

Thanks for your patience with this rambling question.  I can add a photo of Andromeda which is even worse if that would help.

 

Steve

2021_11_04_AS_zoom_lowres.jpg



#2 bobzeq25

bobzeq25

    ISS

  • *****
  • Posts: 26,985
  • Joined: 27 Oct 2014

Posted 02 December 2021 - 01:49 AM

Good evening

 

I have an old Celestron Celestar 8 that I would take out and do some viewing periodically. Yes I know I need a better scope - it is on my list of things to buy someday. I use it for small DSOs but mainly planet obs and some photography of the planets.  I recently used all my loose cash to buy a new Canon R5 (for nature and family photography) and decided to do some DSO astrophotography with it as well. I started out with the obvious - Andromeda, N. American Nebula, Pleidies, Orion, etc.  The Celestron is not going to work with its small field of view (it can just about fit the moon when I attach with a T-ring) but since it has a motor drive I polar align it - mount my camera on top with a 70-300 Canon Ultrasonic zoom lens (1:4 - 5.6) and point it at the DSO and set it to take photos for a couple hours. (I use a diffraction mask to focus) I can go 15-30 second exposures without startrails depending on the alignment.

I take a bunch of Dark, Dark Flat, and Flat frames - get rid of any bad ones - then use DSS to stack them.

According to clearoutside.com (an amazingly useful website for planning) I am under Class 7 skies (sigh) so there is a lot of light pollution. I deal with this in PS with a great plugin tool that I found related to the "Astronomy Tools Action Set". I go through the processing in PS according to the AstroBackyard Imaging Processing Guide.

And the results are.......meh

 

I include an example - - 280 minute exposure. 

 

Clearly I am doing something wrong - maybe a couple things. I chatted with someone I know who gets much better results and here are the culprits we have come up with:

 

1) Using a zoom lens means going through a lot of glass. I should be using a small refractor that is made to take DSOs. Of course I would need a mount to go with it so we are talking $2k - $3k at least. And since I just spent a tidy sum on a new Canon that is going to be a hard sell with my wife. This may be it but I have seen other folks on the internet that have taken DSOs with zoom lenses and they seem to do much better than me. So I want to make sure it is not a different issue. Do you think this is what is holding me back? 

 

2) I am really really bad at PS. Or at least there is something I am doing in the process that is overdoing it and I need a much gentler touch. Very possible. But I have fiddled around with the example files from AstroBackyard and gone through his process and they don't come out as good as his - but not too bad. So I don't think it is #2 - but I DO think I am screwing something else up - and in doing so I try to correct it in PS and I because the input is flawed the PS work goes wonky trying to fix it and make it look good.

 

3) My exposure of 30 seconds is too long. Maybe but I see pretty small star trails so I don't think this is it.

 

4) I am messing up DSS.  Maybe....

 

5) This is my best guess - I am using too high an ISO.  This picture uses an ISO of 6400.  My thinking here is - high ISO more signal. Yes I know it also means more noise and a hotter sensor. My thinking is that the Canon goes to a crazy ISO of 52,000 or something and visual inspection of 6400 doesn't seem too noisy. And the alignment is not perfect so the DSO and stars are moving slightly from one frame to another so when I stack it I subtract the flats and darks and dark flats so that gets rid of the noise on the same pixels and the slowly shifting DSO & stars are on new pixels - so when you stack them the hot pixels will disappear (I think this is called dithering).

This sounds like it should work - but I look at what others use and they seem to all be using ISOs below 1000. But I am guessing that by using an ISO of 6400 I have too much noise and also too much light pollution and subtracting that out is helping kill my image. 

 

So I am guessing - even though I have a Canon R5 with an amazing fast sensor I should bring my ISO way down. So my questions are - - which of the above do you think is the main problems? And if as I suspect it is #5 - what ISO should I be using? If I am really lucky and there is a Canon R5 person out there - - what ISO do you use as I figure it is camera specific what ISO is too high.

 

Thanks for your patience with this rambling question.  I can add a photo of Andromeda which is even worse if that would help.

 

Steve

attachicon.gif2021_11_04_AS_zoom_lowres.jpg

The problem with high ISO is not noise.  It's the limited dynamic range.  The means bright stars are way oversaturated and even not so bright lose color.  That's a lot of what you're seeing there.

 

I recommend you replace DSS and PS with Astro Pixel Processor.  It does calibration/stacking/processing in one software, a serious advantage.  Has an excellent gradient reduction tool for reducing the effects of light pollution.  Has an embedded numbered workflow that actually teaches you processing.

 

This book explains many things, including why high ISO is bad.  Again, highly recommended.

 

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

 

This is a complicated business, and often unintuitive.  The unintuitive nature means people often concentrate on less important things.  Primes (fixed focal length) are better than zooms, but you've got bigger fish to fry right now.  <smile>



#3 John Tucker

John Tucker

    Surveyor 1

  • *****
  • Posts: 1,542
  • Joined: 03 Mar 2018

Posted 02 December 2021 - 09:58 AM

ISO 1600 is pretty standard for Canon camera in astrophotography.  

 

I think your individual subexposures are overexposed.  Your stars are bloated and that's a dead giveaway.  This isn't like terrestrial photography where you put the peak of your histogram in the middle. The contrast is extremely high and so you have to underexpose a bit so that the bright stars aren't burned out and bloated.   Set the peak about 15% to 20% from the left.  This will reduce the prominence and size of your stars relative to the nebula.

 

I agree with bob about the very high ISO being a bad idea due to lack of dynamic range.

 

The diffraction spikes around your stars are pretty normal for standard photography lenses but might be less prominent if you reduce exposure time

 

Disagree with Bob about APP.  For most things Deep Sky Stacker works pretty well. APP is "nice to have" but not critical.

 

I don't think its that bad of a picture given the ad hoc nature of your setup.  

 

I don't think you need to spend $2K to $3K.  Get yourself a SkyWatcher Star Adventurer https://www.amazon.c...2/dp/B07NY44782. They show up used on the classifieds here or at B&H Photo pretty much weekly for $400 or so.  Your zoom lens seems to work fine though you might want to get a prime lens at some point down the road.  

 

Lots of people on this site will answer every question with "go buy this $300 program, this $2000 mount and this $3000 telescope.  Oh, and replace the DSLR with a $1.5K cooled astro camera".  But lots of people take great photos with a DSLR, a regular camera lens, and an astrotracker.  


Edited by John Tucker, 02 December 2021 - 10:01 AM.


#4 unimatrix0

unimatrix0

    Apollo

  • -----
  • Posts: 1,327
  • Joined: 03 Jan 2021

Posted 02 December 2021 - 10:24 AM

 

5) This is my best guess - I am using too high an ISO.  This picture uses an ISO of 6400.  My thinking here is - high ISO more signal. 

Steve

attachicon.gif2021_11_04_AS_zoom_lowres.jpg

Absolutely not.  Higher ISO does not add more signal.  The signal you are getting is linear- meaning your camera receives the same light it receives on any ISO, It cannot force more light photons into your sensor. 
ISO is an amplifier, it amplifies what you already got. And it does not add noise, it amplifies the noise you have along with the signal you have. 

 

It's not like land based photography, where we boost the ISO to have things visible on 1 single frame. 

Don't aim to see something in single shots of the camera.  Most single frames taken with any camera of the night sky looks like nothing. Looks like pitch black with a few stars on it.  The software (NINA, SGP, APT etc) - stretches the image for a preview  

 

Don't mistake that for how the sub looks like.   
When I imaged the Witch Head Nebula, it's 3 stars and total blackness without stretching the image. 

 

When it comes to making pictures of the night sky, it's different is so many ways from landscape photo.  We are not trying to make 1 picture, but it's more like a fraction of the light at each subframe, which later be sandwiched together and then from the sandwitched-stacked frame has to be stretched out. 

In astro-photo, it's about trying to find the perfect balance between the least amount of camera noise and the most amount of signal to capture each shot. 

The sensors are not perfect, so each camera has its own "best" setting for amplifying (ISO).   For Canons, it's ISO 800 or ISO1600.    Others like Nikon and Olympus can go down low as ISO 200. 

Let's say you dial in ISO 800 on your Camera. 
Now , take a 30 second exposure.  Dont' change anything.  Just look at it and determine where the histogram peaks.  Don't look at the image and try to determine if it's good or not, but it's kind of  like navigating a submarine with no windows.

You have to go by the readout of what your camera (or imaging software) tells you what you got.  The image maybe look garbage or black or weird, don't go by that, beside looking at the stretched version of it . 

If it's too dark, meaning the histogram is all the way to the left, it could be still ok, but with DSLRs and mirroless, you can move the histogram a bit towards the right , so it peaks at between 1/4 on the left side. 

Don't change the ISO, but instead, take a 60 second shot and look at the picture histogram again.   Keep doing this until you find the peak at 1/4 to the right or less.  Stick to that exposure settings and take as many subframes as possible. 
 

post-355785-0-17927900-1620009808_thumb.


Edited by unimatrix0, 02 December 2021 - 10:26 AM.

  • bobzeq25 likes this

#5 stevejbo

stevejbo

    Lift Off

  • -----
  • topic starter
  • Posts: 17
  • Joined: 16 Sep 2020

Posted 03 December 2021 - 12:40 AM

These are extremely helpful discussions - I'll drop the iso and give it another try. 


  • bobzeq25 and John Tucker like 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