I'm seeing lots of noise in my M31 images from my Bortle 8-9 backyard in the city, I'm using GraXpert to reduce noise but im still not pleased with the image quality i am getting. Im currently using a stock Canon dslr, star tracker, and an 85mm f1.8 lens. Im new to this hobby so im not exactly sure the cause.
How can I take images with less noise?
The only way to truly reduce noise (increase your signal to noise ratio) is more integration time, or darker skies. (or both).
This statement though assumes you know what stacking is? Are you taking a single image? or stacking multiple sub exposures?
This statement though assumes you know what stacking is? Are you taking a single image? or stacking multiple sub exposures?
What ISO are you using? What is your exposure time? Being a DSLR, more frames is probably your first starting point.
Well Written Concise
@Jake Turner I was using iso 400 so i could shoot longer frames without it being blown out.
@Jake Turner I was shooting around 700 exposures at 5 seconds each
At 700 x 5s you are still at under an hour of total integration time, which is quite short for your light polluted skies. Aim for at least a few hours, the more you have patience for the better it gets.
Since you are using a star tracker you should bump up that exposure time to say, a minute for starters although that might be too long. Your limiting factor here is the core, which can get blown out when exposing for too long. Just test how long you can expose without the core getting blown and choose an exposure time under that (5s is definitely too short). With longer exposures you will have an easier time stacking up to a longer integration time without the process taking all day, at 700 frames you are probably already finding that stacking takes a while.
Since you are using a star tracker you should bump up that exposure time to say, a minute for starters although that might be too long. Your limiting factor here is the core, which can get blown out when exposing for too long. Just test how long you can expose without the core getting blown and choose an exposure time under that (5s is definitely too short). With longer exposures you will have an easier time stacking up to a longer integration time without the process taking all day, at 700 frames you are probably already finding that stacking takes a while.
Well Written Helpful Insightful Concise
@Ethan Sweet I know a place about 40 minutes away from me northeast and its a Bortle 5, would you consider that a good spot for capturing m31?
Sidney Minter:
@Ethan Sweet I know a place about 40 minutes away from me northeast and its a Bortle 5, would you consider that a good spot for capturing m31?
B5 is much better than city skies. Not superb but still much better.
I have two different versions of M31 in my gallery that were taken in my B6 back yard. The one from 2023 is 3.5 hours of data and the one from 2024 is just shy of 10 hours of data. Both were taken with my Canon 70D and canon 200mm lens.
So if you have the ability to take your gear to the darker skies your images will definitely benefit. But you will still need to spend a good amount of time (think hours) on the target.
As @Oskari said bump up your exposure time. You'll have to watch your histogram and make sure it stays 1/4-1/3 from the left. Even 30 sec exposures will reduce your current number of images by 6X.
Canon DSLRs have a lot of read noise below ISO 800-1600, and I'm assuming your using your lens at f/1.8, I would stop that down some to tame the stars, sharpen the image and allow you to run your camera at ISO 800. It will take some experimentation.
Helpful Insightful Respectful Engaging
Ethan Sweet:Sidney Minter:
@Ethan Sweet I know a place about 40 minutes away from me northeast and its a Bortle 5, would you consider that a good spot for capturing m31?
B5 is much better than city skies. Not superb but still much better.
I have two different versions of M31 in my gallery that were taken in my B6 back yard. The one from 2023 is 3.5 hours of data and the one from 2024 is just shy of 10 hours of data. Both were taken with my Canon 70D and canon 200mm lens.
So if you have the ability to take your gear to the darker skies your images will definitely benefit. But you will still need to spend a good amount of time (think hours) on the target.
As @Oskari said bump up your exposure time. You'll have to watch your histogram and make sure it stays 1/4-1/3 from the left. Even 30 sec exposures will reduce your current number of images by 6X.
Canon DSLRs have a lot of read noise below ISO 800-1600, and I'm assuming your using your lens at f/1.8, I would stop that down some to tame the stars, sharpen the image and allow you to run your camera at ISO 800. It will take some experimentation.
Thanks, I will change my camera settings and integration time the next time I shoot on a clear night. And if the quality still isn't ideal I will shoot from a less light polluted area and see how that goes.
Well Written
I agree with the others here. Longer exposures will help even in light polluted skies and more overall integration time will allow the good signal to overcome noise and the bad signal from light pollution.
Well Written Concise
Astrophotography in Bortle 7/8/9 is possible, I shoot under a solid Bortle 8 sky. Since heading to my astronomy club's dark sites isn't really a regular option, I make do with what I have to work with. As others have pointed out, you will need much longer integration times to really start reducing the noise. Start at trying to get a solid 1 hour's worth of data. From there, build off of that to push for more hours. Stopping down the lens is a great idea as well and, if you can, try to take 10s exposures to start off. Experiment incrementally and see if you can move up to 15s-30s–keep experimenting as you gain more experience. Likewise, your exposure length will also be determined with how well your star tracker is aligned and tracking. I also used to shoot between ISO 800-1600 with my old stock Canon T5i with no clip-in filters. If you start narrow band imaging, there are more DSLR clip-in filters now than there were in 2015 (when I started in earnest).
With a stock DSLR, you can do quite a bit, but bear in mind that you will/should take darks and bias frames during each session so factor that into your imaging process. Good luck!
With a stock DSLR, you can do quite a bit, but bear in mind that you will/should take darks and bias frames during each session so factor that into your imaging process. Good luck!
Helpful Respectful Engaging Supportive
Make sure you are shooting RAW.
Keep the ISO low, you are shooting at ISO400 and that’s a good place to be with most DSLRs.
Properly calibrate your data with flats and bias frames. You can take darks but keep in mind that since you have no control over the camera temp, the darks won’t be very accurate. You’ll just have to try it and see if they hurt or help. If you do take darks, do it in the middle of the session.
Many will tell you to make longer sub exposures but I don’t agree. You are under very heavy light pollution so shot noise is going to be the major source of noise for you. I would stick with the 5 sec. you’ve been doing.
As everyone else has noted, you need a LOT of total exposure time under B-8,9. I would try and get a few hours at least. Stopping down a bit might be a good idea as well.
Helpful Concise Supportive
First of all, as all the others have already pointed out, you need more integration time, you need more signal, more "usable data", at least few hours (the more the better). This will help increasing the SNR (signal-to-noise ratio) during stacking. GraXpert is a very good tool but it needs signal/data to work with.
Increasing the exposure time will help getting more signal, but it will also amplify the noise. And as you are in highly light polluted area... (however 5 sec. exposure seems too short to me; I would try to increase it and see what I get).
As @Tony Gondola already mentioned, do not forget to take calibration frames. The Dark frames are the ones intended to reduce the noise, but they are temperature dependent: you need to take them at the same temperature as your light frames. And as you do not have temperature control on your DSLR, again as @Tony Gondola mentioned, the only thing you can do is to try to be as close as possible to the light frames temperature by taking the darks during the session. Me, I used to take at least 20 of them at the end of the session, by the time I was imaging with the DSLR. Unfortunatelly this means time lost for the light frames, but in your case with 5 sec. exposure it might not be such a problem (me I was using exposures of 60 to 120 sec.).
Other ideas:
With some effort, you can get very good results with DSLRs, but do not expect the same as with dedicated astronomy cameras which feature low noise and temperature control.
I used to image with my astro-modified Canon 100D and I was quite pleased with the results, until the moment it broke and I decided to purchase a dedicated camera for astrophotography. For me the difference was, let's say...a big one. And so was the first time I started using filters.
I hope it helps.
Clear skies!
Increasing the exposure time will help getting more signal, but it will also amplify the noise. And as you are in highly light polluted area... (however 5 sec. exposure seems too short to me; I would try to increase it and see what I get).
As @Tony Gondola already mentioned, do not forget to take calibration frames. The Dark frames are the ones intended to reduce the noise, but they are temperature dependent: you need to take them at the same temperature as your light frames. And as you do not have temperature control on your DSLR, again as @Tony Gondola mentioned, the only thing you can do is to try to be as close as possible to the light frames temperature by taking the darks during the session. Me, I used to take at least 20 of them at the end of the session, by the time I was imaging with the DSLR. Unfortunatelly this means time lost for the light frames, but in your case with 5 sec. exposure it might not be such a problem (me I was using exposures of 60 to 120 sec.).
Other ideas:
- Use filters: You can partially address the issue of the light pollution by using broadband/light pollution filters (for deep sky objects like galaxies) or narrowband filters for nebulae. This will allow you increase the exposure time.
- Dithering: at 5 sec. exposure time I suppose you go unguided, but in case you consider going for autoguiding (I don't know if your mount supports it), dithering can help reducing the noise in the final image.
With some effort, you can get very good results with DSLRs, but do not expect the same as with dedicated astronomy cameras which feature low noise and temperature control.
I used to image with my astro-modified Canon 100D and I was quite pleased with the results, until the moment it broke and I decided to purchase a dedicated camera for astrophotography. For me the difference was, let's say...a big one. And so was the first time I started using filters.
I hope it helps.
Clear skies!
Helpful
First of all, as all the others have already pointed out, you need more integration time, you need more signal, more "usable data", at least few hours (the more the better). This will help increasing the SNR (signal-to-noise ratio) during stacking. GraXpert is a very good tool but it needs signal/data to work with.
Increasing the exposure time will help getting more signal, but it will also amplify the noise. And as you are in highly light polluted area... (however 5 sec. exposure seems too short to me; I would try to increase it and see what I get).
As @Tony Gondola already mentioned, do not forget to take calibration frames. The Dark frames are the ones intended to reduce the noise, but they are temperature dependent: you need to take them at the same temperature as your light frames. And as you do not have temperature control on your DSLR, again as @Tony Gondola mentioned, the only thing you can do is to try to be as close as possible to the light frames temperature by taking the darks during the session. Me, I used to take at least 20 of them at the end of the session, by the time I was imaging with the DSLR. Unfortunatelly this means time lost for the light frames, but in your case with 5 sec. exposure it might not be such a problem (me I was using exposures of 60 to 120 sec.).
Other ideas:
- Use filters: You can partially address the issue of the light pollution by using broadband/light pollution filters (for deep sky objects like galaxies) or narrowband filters for nebulae. This will allow you increase the exposure time.
- Dithering: at 5 sec. exposure time I suppose you go unguided, but in case you consider going for autoguiding (I don't know if your mount supports it), dithering can help reducing the noise in the final image.
With some effort, you can get very good results with DSLRs, but do not expect the same as with dedicated astronomy cameras which feature low noise and temperature control.
I used to image with my astro-modified Canon 100D and I was quite pleased with the results, until the moment it broke and I decided to purchase a dedicated camera for astrophotography. For me the difference was, let's say...a big one. And so was the first time I started using filters.
I hope it helps.
Clear skies!
Excellent points regarding autoguiding and dithering, something I completely forgot about in my post.
OP, if you do wind-up being able to take longer exposures (with or without a filter) and your star tracker is capable of guiding, you’ll definitely want to dither.
Well Written Supportive
If for some reason you cannot take longer exposures, you can "clean" your images very efficiently with a denoising script. A very powerful one isTopazDenoiseAI.
It can suppress noise and strongly enhance sharpness. Very impressive ! See examples their website.
Philippe
It can suppress noise and strongly enhance sharpness. Very impressive ! See examples their website.
Philippe
Concise
From my experience, there was a huge difference between 50sec exposure and 300 sec exposure for 4 hours in terms of noise and quality. M31 has a lot of dust and nebulosity around its arms. Also, for noise removal, NoiseXterminator from Russell Croman helps a lot in processing stage.
Helpful
ed bat · Aug 14, 2025, 12:37 PM
From my experience, there was a huge difference between 50sec exposure and 300 sec exposure for 4 hours in terms of noise and quality. M31 has a lot of dust and nebulosity around its arms. Also, for noise removal, NoiseXterminator from Russell Croman helps a lot in processing stage.
in B8-9?
B4-5
Those are very different conditions. Long subs do not work well under 8-9. The OP can try it but I doubt he’ll be happy with the results, especially with such a fast lens.
Oskari Nikkinen:
At 700 x 5s you are still at under an hour of total integration time, which is quite short for your light polluted skies. Aim for at least a few hours, the more you have patience for the better it gets.
Since you are using a star tracker you should bump up that exposure time to say, a minute for starters although that might be too long. Your limiting factor here is the core, which can get blown out when exposing for too long. Just test how long you can expose without the core getting blown and choose an exposure time under that (5s is definitely too short). With longer exposures you will have an easier time stacking up to a longer integration time without the process taking all day, at 700 frames you are probably already finding that stacking takes a while.
You can also use HDR composition, that is long subs for the fainter portions and short subs for the brighter areas. If you use Pixinsight it easy, there is an HDR integration tool that combines the images. I use it to retain color in the core of stars, which tend to blow out quickly. So I take 100 or so 10 sec subs in each color (I shoot with a mono camera), and Pixinsight combines the subs and uses the 10 sec subs for the stars. Actually, the tool creates masks that are applied to the image during integration which applies the short sub data to the brightest areas. I used this method for M31 due to the core and it worked well. I used it for M42 when I shot it years ago. Even without Pixinsight HDR composition is a thing, though I wouldn't know how to do it without the PI tool. Maybe Pixel Math?
Helpful
A Comparison - Fast vs. Large
Of the many diatribes in AP world one of the most recurrent is the role that "fast" optics plays in achieving a given result or more generally speaking whether being "fast" is better than being "large". And here is the c...
Review: ToupTek 36mm LRGB and SHO Filters.
Thought I'd share my thoughts about the ToupTek Astro filters I've been using. Brilliant budget friendly filters. ToupTek Astro 36mm LRGB + SHO Filter Review. Introduction.We all know how expensive astronomy can be. Further to that, we all kn...