Back to forums

Can anyone try processing my M101 to see if I just have bad data (poor seeing/transparency etc.)

Processing Deep Sky M27 M101 Messier Objects
17 replies•1.3k views
Fabian Butkovich avatar
Fabian Butkovich avatar
I recently collected up to 15.2hrs of 2-minute long subs on M101 across 4 non-consecutive nights within the last month. So far the results are ok but I really wish I could pull more out of the outer arms of the pinwheel galaxy which is what really gives it it's distinctive spiral structure. 

I noticed that the subs on another DSO I've been imaging the same nights as M101, M27 the Dumbell Nebula, have a lot clearer looking appearance and better star color and improved contrast, this is at the same ISO, same exposure length focal length etc. 

So my fear is that currently at this time of year in my location M101 is not really the best choice to photograph due to it's position in the sky there might be more light pollution towards the northern part of the sky in the Bortle 7-8 zone that I'm in. 

Here is M27 captured in the early morning hours


And for comparison this is M101 captured between 10PMish-2AMish every night


This is my final processed image with 12h30m of total integration time (375x2min stack)


Here is the link to the .tif straight out of DSS:
https://drive.google.com/file/d/1WhyUZ76kYf_MbI2MwzK9rbVDepGTiJOl/view?usp=sharing

I'm excited to see if anyone else can process this better than me
Helpful Engaging
andrea tasselli avatar
andrea tasselli avatar
So, it is out of focus and not calibrated properly or at all. I congratulate you for the image then as it wasn't easy at all to come up with something worth the effort.
Fabian Butkovich avatar
Fabian Butkovich avatar
andrea tasselli:
So, it is out of focus and not calibrated properly or at all. I congratulate you for the image then as it wasn't easy at all to come up with something worth the effort.

Focusing on my Sigma 150-600 f/6.3 lens was as sharp as it could be since I used zoomed-in live view focusing aid on the camera. I will have to eventually use a bahtinov mask to see if I could get anything better, but I believe poor focus is not the sole issue on why I can't seem to extract more data from my stack of M101.
Andy Wray avatar
Andy Wray avatar
Quick question if I may?  Are you taking any flats and darks or are you just stacking photos?

FWIW:  I was only just learning about flats and darks and this is the best I could get with my mirrorless camera on that target:


M101 on NEX5N

Also:  I did try re-processing your image, but struggled to improve on what you did.
Helpful Respectful Engaging Supportive
andrea tasselli avatar
andrea tasselli avatar
Fabian:
Focusing on my Sigma 150-600 f/6.3 lens was as sharp as it could be since I used zoomed-in live view focusing aid on the camera. I will have to eventually use a bahtinov mask to see if I could get anything better, but I believe poor focus is not the sole issue on why I can't seem to extract more data from my stack of M101.


Live focusing isn't the way to go. Use either the bahtinov mask on a bright stars or connect the camera to a PC and look at the in focus image of the stars and fine tune from there.  BTW, what camera were you using then?

The other reason of the poor data extraction is that you didn't calibrate the images properly or at all. Using DSS doesn't help either. You should invest some time in learning better programs for handling this line of work. I suggest Siril is your next port of call. Not terribly easy but very good and very free.
Helpful
Fabian Butkovich avatar
Fabian Butkovich avatar
Andy Wray:
Quick question if I may?  Are you taking any flats and darks or are you just stacking photos?

FWIW:  I was only just learning about flats and darks and this is the best I could get with my mirrorless camera on that target:


M101 on NEX5N

Also:  I did try re-processing your image, but struggled to improve on what you did.

I've tried taking darks and flats before with ambiguous results, however that was with different optics than I have now. Perhaps it wouldn't be a bad idea to get into that habit of taking calibration frames now that I have an improved optics setup than I did previously.
Andy Wray avatar
Andy Wray avatar
Fabian:
Andy Wray:
Quick question if I may?  Are you taking any flats and darks or are you just stacking photos?

FWIW:  I was only just learning about flats and darks and this is the best I could get with my mirrorless camera on that target:


M101 on NEX5N

Also:  I did try re-processing your image, but struggled to improve on what you did.

I've tried taking darks and flats before with ambiguous results, however that was with different optics than I have now. Perhaps it wouldn't be a bad idea to get into that habit of taking calibration frames now that I have an improved optics setup than I did previously.

I can promise you that it makes a huge difference.  You only need to take darks every few months or so, however flats are quite key and should be taken quite frequently; definitely if you change any rotation in your optical train.  I don't do either anywhere near enough.
Rick Veregin avatar
Rick Veregin avatar
My advice is mostly similar to that given so far: you must learn to do darks and good flats. Darks are easy, flats need some work–there are a lot of discussion here on Astrobin about how to do flats, or search online. As far as I can see any of the methods will work well for flats, once you understand the requirements, just work on it until you get them right.

As for registering and stacking there are many options. I doubt personally that DSS is your issue, I have never had an issue in my own work due to DSS (aside from my mistakes is settings). Make sure you learn the ins and outs of DSS, before you try something else. I find DSS easy to use compared to some programs, and if you switch you will be back to square one. And again, from what I can see register and stack is not the issue, you are on a reasonable track with what you are doing. One way of checking is to look at your average FWHM in DSS for each of your images–I copy the registered DSS file list to Excel and then do an average there. Then when you have your  saved stacked file, load that as a picture and register it. Is the FWHM of the stack the same or even better than the single image–it should be. Look at stars at the corners, where there may be abberations from your optics and from stack errors–are shapes of the stars similar in the final stack as in single images? Again, generally star shapes are better in the stack than a single image. 

Especially in high Bortle regions (I am in Bortle 8 myself), I try to image when targets are high overhead. The sky is much brighter and the stars more bloated the lower you go in elevation. I rarely image below 45 degrees elevation–though I break down as there are some nice targets that are lower than this. But work as high as you can and try to avoid nights with low transparency, the sky is so much brighter then.

Sadly, if you are in Bortle 7 or 8 lower your expectations a bit, you will need longer exposures than those with darker skies to get the same result, and the faintest stuff may be very difficult. If you compare to someone with darker skies you will always fall short. 

Narrowband is one way to close the gap due to sky glow, though you will always need more exposure than if you had darker skies. 

Make sure your camera is in the best iso/gain range for lowest noise, critical in bright skies as you exposures are short and sky noise is high.

Also with time with Bortle 7/8 data you will get better at stretching and getting the ultimate from the image, but you will need to work harder than if you skies were dark. 

Rick
Helpful
Fabian Butkovich avatar
Fabian Butkovich avatar
Rick Veregin:
My advice is mostly similar to that given so far: you must learn to do darks and good flats. Darks are easy, flats need some work--there are a lot of discussion here on Astrobin about how to do flats, or search online. As far as I can see any of the methods will work well for flats, once you understand the requirements, just work on it until you get them right.

As for registering and stacking there are many options. I doubt personally that DSS is your issue, I have never had an issue in my own work due to DSS (aside from my mistakes is settings). Make sure you learn the ins and outs of DSS, before you try something else. I find DSS easy to use compared to some programs, and if you switch you will be back to square one. And again, from what I can see register and stack is not the issue, you are on a reasonable track with what you are doing. One way of checking is to look at your average FWHM in DSS for each of your images--I copy the registered DSS file list to Excel and then do an average there. Then when you have your  saved stacked file, load that as a picture and register it. Is the FWHM of the stack the same or even better than the single image--it should be. Look at stars at the corners, where there may be abberations from your optics and from stack errors--are shapes of the stars similar in the final stack as in single images? Again, generally star shapes are better in the stack than a single image. 

Especially in high Bortle regions (I am in Bortle 8 myself), I try to image when targets are high overhead. The sky is much brighter and the stars more bloated the lower you go in elevation. I rarely image below 45 degrees elevation--though I break down as there are some nice targets that are lower than this. But work as high as you can and try to avoid nights with low transparency, the sky is so much brighter then.

Sadly, if you are in Bortle 7 or 8 lower your expectations a bit, you will need longer exposures than those with darker skies to get the same result, and the faintest stuff may be very difficult. If you compare to someone with darker skies you will always fall short. 

Narrowband is one way to close the gap due to sky glow, though you will always need more exposure than if you had darker skies. 

Make sure your camera is in the best iso/gain range for lowest noise, critical in bright skies as you exposures are short and sky noise is high.

Also with time with Bortle 7/8 data you will get better at stretching and getting the ultimate from the image, but you will need to work harder than if you skies were dark. 

Rick

Rick, I appreciate your extensive advice. For me imaging M101 high overhead hasn't been an issue this time of year as it is at least 70 degrees overhead by the time it gets dark. 

I also do not think integration time is an issue either because I have more than 10hrs to play with so far, so I should have a better defined spiral arm structure. 

I will take some darks and flats along with 2-3 hours more of lights between tonight and tomorrow night when it is clear and stack that to see if there is any improvement in the amount of signal I can pull out. Unfortunately, the temperatures have gone up significantly since the first time I imaged this target for the season, so I'm not sure how well my dark frames will apply to previous nights data.
Helpful Insightful Respectful
Rick Veregin avatar
Rick Veregin avatar
Fabian:
Rick Veregin:
My advice is mostly similar to that given so far: you must learn to do darks and good flats. Darks are easy, flats need some work--there are a lot of discussion here on Astrobin about how to do flats, or search online. As far as I can see any of the methods will work well for flats, once you understand the requirements, just work on it until you get them right.

As for registering and stacking there are many options. I doubt personally that DSS is your issue, I have never had an issue in my own work due to DSS (aside from my mistakes is settings). Make sure you learn the ins and outs of DSS, before you try something else. I find DSS easy to use compared to some programs, and if you switch you will be back to square one. And again, from what I can see register and stack is not the issue, you are on a reasonable track with what you are doing. One way of checking is to look at your average FWHM in DSS for each of your images--I copy the registered DSS file list to Excel and then do an average there. Then when you have your  saved stacked file, load that as a picture and register it. Is the FWHM of the stack the same or even better than the single image--it should be. Look at stars at the corners, where there may be abberations from your optics and from stack errors--are shapes of the stars similar in the final stack as in single images? Again, generally star shapes are better in the stack than a single image. 

Especially in high Bortle regions (I am in Bortle 8 myself), I try to image when targets are high overhead. The sky is much brighter and the stars more bloated the lower you go in elevation. I rarely image below 45 degrees elevation--though I break down as there are some nice targets that are lower than this. But work as high as you can and try to avoid nights with low transparency, the sky is so much brighter then.

Sadly, if you are in Bortle 7 or 8 lower your expectations a bit, you will need longer exposures than those with darker skies to get the same result, and the faintest stuff may be very difficult. If you compare to someone with darker skies you will always fall short. 

Narrowband is one way to close the gap due to sky glow, though you will always need more exposure than if you had darker skies. 

Make sure your camera is in the best iso/gain range for lowest noise, critical in bright skies as you exposures are short and sky noise is high.

Also with time with Bortle 7/8 data you will get better at stretching and getting the ultimate from the image, but you will need to work harder than if you skies were dark. 

Rick

Rick, I appreciate your extensive advice. For me imaging M101 high overhead hasn't been an issue this time of year as it is at least 70 degrees overhead by the time it gets dark. 

I also do not think integration time is an issue either because I have more than 10hrs to play with so far, so I should have a better defined spiral arm structure. 

I will take some darks and flats along with 2-3 hours more of lights between tonight and tomorrow night when it is clear and stack that to see if there is any improvement in the amount of signal I can pull out. Unfortunately, the temperatures have gone up significantly since the first time I imaged this target for the season, so I'm not sure how well my dark frames will apply to previous nights data.

Thats great, you are doing mostly the right things so far with your imaging.

If temperature is changed, you must redo your darks. It is easy to do darks, just cover your lens at the end of the night at the same temperature and do your new darks. It is important to have the darks temperature and iso/gain matched, otherwise you will not correct out the dark side of your image, and even your flats will not be quite right.

I would respectfully disagree that you have enough data (you asked for advice, don't discount my experience in bright skies without proof) . In high Bortle regions M101 is an exceptionally unreasonably difficult target! It is huge in arc-seconds, but for that reason it is not bright per arc second (each pixel gets little light). I use Telescopius which shows the surface brightness (brightness/arc-sec) of 14.9 magnitude/per arc square for M101!!! I did not see mention of what F-ratio you are using, a very low F-ratio would enable this target.

Under 13 in your Bortle is relatively easy (10 hrs is enough probably for a great image under good conditions--I use F6.3), 13 to 14 needs 20 hours at least, while M101 in my Bortle 8, I would not expect a really good image without very, very long exposures for this diffuse a target. I have not done anything in my bright skies over 14 yet--doable, but lots of hours to make a really decent image.

Is there a reason you want M101, there are many, many wonderful galaxies that are not nearly so difficult in the same area??

I suggest you choose a more amenable target to your light conditions (you can use Telescopius to find them), and my bet is you will be really happy with the result. 
Rick
Helpful Insightful Engaging
Fabian Butkovich avatar
Fabian Butkovich avatar
@Rick Veregin I added 25 flats, 15 darks, and 2 hours more worth of data to my original dataset and this was the final result, a huge difference in the amount of detail in the spiral arms of the galaxy if you ask me. 

Padraig Farrell avatar
Padraig Farrell avatar
Hi Rick 

just wondering how do you calculate …bortle vs surface brightness = integration time. 
A lot of images seem to be around 10 ‘ish hours. 
are you looking for something in particular in you image to go longer ? 

sorry for the hijack Fabian 🫣
Observatório Astrográfico do Boqueirão avatar
Observatório Astrográfico do Boqueirão avatar
I file went to trash. I can't give a try
Rick Veregin avatar
Rick Veregin avatar
Fabian Butkovich:
@Rick Veregin I added 25 flats, 15 darks, and 2 hours more worth of data to my original dataset and this was the final result, a huge difference in the amount of detail in the spiral arms of the galaxy if you ask me. 


Yes, indeed--great illustration of the importance of calibration--very nice!
Well Written Respectful Supportive
Rick Veregin avatar
Rick Veregin avatar
Padraig Farrell:
Hi Rick 

just wondering how do you calculate …bortle vs surface brightness = integration time. 
A lot of images seem to be around 10 ‘ish hours. 
are you looking for something in particular in you image to go longer ? 

sorry for the hijack Fabian 🫣

Actually, that is just my experience that I mention that. For each target I do I note the surface brightness and then what my image looks like after different lengths of time. And it is arbitrary, early on in my experience I would have said it can be done in shorter time, but well I like to push the limits of what I can capture, so my standards for the image are higher now.

What am I looking for:
-Noise overall: the areas with detail I like to be nearly noise free, so no need to do any noise control there that will kill detail. And on the other hand to pull out more detail using decon or sharpening or wavelets requires very low noise, otherwise you just magnify the noise.

-Faint areas in the target are visible without excessive stretching. You can push this as far as you want, takes longer to get tidal tails in most galaxies for example.

-If there are faint areas with detail, then look for low noise in the faint areas too. One can tolerate noise in the faint areas if there is not much detail there, then just use more noise suppression in those areas. Depends on the target.

-Color is much better when you have enough integration, and easier to get in processing. And stretching is much easier, if you need to stretch too much, image won't look as good.

Fact of the matter is longer is always better, in the end it is what you want to get out of the target and how much time you want to put in. I spend a lot of time getting the processing right. If you spend a lot of time in processing and aren't getting anywhere, you need more exposure.

My own philosophy is to try to make my image the best there is, so I compare to the best images to see how I am progressing. Some targets don't continue to improve much, so you can stop earlier. Others you see improvement continuously, gets hard to stop.  I generally only do two targets at a time--one at least 3 hours east of the other--so early in the night I do one while it is high in the sky, then later switch to the other. For many targets this means I do about 2 targets in 2 to 3 months, maybe as many as 8 to 15 nights of as much time as I can manage. I stop when I run out of sky and need to switch to new targets. Others like to do a reasonable job on many targets if they can, depends what you are looking for in this.

The other thing I do is to check what others have done with the target in a similar Bortle, and use that as a reference for what you will need to do to be as good, or better!
CS
Rick
Helpful
Padraig Farrell avatar
Padraig Farrell avatar
Thanks for explaining that Rick,
I must keep a notebook of when imaging.
Fabian Butkovich avatar
Fabian Butkovich avatar
Rick Veregin:
Padraig Farrell:
Hi Rick 

just wondering how do you calculate …bortle vs surface brightness = integration time. 
A lot of images seem to be around 10 ‘ish hours. 
are you looking for something in particular in you image to go longer ? 

sorry for the hijack Fabian 🫣

Actually, that is just my experience that I mention that. For each target I do I note the surface brightness and then what my image looks like after different lengths of time. And it is arbitrary, early on in my experience I would have said it can be done in shorter time, but well I like to push the limits of what I can capture, so my standards for the image are higher now.

What am I looking for:
-Noise overall: the areas with detail I like to be nearly noise free, so no need to do any noise control there that will kill detail. And on the other hand to pull out more detail using decon or sharpening or wavelets requires very low noise, otherwise you just magnify the noise.

-Faint areas in the target are visible without excessive stretching. You can push this as far as you want, takes longer to get tidal tails in most galaxies for example.

-If there are faint areas with detail, then look for low noise in the faint areas too. One can tolerate noise in the faint areas if there is not much detail there, then just use more noise suppression in those areas. Depends on the target.

-Color is much better when you have enough integration, and easier to get in processing. And stretching is much easier, if you need to stretch too much, image won't look as good.

Fact of the matter is longer is always better, in the end it is what you want to get out of the target and how much time you want to put in. I spend a lot of time getting the processing right. If you spend a lot of time in processing and aren't getting anywhere, you need more exposure.

My own philosophy is to try to make my image the best there is, so I compare to the best images to see how I am progressing. Some targets don't continue to improve much, so you can stop earlier. Others you see improvement continuously, gets hard to stop.  I generally only do two targets at a time--one at least 3 hours east of the other--so early in the night I do one while it is high in the sky, then later switch to the other. For many targets this means I do about 2 targets in 2 to 3 months, maybe as many as 8 to 15 nights of as much time as I can manage. I stop when I run out of sky and need to switch to new targets. Others like to do a reasonable job on many targets if they can, depends what you are looking for in this.

The other thing I do is to check what others have done with the target in a similar Bortle, and use that as a reference for what you will need to do to be as good, or better!
CS
Rick

Thank you for sharing your extensive field experience Rick. I think many of us feel like our images will never be good enough lol, were always pushing our skills and equipment limits. I've been recently acquiring two targets a night, M101 eventually disappears behind the west side of my house and then I'll switch to M27 which is rising in the east at this time of year. 

Fun fact, I recently tested my wife's knowledge just to see if she even pays attention to me when I babble about AP, I asked her why couldn't I image the Orion Nebula now if I wanted to, and she passed the test lol (I'm in the Northern Hemisphere). 

Appreciate the input
CS
Fabian
Respectful Engaging Supportive
Fabian Butkovich avatar
Fabian Butkovich avatar
Padraig Farrell:
Hi Rick 

just wondering how do you calculate …bortle vs surface brightness = integration time. 
A lot of images seem to be around 10 ‘ish hours. 
are you looking for something in particular in you image to go longer ? 

sorry for the hijack Fabian 🫣

No worries Padraig, I'm glad my forum post has been very popular topic and helpful