Jump to content

Narrowband

M31 with 60mm


Scooot

Recommended Posts

This is my second deep sky image with the baby tak on the sky adventurer and my first of M31. Considering what the lights looked like I was surprised at the end result. I stacked 34 thirty second images taken on ISO 800 and calibrated, integrated, and processed them in pixinsight. I think more lights would have been good but for a 60mm scope it's picked up a lot of data. All suggestions welcome :) 

M31_f6_2_iso_800_30secs_DBE.jpg

 

Link to comment
Share on other sites

14 minutes ago, wimvb said:

For only 30 second exposures that's very good.

More data (ideally longer exposures) will lower the noise level.

Thanks wim, this was one of the lights straight out the of the camera. You can see why I was surprised with the outcome :) 

I'm not sure how much longer I could go with the unguided star adventurer, maybe 60 seconds if I'm more careful with the polar alignment. I was wondering whether to mix up the exposures and ISO setting, maybe some at ISO 400 to reduce noise and increase the dynamic range and combine with HDR.

m31 IMG_9730.JPG

Link to comment
Share on other sites

That's quite faint. If you can't increase the exposure time, try increasing ISO and taking more exposures. Don't bother about "diminishing returns", take as many images as you can. 80 30 second images is still only 40 minutes of data. Even double that, 160, shouldn't be a problem. The idea is that if you can't increase the signal, you should aim for decreasing the noise. Lower noise will allow you to stretch the  image more.

As for the image at hand:

You can get more detail, without overstretching the core, by using MaskedStretch as the first stretch. Then you use histogram transform to stretch more. Masked stretch will stretch faint parts more than stars and the core, but it will result in a flat image. HistogramTransform will increase the range.

After that, try HDRMultiscaleTransform to bring out detail in/near the core. This technique works best if your data has a low noise level.

I find that as long as the stacked image isn't over exposed (and it rarely is), doing HDR combination isn't necessary in PI. There are various methods that will allow you to create an HDR image with just one source image. If you want to do HDR combination, you can also just use the original image. Stretch one copy to bring out detail in the outer regions, and one copy to bring out detail in the core. Then combine these images using a lightness mask and PixelMath.

Good luck

Link to comment
Share on other sites

 

7 hours ago, wimvb said:

That's quite faint. If you can't increase the exposure time, try increasing ISO and taking more exposures. Don't bother about "diminishing returns", take as many images as you can. 80 30 second images is still only 40 minutes of data. Even double that, 160, shouldn't be a problem. The idea is that if you can't increase the signal, you should aim for decreasing the noise. Lower noise will allow you to stretch the  image more.

As for the image at hand:

You can get more detail, without overstretching the core, by using MaskedStretch as the first stretch. Then you use histogram transform to stretch more. Masked stretch will stretch faint parts more than stars and the core, but it will result in a flat image. HistogramTransform will increase the range.

After that, try HDRMultiscaleTransform to bring out detail in/near the core. This technique works best if your data has a low noise level.

I find that as long as the stacked image isn't over exposed (and it rarely is), doing HDR combination isn't necessary in PI. There are various methods that will allow you to create an HDR image with just one source image. If you want to do HDR combination, you can also just use the original image. Stretch one copy to bring out detail in the outer regions, and one copy to bring out detail in the core. Then combine these images using a lightness mask and PixelMath.

Good luck

Thanks for the guidance. I actually used the adaptive stretch which I'm now getting fairly used to. I might be mistaken but when I use this it doesn't seem to give much room to further stretch it without blowing out the image. I'll have a go at the masked stretch and compare. I'll also try a combination of two stretches of the same stack, I've done a similar thing on some landscape images to bring out the sky but hadn't considered it with Astro.

I used the subframe selector script to analyse the lights and tried to stack them according to the assessed weights. However I kept getting an error message to the effect that imageintegration  couldnt find the fits keyword that was embedded in the subs. I couldn't work this out because when I opened the properties of the individual subs the keyword was clearly there. I wonder whether it was because I was working with xisf files and not fits. I'll have to ask on their forum. I ended up stacking according to the weight of the noise, I don't suppose it made much difference but I'd like to get to the bottom of it.

Anyway, it's all an interesting learning curve :) 

edit, I think I might have found the problem, just read a post on the pi forum about negative weights, i need to check but I think some of my weights were negative.

Link to comment
Share on other sites

I've reprocessed the stacked image using a masked stretch, followed by a light stretch with histogram transfer then HDRMultiscale transform a bit more aggressively on the core. then combined it with the original image using pixelmath. This was a bit of trial & error as I didn't really know what I was doing but it seems to have worked. :)  The core is much better and the overall image is a bit softer.

 

 

HDR.jpg

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.