Capturing fainter data

[cuivenion]

Here's what I have so far. This is a fairly quick process, 2 hours of 30 second subs:

I'm fairly happy with the central part of the galaxy but I definately want to improve the outer arms. What's the best way to go about this when capturing? Do I use lower gain and go for longer exposures, or do I just keep collecting data like I'm already doing?

I'm capturing with a ZWO ASI224, this image was captured with 30s exposures at fairly high gain (255). Imaging scope is a Skywatcher 130PDS.

[almcl]

Perhaps longer exposures?

This might/will probably burn out the core so when you have the outer arm image  to your satisfaction, try layering it together with your core detail image?

Hours of fun to be had with your image processing software of choice!

[cuivenion]

Cheers almcl.

[ollypenrice]

In CCD you'd certainly take longer subs. With CMOS maybe not. I don't know how upping the gain plays out but I thought the best of the CMOS imagers were working at unity gain? I'm not sure about this though. The outer parts of M106 are very faint and have always needed a lot of time in my experience. How about trying 60 second subs at unity?

The background sky is looking fairly mottled. Are you dithering? This might help, maybe not between every sub but after shorter batches.

Olly

[drjolo]

How dark sky do you have? Faint galaxy arms can be a challenge under light polluted skies.

[cuivenion]

I haven't dithered so far as I don't have a large field of view, but I see your point on the background. I've hopefully got a clear sky tonight so I'll try the longer exposures with dithering.

[cuivenion]

4 minutes ago, drjolo said:

How dark sky do you have? Faint galaxy arms can be a challenge under light polluted skies.

Not that dark unfortunately.

[cuivenion]

Clear outside gives it as class 5 bortle.

[wimvb]

CMOS imaging is a bit more complex than CCD imaging, because you have a variable gain to deal with. While high gain will generally give you lower read noise, it will also result in a lower dynamic range (down to 8 stops or lower). You will need a lot of exposures to compensate for this. This is why many astrophotographers will use longer exposures at lower gain. Very often well below unity gain.

Also, in order to compare exposure times at a certain gain, you should compare gain in e/ADU (where unity gain is 1 e/ADU), and not in "ZWO units" or "QHY units". Imo, at any gain you should use the longest exposures that don't saturate any part of the galaxy, but will get you above the read noise floor. Your tracking/guiding must allow for these exposures, of course.

An advantage of CMOS imaging is, that because of the variable gain, you can use short exposures if e.g. guiding is working against you.

Finally you should consider what I like to call the Lodriguss-factor: for every magnitude you lose due to light pollution, you have to increase your total integration time by a factor of 2.5 to end up with the same signal to noise ratio. 1 hour of data under magnitude 22 sky is equivalent to 2.5 hours of data under mag 21 sky, and 6.25 hours under mag 20 sky, etc. (While the exact numbers may be debatable, the essence of the Lodriguss-factor is valid.)

In the end it's sky darkness and total integration time that will give you the best result as far as signal to noise is concerned, while tracking and focus will determine star shape and sharpness.

If you opt for short exposures at high gain, you will end up with many subs to calibrate and stack. This will eat storage space and processing time. Your raw image files should be about 2.4 MB in size. 250 30 s subs (about 2 hours worth of data) will occupy 600 MB of hard drive space. 8 hours of data will occupy almost 2.4 GB. If you use 300 s subs at low gain, 8 hours of data will occupy less than 240 MB.

[cuivenion]

Thanks Wim. I had the best seeing for a very long time last Sunday and got about 5 hours of good data on M106, making sure I dithered every 5 minutes. The data's that good (for me) I'm thinking of completely discarding the previous data I had because it just seems to mottle and brighten the background.

I'm going to make sure I dither on each project from now on. I wish I could get away from my urban backgarden but unfortunately it's not an option at the moment, I can see how dark skies would make a huge difference.

I use Sharpcap for image capture and it has an interesting feature called the Brain. It analyses the sky brightness and gives you an optimal exposure time and gain. You have to have previously analysed the camera with Sensor Analysis tool. I got a low gain (18) with a 50 second exposure last time out that worked very well. The relatively low exposure time points to the low read noise of the camera and, I would imagine the light pollution in my area. Would a light pollution filter help with this? I was thinking of going for this:

https://www.firstlightoptics.com/baader-filters/baader-uhc-s-filter.html

[wimvb]

6 hours ago, cuivenion said:

I got a low gain (18) with a 50 second exposure last time out that worked very well. The relatively low exposure time points to the low read noise of the camera and, I would imagine the light pollution in my area. Would a light pollution filter help with this? I was thinking of going for this:

https://www.firstlightoptics.com/baader-filters/baader-uhc-s-filter.html

50 s at low gain seems very short. At that gain setting, I usually can expose for several minutes. But my camera has a higher read noise than yours. If it works, go for it. Just remember that it's the total integration time which is important.

Regarding the lp filter, I used that type before I moved to a dark site. It worked ok for me. From discussions here on sgl, I think that IDAS lp filters may work even better.

You can probably keep the colour data from your old set, since colour is much less critical than luminance when it comes to detail. Unless the framing is very different, there's no harm in using it.

[cuivenion]

Cheers Wim.