9 replies to this topic

[CrookedEyes]

Good day, everybody!

 

I was able to finally acquire some images on 25 November 2020. My intent for this night was only to acquire some images, and if I had the time, I would do calibration frames. Well, again, as a new guy I ran into some speed bumps and hurdles having a hard time getting the NINA auto focus routine to run properly this time. So, this image has no calibration frames, and not a whole lot of data. On top of that, clearly the processing also needs some major work. But, that is expected until I develop myself a normal workflow and get used to using Photoshop again. I was trying for SHO but the last filter set of images in the sequence, my SII, failed to image anything and I just had streaks of stars. I am unsure if I got to the meridian and the mount said, "Nope" or if I lost guiding. I am sure there is a way to figure that out, that I unfortunately do not know. The image I ended up with is HOO that, unfortunately, doesn't have any of the blue tinting that one would expect. Maybe I did something incorrectly? Anyway, this is Ha - 15 images @ 240s and 175 gain, and same for OIII. Tonight is supposed to be a decent night and I would love to try another target, but I think it is best to just collect more data for the Rosette as well as get some darks taken. Temps are supposed to be equivalent to the night of the 25th, so it should benefit. 

 

I was looking through Astrobin to get some ideas of exposure and gain times. It seems to me that people are able to get very good data at very low gain. One had gain at like 60. However, it was great looking. They were running 600 second exposures and had 10+ hours of integration, so I am assuming that is why. But, is there a way to gauge your gain per target? Is it just by taking test/framing shots and going from there? Or is it taking test shots and looking for something specific to help choose the gain setting. I was reading that the higher the gain, the more noise, but if I had my gain down to 60, I feel as though I wouldn't be able to see any data. This is a fun learning experience and even though my first light image isn't spectacular, the whole process was exhilarating to do. 

 

If you would have the time to do so, please do toss out some tips and critiques, positive and/or negative. Maybe folks who have been doing this a while can tell what I have done either wrong or poorly just from the resulting image. Your time is greatly appreciated. Thank you all very much for taking a look! Clear skies!

 

 

 

Attached Thumbnails

• 

[KTAZ]

As a beginner, the rush is always to get anything on disk, so keep grabbing data!

 

However, there are so many things to learn. And you've not given any information on your rig or imaging chain, so, little to help with.

 

With regard to gain, if a CMOS your camera should state a "unity" gain in it's manual (where e-/ADU=1). This is where you should start. It provides the lowest "read noise".

[sterec]

I have the same camera as you and find that 300 s lights ,at unity gain for NB all the time, and it gives me the images that I want.  I do tend to try and take 15-18 exposures in each set of NB pictures.  Also calibration frames are a chore but make a startling difference to the final image.  To quote Dylan O'Donnell - Dither or Die.  I don't think that you are too wide of the mark.  I would be pretty happy to own this image - TBH.  the dark tendrils at 5-7 o'clock of the nebula look sharp.  I still continue to struggle with the "intuitive" software but having stuck with one, get better with practice.  I bet if you did just the Ha in black and white you would see that you have an image that is compatible with anything out there.  Also you are likely to be the most unappreciative of your image. So in summary just take more images.

[bobzeq25]

 

With regard to gain, if a CMOS your camera should state a "unity" gain in it's manual (where e-/ADU=1). This is where you should start. It provides the lowest "read noise".

Depends on the camera.  You need to look at the actual specifications.

 

And, there are tradeoffs.  You don't necessarily want the lowest read noise.  It often comes with lower dynamic range, because the full well capacity is so much lower.
 

[f430]

If you used the ASI 1600 MM camera 'Unity' is 139 gain - 21 Offset. That's a good place to leave it, until you find a reason to change it. Some targets will do better at different settings, but most all targets will look fine at 'Unity'.

I use Sequence Generator Pro, but I'd guess that NINA has a similar setup for helping to judge exposure times. In SGP, under 'Image Statics' are several parameters; Minimum, Maximum, Mean, etc, and Half Flux Radius. This could also be named FWHM, or similar. 65509 is the highest value it can show, and is when a star for instance, is completely over exposed and will show as pure white. 

After setting your camera to 'Unity' take a single image with your Ha filter. With that image on the computer screen, move your mouse around and onto a bright star. With the mouse pointer on the center of the star, look at the Half Flux Radius, FWHM or similar and see what value it shows. If it reads full value (65509) move the mouse pointer to a smaller star and check its reading. If most or all the stars read that Max value, your exposure time is too long. Take another exposure at say 60 second less, and do the same test. 

What you want to end up with is an exposure time where mid range size stars read around 30,000 to 45,000, and if a few of the really bright stars are still at 65509, that's okay.

What you'll end up with is that the stars Below the Max value will show color when you process the images, and the stars at Max value will show as white. If you can get all the stars below Max Value, so much the better, but then your image might suffer from too short of an exposure time.

Best for now to not worry about a few large over exposed (blown out) stars. 

HOO is a good combination to learn with. The SHO is more colorful, but also more difficult to pull off.

One I did was 20 images each of Ha and OIII, at 120 sec, unity 139-21, with a 80 mm Explore Scientific scope. 

Edited by f430, 28 November 2020 - 07:19 PM.

[Madratter]

Depends on the camera.  You need to look at the actual specifications.

 

And, there are tradeoffs.  You don't necessarily want the lowest read noise.  It often comes with lower dynamic range, because the full well capacity is so much lower.
 

 

Yes there are. That is something that is often lost in the discussion around here. A good example is actually the aforementioned 1600mm. At unity gain it has a full well of just around 4000. At lowest gain setting (0), it has a full well of 20,000. That is 5x better.

 

Meanwhile noise goes from 1.75 at unity to just over 3.5. That is only twice as much noise for a gain of a factor of 5 with full well.

 

And a less than totally appreciated fact around here is that snapping lots and lots of short photos builds up read noise faster than long, slow exposures simply because you aren't doing that reading near as often.

 

It especially makes no sense to stay at unity gain with the 1600 if your exposure time you are using already is well over what you need to be sky limited. And that is probably the case for a whole lot of people, especially those with significant light pollution (or just the moon being up).

 

In other words, as you say, it very much depends on the camera. It also depends very much on your conditions. The 533 color is a different kettle of fish.

Edited by Madratter, 28 November 2020 - 08:26 PM.

[f430]

Here is a short article regarding various Gain and Offset settings for the ZWO ASI 1600 MM and MC cameras, from ZWO. This is straight forward and easy to understand for new users. 

https://astronomy-im...com-driver.html

[Madratter]

I have to say that article is pretty dreadful for a whole bunch of reasons. It isn't that it is wrong. The problem is that it doesn't give you very actionable information. And it neglects some rather obvious effects.

 

The biggest problem is that no one in their right mind who actually understands what is going on would use the same exposure times for those different settings. The article does demonstrate what happens if you do. I'll not dispute that.

[f430]

Crookedeyes, Here is a quick re-adjustment of your image. 

It shows that you have a lot of nice color in there if you play with it a bit.

Attached Thumbnails

• 

Edited by f430, 30 November 2020 - 07:11 PM.

[CrookedEyes]

I collected more data a couple days later, the night of 28-29 November...more days than the first night, as well as spent much more time processing in Photoshop. It is taking a bit to get used to Photoshop again, but it's slowly coming back. I used only the second set of data because I actually took calibration frames as well that night. I would have added the first night into the stack but wasn't sure how much it would degrade the final stacks if I didn't have calibration frames for that night. I feel better about this image. I feel like it turned out pretty well for the second time out...but I know it can be better. Hope y'all like it. 

Attached Thumbnails

• 

Edited by CrookedEyes, 07 December 2020 - 03:52 PM.