Orion - Flame Nebula, Horse's Head, M42 etc. and lots of hydrogen

[Gina]

Captured on Friday night last week (3rd January) with 3nm Ha filter, 135mm f2.5 lens and ASI 1600MM-Cool camera on EQ8 mount without guiding.  55 subs of 4m each giving a total of 3h 40m.  Captured with RPi 3 running INDI firmware and saved in Linux Mint desktop indoors running KStars/Ekos.  Simple processing in PixInsight for star alignment, stacking and curves, followed by cropping and scaling for posting here.  No calibration frames were used. 

Here are three different fields of view, obtained by cropping the image.  First is about a third of the captured frame and third is full resolution (4656x3520 pixels cropped to about 1200).

Edited January 8, 2020 by Gina

[Pete Presland]

Some lovely faint dust visible there, especially around the Horsehead!

[DaveS]

Some lovely faint detail in th Curtains Gina. I see Alnitak showing some interesting diffraction patterns, was the lens stopped down at all?

Oh, and it's great to see you posting in the Deep Sky Imaging section again!

[Gina]

Than you Dave.  The lens wasn't supposed to be stopped down but I noticed the diffraction pattern and wondered I I managed to turn the iris ring by mistake.

Yes, nice to be back.  Makes a change to have something I think worth posting!

[carastro]

The Horsehead is smashing. off looking Alnitak though.  Nice to see you imaging again.

Carole 

 

[Gina]

Next dry day I'll pop out and check the lens.

[DaveS]

6 minutes ago, Gina said:

Than you Dave.  The lens wasn't supposed to be stopped down but I noticed the diffraction pattern and wondered I I managed to turn the iris ring by mistake.

Yes, nice to be back.  Makes a change to have something I think worth posting!

I did the same thing when I made a LRGB image of the Sword and Flame, accidentally stopping the lens down from f/3.4 to f/5.6 when I made the Luminance exposures, so not only did I have huge diffraction spikes, but the L didn't bring as much to the party as I'd hoped. Dur...

 

[Gina]

Just been out and checked the lens iris and it's wide open - nothing I can see that would cause a diffraction pattern.

[cuivenion]

Telephone line are some such thing?

[Gina]

Nothing like that in the FOV.

[wimvb]

Nice to see you in this section again, Gina. And also nice to see different clouds than the terrestrial kind.

I think that the diffraction may very well be caused by minor flaws/intrusions near the edge of the lens. How old is the lens, and especially, how clean is it on the inside? Alnitak is so bright compared to the rest in this image, that it doesn't take much to cause diffraction. 

[Gina]

Thank you Wim.  The lens is a vintage SLR film camera lens.  Asahi Super Takumar as used with the Pentax cameras.  42mm x 1mm thread.  There was a smear on the front of the lens which I have removed - the insides look clean.  Alnitak would probably be improved by shorter exposures and/or lower gain.  I'm still experimenting to get the best settings for this camera.

[wimvb]

8 minutes ago, Gina said:

Alnitak would probably be improved by shorter exposures and/or lower gain.  I'm still experimenting to get the best settings for this camera.

The sensor has its highest dynamic range at low gain (0 - 20), that's what I normally use with my ASI174MM-Cool. But that's for rgb imaging. Most people use a high gain (250) for nb imaging, probably to keep the exposure time short. Personally, I'd start with an exposure time that has Alnitak somewhat over exposed (but not too much), and the flame/horsehead very weak in a single sub. Cooled cmos do better with many but short exposures than fewer long exposures. 

[Gina]

Went back and checked - gain was 360 which I think is too high.  Not sure about 240s exposure but I don't want to lose the faint stuff.  Maybe reduce gain to the next "step" down of 240 (I know you can set any value up to 600) but keep to the same exposure.  Advice from ASI1600 (or other CMOS) experts always very welcome.

[Gina]

I was considering a 200mm lens for more resolution but these are f4 - almost 2 stops slower.  That would mean at least 6-8 hours total imaging time.

[wimvb]

8 minutes ago, Gina said:

Not sure about 240s exposure but I don't want to lose the faint stuff.

That's the challenge with cmos (or in general, really). Faint signal seems hidden in the noise, but with many exposures, it's revealed again. Jason Guenzel on Astrobin once wrote that he never shoots fewer than 80 subs per filter, and would rather have 100 or more subs. Because the dynamic range of cmos cameras is low at high gain, you need many subs to recover it.

2 minutes ago, Gina said:

I was considering a 200mm lens for more resolution but these are f4 - almost 2 stops slower.

135/2.5 = 54 mm aperture. 200/4 = 50 mm aperture. But the 200 mm lens has a different pixel scale.

Light gathering power on a pixel is proportional to (rD)^2, where r is pixel scale and D is aperture, or (pF)^2 where p is pixel size and F is F-number.  The faster lens is better as far as light gathering power is concerned. But if the longer lens is of better quality, you will still get a better image out of it.

You could use a stop down ring for the 135 mm lens to see if that cleans up any diffractions. If the ring takes 4 mm off the aperture, you still have a f/2.7 lens. You could go down to a 45 mm aperture and still have a f/3 lens. That is, if the diffraction around Alnitak is caused by something in or on the lens.

[Gina]

I presume you mean an external aperture ring as stopping down with the iris causes diffraction spikes. 

The 200mm lenses are probably about the same quality being the same make and similar vintage.

[Gina]

Would I get as much very faint nebulosity with say 100 2m subs as 50 4m subs, other things being equal?  I know that a lot of my experience with CCD sensors doesn't apply to CMOS.

[wimvb]

22 hours ago, Gina said:

presume you mean an external aperture ring as stopping down with the iris causes diffraction spikes. 

Yes. Step down ring for filters will do the job. You could 3d print a ring, but if its inner edge isn't smooth it will cause its own diffraction. 

[wimvb]

21 hours ago, Gina said:

Would I get as much very faint nebulosity with say 100 2m subs as 50 4m subs, other things being equal?  I know that a lot of my experience with CCD sensors doesn't apply to CMOS.

For ordinary rgb imaging, where you stay safe of the read noise floor, I'd say no problem. But nb imaging is a bit trickier, since the signal levels are much lower. With a f/2.5 or f/3 lens, I would definitely try it. The theory is the same for cmos as for ccd; you need to stay above the read noise floor. Ie, your exposures need to be dark noise or light pollution noise limited. Cmos (except the very new generation) generally have higher dark current (noise) than ccd. This, combined with the lower read noise, allows the shorter exposures. But you need the total integration time. "There still ain't no free lunch."

[Gina]

Thanks Wim.  Yes, I know it all depends on total integration time.  What I really want to know is whether there's any benefit from using 2m exposures rather than 4m in view of the greater disk space required. 

[wimvb]

I agree that 100+ images will fill any hard drive in no time. Do you save lically on the Pi while imaging?

Can you post a single sub with only STF stretch applied? 

[Gina]

No I save on an SSD in my indoor desktop running Mint.  I move files off that onto a hard drive every so often to make space.  I still have 200odd GB free on the HD. 

I'll sort out a single sub...

Edited January 11, 2020 by Gina

[Gina]

One of the best subs with STF applied.  Firstly a screenshot of the PI screen then a PNG format of the full size, full resolution image (13.4MB).  Or did you want the XISF format (32.8MB)?

[wimvb]

Png + screen shot were fine, no need for xisf. It seems to me that you can reduce the exposure time, or the gain if you don't want to fill your hard drive/sd card. Unlike ccd, with cmos there's one more parameter to optimise, gain. And that parameter affects others, unfortunately.

You could look around on astrobin what others have done with similar equipment.