Looking for process feedback about this picture of Orion nebula

[morphosis7]

Hey everyone,

I'm looking for some advice on ways to improve my results on an image like this - and to confirm some suspicions of mine - specifically around the capturing of images. My setup is:

• stock Canon a6000 camera body, on a tripod but no skytracker mount, in either manual or shutter-priority mode
• 210mm lens, manually focused at the point where the a6000 first reports "infinite" distance (so not buried all the way)

The images are:

• 146 light images - RAW format, 2 sec exposures, f/6.3, and ISO 3200 (the JPEG version of one of these is attached)
• 11 dark images
• 10 bias images

Processing is in DeepSkyStacker, with post-processing in GIMP.

I took these just outside Tusayan, Arizona a few days ago - I don't remember the specific conditions from ClearDarkSky but it was at least decent.

Areas for improvement:

1. The nebula's structure is largely illegible. I _think_ this can be improved with a lower ISO speed.
2. The background is super noisy. I suspect this comes down to post-processing, and ISO speed.
3. The image is basically monotone, rather than color. Honestly, I'm a bit lost here.

Thoughts on possible tweaks:

• Without a skytracker mount (is that the right term?), I don't think going to a longer exposure would help. At 210mm, the star trails show up really quickly - and I don't think there's a way to counteract them in post-processing.
• The ISO speed was probably too high, which left me collecting so much light that the main body of the nebula gets blown out. Maybe dropping to ISO 800 or so would help preserve the nebula's structure, and possibly even out the background.
• I'm not using flat images, in part because I've been unable to figure out how I should capture them. I drive out to a capture site, and need to pack the camera up to head home - but the documentation I've seen on flats is that you need the same physical setup pointed at a brightly-lit flat white surface? Do people image those in the field? Or are they meant for rigs that don't dissassemble?
• Am I right in thinking that focus isn't a problem I should focus on? It may not be perfect, but the fainter stars in this image seem okay.
• It isn't lost on me that many people here with great images are using much longer exposures than I am (60 minute exposures, or more). That would probably help as well, as ~5 minutes of light time is a pretty small exposure.

 

[PeterCPC]

Hello and well done for trying this. The problem is that, if you are limited to such short exposures, you are never going to get much detail. What you need is to be able to track and have long exposures over a long period of time. Tracking is essential IMO. I would limit ISO to 800 or less. You could try a bit of "lucky imaging" where you take hundreds of images of short duration and stack them. This would improve on what you have so far. Have a look here.

Good luck.

Peter

[happy-kat]

A member very recently took the same target using 2 second images but they took 500 images the image is in the No "EQ" DSO Challenge thread. Isn't the 6000a a Sony camera and not canon? Ideally if you are going to use calibration files aim for at least 20 of each. I use my 7 inch tablet running application lightbox to take flats at the end of the imaging with darks, bias and dark flats. Also use something to remotely trigger the shutter. After using DSS you will need to crop the edge of the image at the start of processing. I think the image is noisy only because of keenness to pull as much as you could but you need hundreds more images of that exposure length. So keep the data just get some more and add to it.

[steppenwolf]

This is a great attempt but your biggest limitation here is the exposure lengths and the only way to increase them is to use an equatorial tracking mount of some kind.

6 hours ago, morphosis7 said:

It isn't lost on me that many people here with great images are using much longer exposures than I am (60 minute exposures, or more). That would probably help as well, as ~5 minutes of light time is a pretty small exposure.

Although it is true that I routinely capture half hour exposures, I am working in narrowband - a five minute exposure will capture huge amounts of lovely data and a simple tracking mount camera mount like the Sky-Watcher Star Adventurer will allow you to do this.

[knobby]

As above comments, take way more exposures. Also looking at the cropped image the donut looking stars makes me think it's slightly out of focus. Try a DIY Lord mask or Y mask for focusing.

Maybe Google DIY barn door tracker, even a basic clockwork or balloon driven type would allow maybe 15 seconds if polar aligned.

But keep at it, interesting project.

[SamAndrew]

The 2nd picture isn't blown out so the ISO setting isn't too high. Higher ISO will generally result in less read noise, which will be significant with such short exposures, I would keep the ISO high.

You've got a good image there, it just needs better processing. The core has been blown out in the processing. If you can post the autosaved file from deep sky stacker I'll have a go at processing it when I'm home later.

I wouldn't bother with bias files, but I would add more dark files (but these need to be taken at a similar ambient temperature to the lights). The bias signal is in the dark file, although with 2 second exposures there's probably very little difference between the two.

Flats can be taken at home - ideally don't remove the lens, and don't change the focus, put a piece of paper or a t-shirt over a computer monitor and put the camera up against the screen, turn the ISO down, maybe turn the monitor brightness down, you want the exposure to be several seconds long otherwise the image will be unevenly illuminated due to the screen flickering. You then want to add Dark flats files that are the same length and ISO as your flat files.

[don4l]

I agree with Sam. 

There is a lot more information in your image than we can see in the close up.  The widefield has colour that has been lost in processing somewhere.

When you are processing in Gimp, one of the first things to try is Colours -> Levels -> Auto Input Levels.   This will give you an idea of what is actually there.  You can almost always get a better result adjusting the curves manually.

It would be a good idea to post up the stacked data before you process it in the Gimp and let others have a look at it.  If you do this, then it would be a good idea to crop the image for people who don't have high speed internet access.

 

One final thought, make sure that you have a 16 bit version of the Gimp (Version 2.10 or later).

[happy-kat]

I found the post here link here and regarding a home made barn door, I made one here and it is totally manual but can get 3 minute exposure s reliably when using a 40mm camera lens.

[SteveNickolls]

Hi and thanks for posting your images. You have done well with your equipment and it's exciting what a DSLR and lens can show in the night sky. As others have said with a static tripod and camera/lens you will be severely limited in how long you can expose for without star trailing showing up. I can vouch for the use of a relatively inexpensive EQ mount such as the SkyWatcher Star Adventurer (SA) mount, you should be able with practice to take much longer images. There is a thread on here about imaging with the SA mount-

 

I built a barn door tracker for next to nothing following this template in a past copy of the Sky at Night magazine-

http://www.skyatnightmagazine.com/feature/how-guide/how-build-tabletop-tracking-mount you can download templates from here. Like happy-kat I was able to get multiple minute exposures at short FL's and it is a great practical way to introduce polar alignment etc.

Now it's best to take your time when focusing the camera/lens on a bright star, slightly moving in and out of focus until you are happy you have things as precise as possible. At 210mm bright stars should easily permit you to focus accurately, it's more difficult when you use shorter FL lenses. I use Canon DSLR's and cannot speak for the A6000 Sony model of yours but if you have a magnify capability and 'Live View' these will help you judge focus better. Using a higher ISO when focusing will help show stars up but using too high an ISO when taking the actual exposures will introduce noise into the frame and actually doesn't add anymore detail into your image, it just looks lighter. Experiment with ISO.

Good luck with your future imaging, there's much you can do with a limited amount of equipment!

Cheers,
Steve

[happy-kat]

Let the light histogram guide your ISO selection, you are looking for the histogram peak (all of it) to be clear of the left edge even up too the middle is workable but I prefer to aim for the peak to be left of the middle.

[morphosis7]

Hey everyone, thanks for the encouragement and the recommendations that I've seen so far. There's a few things that people have said that I wanted to respond to directly...

11 hours ago, happy-kat said:

Isn't the 6000a a Sony camera and not canon?

Yeah, you're right. The camera is a Sony, not Canon. Silly mistake...

10 hours ago, SamAndrew said:

If you can post the autosaved file from deep sky stacker I'll have a go at processing it when I'm home later.

Thanks! I'm attaching the autosave file here. I'd love to see what more skilled hands than mine can achieve with this. I've attached the original autosave at full resolution, and then a cropped version that pretty closely reflects what I posted above.

10 hours ago, SamAndrew said:

I would add more dark files (but these need to be taken at a similar ambient temperature to the lights)

Yeah - my process has been to capture the light frames in the field, and leave some time at the end to capture dark frames immediately afterwards; still in the field, nothing has touched the camera other than me adding the lens cap. Then I drop the exposure time to 1/4000 second and fire off the bias frames so all three batches come in one timeframe.

I can hear the feedback that maybe the bias images aren't needed with a 2-second exposure. Is there a threshold at which the bias frames become necessary? When I use a shorter lens, I'll bump up exposure times - getting up to 15-30 seconds if I've got my wide angle lens on. By that point I'm assuming that a bias frame is more useful?

10 hours ago, SamAndrew said:

put a piece of paper or a t-shirt over a computer monitor and put the camera up against the screen, turn the ISO down, maybe turn the monitor brightness down, you want the exposure to be several seconds long otherwise the image will be unevenly illuminated due to the screen flickering.

Thanks for this - I'll give this a shot soon and see how it affects the outcome.

10 hours ago, don4l said:

One final thought, make sure that you have a 16 bit version of the Gimp (Version 2.10 or later).

Just checked - I've got the latest version, 2.10.8.

12 hours ago, happy-kat said:

Also use something to remotely trigger the shutter.

Yeah, I forgot to mention this in my setup. I've got an infrared remote to trigger the exposures, which has helped speed things up quite a bit. I'd originally used the 2-second delay timer to let my finger get clear of the camera and stop moving, but that's a lot of wasted time - and even then things don't always stop moving in time.

1 hour ago, happy-kat said:

Let the light histogram guide your ISO selection, you are looking for the histogram peak (all of it) to be clear of the left edge even up too the middle is workable but I prefer to aim for the peak to be left of the middle.

Ooh...that's a good piece of advice. I'm headed back out tonight to a place east of Phoenix, so I'll try this out and see how things go.

For everyone who has shared links for sky trackers and barn doors, thanks - I'll look to see what I can put together on this front. A DIY approach is probably going to be my friend until I can piece together the money for something commercial, but it is helpful to know what my next purchase should be.

Autosave.tif

Autosave_crop.tif

[happy-kat]

There a link to my barn door build in my signature when seen on a desktop web browser. Good luck with your captures tonight.

[don4l]

I've had a play with your cropped image and here are some comments.  I must stress that I am not an expert.

My opinion is that you should concentrate on one thing at a time.  If you can get up to 10 second exposures I think that you will see a great improvement.   After that you can play with things like ISO settings or longer exposures.

BTW Something appears wrong with your tiff file format.  I was able to open it in the Gimp, but many functions would not work properly(weird!!!).  Windows Picture Viewer reports that "It appears that we don't support this format.".  Picture Window Pro wouldn't open it at all.  Does it display properly in your file browser?

 

[SamAndrew]

Not going to go too deep with the processing, this is Auto Background Extraction and then Masked Stretch in PixInsight, and then I've bashed it quite hard with Colormancer Noise reduction in Photoshop and then reduced the star sizes and upped the vibrance/saturation, and then back into PI for some final curve and contrast tweaks. The stars on the right are quite round, so the stretching of the ones on the left is caused by the lens, not the lack of tracking.

[tooth_dr]

8 minutes ago, SamAndrew said:

Not going to go too deep with the processing, this is Auto Background Extraction and then Masked Stretch in PixInsight, and then I've bashed it quite hard with Colormancer Noise reduction in Photoshop and then reduced the star sizes and upped the vibrance/saturation, and then back into PI for some final curve and contrast tweaks. The stars on the right are quite round, so the stretching of the ones on the left is caused by the lens, not the lack of tracking.

That’s pretty impressive processing ??

[morphosis7]

On 29/12/2018 at 04:04, SamAndrew said:

Not going to go too deep with the processing, this is Auto Background Extraction and then Masked Stretch in PixInsight, and then I've bashed it quite hard with Colormancer Noise reduction in Photoshop and then reduced the star sizes and upped the vibrance/saturation, and then back into PI for some final curve and contrast tweaks. The stars on the right are quite round, so the stretching of the ones on the left is caused by the lens, not the lack of tracking.

Wow! This is a marked improvement over what I was able to achieve. Thanks for sharing this.

Seeing that an image like this is hiding in the data I've captured is really encouraging - and gives me a path forward to potentially re-process some of the images I've already captured.

The other night I was able to capture about 200 more images of the nebula - bringing my total up to 324 frames, or almost 11 minutes of light time. I also tried to capture to flat images using the process that Sam described. I'm working on re-processing the total set of images now, and will keep experimenting over the next few days.

Thanks for all the help in this thread, everyone!

[SamAndrew]

Turns out flats weren't too critical at f6.3, it didn't pose a problem in processing. To get more data you could try upping the aperture, try f4 or faster if the lens allows.

[morphosis7]

Thanks for that tip - I'll give that a shot the next time I'm under clear skies (my trip to Arizona is over, so I'll back in Boston now)

[Chris33benoit]

Wow, I am just starting too to shoot Orion and so far I am please with the results.  I don't have a tracker (yet) and I shoot with a tripod.  Settings are Nikon D3200 DSLR camera, ISO3200, metering matrix, 4 sec shot, focus manually (took me about 2 minutes just for focusing), white balance auto, lens sigma 70-300mm @ 300mm, vivid color, processed in DeepSkyStacker and Photoshop.  The first picture is from a 500mm (@500mm) fixed lens by Opteka on my D3200.  Does anybody talk about the barn door NYXTECH star tracker by any chance?  I thought about buying one soon, might help with my pictures.  The price for those trackers are about 130$ US.  Thanks!  

[happy-kat]

Do you have a link to the NYXTECH please?

That's a good start with your images. At those  focal lengths you really need either 1 or 2 second exposures and take around 200 images. It's surprising what can be revealed on this bright target.

[Chris33benoit]

https://nyxtech.us

Here is the link for that product.  Again if someone would gave their opinions about this tracker it would be awesome. Thanks again.