Home · Register · Join Upload & Sell

Moderated by: Fred Miranda
Username  

  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Post-processing & Printing | Join Upload & Sell

1
       2       end
  

Archive 2015 · Star stacking - tougher than I thought? :)

  
 
charlesk
Offline
• • • • •
Upload & Sell: Off
p.1 #1 · p.1 #1 · Star stacking - tougher than I thought? :)


Went to shoot the Milky Way deliberately for the first time last night. Noticed quickly that I could not get a long enough exposure without trails developing. So I decided to take a number of 15 second images and then try to stack them.

So far all my results have been awful. Everyone seems to use Deep Sky Stacker but every attempt thus far has given me a blurry, mushy, monochrome mess.

I am not sure if there's a learning curve here.. or maybe I am using the wrong tool (perhaps it is only really intended for use with telescopes to see deep sky objects, as the name suggests.) I am trying to take a more conventional landscape shot with the Milky Way above.

Your thoughts appreciated!



Jun 25, 2015 at 12:53 PM
D_Sh
Offline
• •
Upload & Sell: Off
p.1 #2 · p.1 #2 · Star stacking - tougher than I thought? :)


Due to earth rotation stacking 15 second images are not going to work, you are going to have star trails. For Milky Way I would advise to go wide open and high ISO. I would stack few images (usually two for me) if there is something on foreground to have everything in focus.
http://club.foto.ru/gallery/photos/2331154/?&author_id=23881&sort=date&next_photo_id=2332925&prev_photo_id=2330893



Jun 25, 2015 at 01:55 PM
dgdg
Offline
• • • • •
Upload & Sell: Off
p.1 #3 · p.1 #3 · Star stacking - tougher than I thought? :)


Great web site and images Charles!

DSS has a small learning curve, but it works very well for wide field astro-landscapes.
The DSS final sky image will look bright, washed out, and monochrome - this is expected. You process the image in PS. Below is a composite image using DSS (six 4 minute exposures).

15 seconds untracked is actually too short for 24mm focal length (full frame) or wider. Use the rule of 600/focalLength (or 500) to determine longest possible shutter speed.

Easiest way on a forum to sort this out is to post the following:
1. jpeg of a single sky image with exif, unprocessed
2. jpeg of the DSS TIFF output, unprocessed
3. jpeg of your land + sky composite, processed, if you have one
3. Gear used and exposure/camera settings
4. DSS settings used for each window (screen shots may be the easiest)

David


http://davidsphotography.zenfolio.com/img/s7/v169/p1102015759-5.jpg



Jun 25, 2015 at 02:05 PM
Ian.Dobinson
Offline
• • • • • •
Upload & Sell: Off
p.1 #4 · p.1 #4 · Star stacking - tougher than I thought? :)


OK I'll ask this question in ignorance as I will confess I've never done anything like this .

but here goes anyway

If your stacking multiple underexposed images to get a fully exposed one (which leaves you a mushy mess )can't you use the same image stacked multiple times ? .
sort of like taking a single image and duplicating onto multiple layers and then using the screen overlay .



Jun 25, 2015 at 03:06 PM
dgdg
Offline
• • • • •
Upload & Sell: Off
p.1 #5 · p.1 #5 · Star stacking - tougher than I thought? :)


Hard to say what the mush is without more info.
Stacking software like DSS does not increase the brightness (signal) of an image. It increases the signal to noise ratio.



Jun 25, 2015 at 05:03 PM
charlesk
Offline
• • • • •
Upload & Sell: Off
p.1 #6 · p.1 #6 · Star stacking - tougher than I thought? :)


Thanks for the replies all. dgdg you are very kind and I am impressed by that shot.

I guess I expected it would stack the images together so the stars would get brighter. If it improves the SNR then that's basically the same thing. But what I actually saw as output was considerably noisier.

I was shooting at 16mm. I have noticed trails start on this lens around 30s and I did take some regular shots at 20s but they are quite dark. I went to 15s to "be safe" figuring since it's going to stack things up anyway the length of the individual shots didn't matter.

This was somewhat spur of the moment and I guess I should have prepared myself better. I may have one more opportunity tonight to try again though the moon is getting pretty bright.



Jun 26, 2015 at 10:18 AM
dgdg
Offline
• • • • •
Upload & Sell: Off
p.1 #7 · p.1 #7 · Star stacking - tougher than I thought? :)


Hard to say anything about your recent output without more info.

Since stacking primarily reduces noise but does not increase the signal, you do need to make sure every capture is optimally exposed. Here are my basic tips.

Shoot with aperture wide open. Some lenses may need to be stopped down a bit for sharpness/coma, but it is preferred to have a lens that performs well wide open. Now Jim Fox often uses f/4 on a f/2.8 and will focus on the foreground. This will make both the stars and foreground look reasonably sharp with a single capture. If you are making a composite, then I recommend using the largest aperture since you will be focusing on the foreground and sky separately.

Pick the maximum shutter speed possible for your field of view. 500/16 would give about 30 seconds with minimal star trails on a full frame body. You might see a tiny trailing at 100% but it should be trivial. Don't mistake bad coma for star trails.

Then look at your histogram. You want the luminence peak to sit in the middle of the 20% and 40% marks. Now adjust your ISO to move the histogram peak to the proper place. It will look too bright on the lcd preview, but you have to ignore this and go only by the histogram.

Now you have systematically determined the optimal exposure settings for your star shot.

David



Jun 26, 2015 at 12:27 PM
charlesk
Offline
• • • • •
Upload & Sell: Off
p.1 #8 · p.1 #8 · Star stacking - tougher than I thought? :)


Thanks David.

Unfortunately it looks like cloud cover tonight. I will try to post some of the shots I already have and then maybe you can suggest what I can do. It will be good education for the next try.

I would love to get the new 11-24 but it's pretty pricey. The 16-35 is a decently good lens but not exceptional. It definitely shows a lot of distortion on the edges at 2.8 or 4 and you are correct that I may be confusing that for trails. 30 seconds has traditionally been my "limit" unless deliberately trying for trails.



Jun 26, 2015 at 02:09 PM
dgdg
Offline
• • • • •
Upload & Sell: Off
p.1 #9 · p.1 #9 · Star stacking - tougher than I thought? :)


Sure!
The 11-24 is very nice (sharpness and low coma), but at f/4 you need a higher ISO compared to f/2.8 lenses.

The Samyang 14mm 2.8 is cheap and if not decentered, a wonderful lens for wide field milky way shots. Coma is nil and distortion is not noticeable with a night sky. Everyone that shoots wide field astro should have this lens if available for their camera mount.
Some lenses to consider in addition -
The 17-40mm f/4 has low coma and not too expensive.
The 24-70mm f/2.8 Mark II. Fabulous
The new tamron 15-30 f/2.8 is sharp with low coma. Used it is about $1k on amazon.
The sigma Art 35mm is worth considering but not really wide field for the milky way.

Good luck!

David



Jun 26, 2015 at 02:22 PM
OntheRez
Offline
• • • • •
Upload & Sell: On
p.1 #10 · p.1 #10 · Star stacking - tougher than I thought? :)


David,

Forgive my ignorance, but I'm not sure what you are referring to when you note "low coma." I've been fiddling with the Samyang 14mm which is an amazing bang for the buck, but have yet to get anything worthy of keeping. Staying at it though.

Robert



Jun 27, 2015 at 10:25 AM
charlesk
Offline
• • • • •
Upload & Sell: Off
p.1 #11 · p.1 #11 · Star stacking - tougher than I thought? :)


Robert, google "coma distortion". It's a type of lens distortion that causes comet-shaped sorts of results, which is where the name comes from.

David, I didn't get to shoot again but I have some samples here that will hopefully illustrate my confusion with DSS.

Here's one of the original 15s images (resized):

http://i.imgur.com/fhalR9b.jpg

Here's what DSS did with 4 of these:

http://i.imgur.com/8BBaDE6.jpg

Now here's the result of manually "stacking" these simply by pixel-shifting them in PS and using "screen" blend mode:

http://i.imgur.com/AyPRuAJ.jpg

What I expected from DSS was something like what I got with my quick manual test, only even nicer. I am at a loss to explain the DSS output.

Thanks!



Jun 27, 2015 at 09:38 PM
OntheRez
Offline
• • • • •
Upload & Sell: On
p.1 #12 · p.1 #12 · Star stacking - tougher than I thought? :)


Charles,
Thanks for contributing to my continuing education. (And I'm not being sarcastic.) All I could think of was chroma and that didn't seem to be the issue here. So coma is actually sort of like comma - a dot with a tail. As I said, still haven't succeeded in this field but with the help of FM experts and continued experimentation, I will figure it out.

Robert



Jun 28, 2015 at 09:04 AM
AmbientMike
Offline
• • • • • •
Upload & Sell: Off
p.1 #13 · p.1 #13 · Star stacking - tougher than I thought? :)


I haven't really done much AP, but I asked a lot of questions on an astronomy forum, heading in the direction of AP.

Yes, I think it's quite hard. Expect a steep learning curve.

You're supposed to get the exposure right, or as right as possible, then put the image in DSS. I think processing in DSS would lower the noise and make it possible to brighten the image, though.

You kind of have to get what you can & process it later with dim objects, alough this isn't so true with the Milky Way, it's bright. A lot of processing in AP.

A telescope is a camera lens & vice versa.

If you can't get a bright enough image, you need a drive, I didn't have one, a major reason I haven't done much AP.

You can run the ISO up, and use a 24/1.4 on ff at 3200.

I got a nice Milky Way shot by running up the contrast into the stratosphere and converting to B&W.

Hers a good overview imo:

Astrophotography without a star tracker: http://youtu.be/e0JSTF8SGi4

AP just doesn't seem to have the same sharness requirements as regular photography. Everything is moving, it's hard to get the drive aligned, etc.

You can learn a lot from an astronomy forum, plus you can ask a lot of questions on the problems that are virtually guaranteed to pop up



Jun 28, 2015 at 09:49 AM
mikedefieslife
Offline
• •
Upload & Sell: Off
p.1 #14 · p.1 #14 · Star stacking - tougher than I thought? :)


I've always wondered, how does stacking multiple underexposed images result in one bright image? does the noise tent to move around so is different in each individual image?

How does DSS track the star? If I shoot 20 15sec images, the stars in images 20 are in a significantly different place to image 1.

I need to learn more about this stuff.



Jun 28, 2015 at 12:15 PM
dgdg
Offline
• • • • •
Upload & Sell: Off
p.1 #15 · p.1 #15 · Star stacking - tougher than I thought? :)


mikedefieslife wrote:
How does DSS track the star? If I shoot 20 15sec images, the stars in images 20 are in a significantly different place to image 1.

I need to learn more about this stuff.


Your electronic tracker tracks the stars so that they stay pin point instead of trailing (ex Ioptron, Vixen Polarie, Astrotrac).

DSS maps a certain number of bright stars for each image, allowing it to rotate and align all the images on top of each other in order to reduce the noise. Depending on the amount of overlap, your image may be effectively cropped some.

David




Jun 28, 2015 at 08:30 PM
dgdg
Offline
• • • • •
Upload & Sell: Off
p.1 #16 · p.1 #16 · Star stacking - tougher than I thought? :)


Charles, the DSS output looks ok.

The glaring issue (excuse the pun), is you greatly underexposed. The deep (dark) sky histogram luminence peak needs to be around 35%. Underexposure will offset some/much of the noise reduction benefits of stacking. That being said, your exposure here is limited by the light pollution.
The city light pollution is tough. You can only do so much with this washing out the milky way and causing an uneven and colored sky illumination. Maybe a single exposure from an A7s would be better. Furthermore, it is hard to use luminence masks to process the deep sky and milky way separately since the light pollution is very bright. Luminence masks are a powerful tool for milky way processing. You need darker skies.
I detect some vignetting? Make sure to correct for vignetting in all of your raw images prior to exporting TIFFs from LR. It cam be hard to correct later if you image long enough since the the stacked image's field illumination won't fit a simple lens profile anymore.

I played with the small file you uploaded. I spent just a few minutes with it. I typically will spend about an hour or three over several days processing my images.
I made a few adjustments.
black point (adjusted using levels on each rgb channel separately)
levels middle slider to tweak color balance of each channel
minimize filter by 1 pixel after selecting the highlights, expanding, feathering.
noise reduction
unsharp mask
high pass filter with overlay blend, opacity ~15%
I don't usually add much saturation, if any, but this one seemed to benefit from it.
The stacking did create some posterization artifacts on the lower left sea level area due to underexposed input images.

http://davidsphotography.zenfolio.com/img/s7/v152/p1285792999-5.jpg




Jun 28, 2015 at 08:51 PM
charlesk
Offline
• • • • •
Upload & Sell: Off
p.1 #17 · p.1 #17 · Star stacking - tougher than I thought? :)


First, thank you so much for the time you took to play with that.

I guess I'm still unclear on the concept here in general. I thought the entire point was that instead of taking a longer image that would have more noise and be more likely to have elongation, I take shorter images and then the software aligns the stars and stacks them, providing a longer effective exposure. That's basically what I did manually within Photoshop.

That doesn't seem to be what's going on here, if my 15s exposures are too short. Sure they are underexposed individually, but all 8 stacked together should have plenty of virtual photons.

To my eye my manual effort has a better output than what even you got from DSS.. and it was done with no processing at all (thus the vignetting).

So .. I guess I still am confused as to what DSS is buying me, at least in this case. I couldn't have gone much longer on the exposure (maybe 30s but that's pushing it) and the camera is too old to go much higher on ISO.



Jun 29, 2015 at 08:36 PM
dgdg
Offline
• • • • •
Upload & Sell: Off
p.1 #18 · p.1 #18 · Star stacking - tougher than I thought? :)


Sure!. Good questions. Your manual effort does seem better. I don't think DSS is helping you in this situation.

I don't understand the internal workings of stacking software, but I am pretty certain that they all work on improving the S/N ratio by reducing noise in the composite. Noise is random, signal is not, so after comparing all the individual photos the program can remove noise pixels and replace them with signal pixels. The actual brightness of an image is not changed even though the DSS output appears bright and washed out. I think stacking is a bit of a misnomer, but I'm not sure there is a better term for the process.

DSS does have some FAQs that discusses how much data you need for an expected result. Worth a read. Some deep sky imagers will spend many hours of total exposure time on an object. A 5 minute total exposure time with a camera is still quite short.

Here is a before and after example of a recent stack.
The first image is a single capture - 14mm, f/2.8, ISO 800, 4 minutes using an Astrotrac tracker.
The second image is the DSS output after stacking six images together with darks and bias frames.
The final processed composite image is posted above!
On my monitor, I do see subtle improvements in the milky way core details, but it is not a massive difference between the 4 and 24 minutes exposure times.

The major keys to the improved IQ are good skies and a low ISO. Only way to drop your ISO is with a small tracker. If you want your night skies images to pop like your wonderful daytime images, you should consider something like the Ioptron skytracker. It is cheap ($300 on Amazon), portable, easy to align with the phone app, and if used properly, works wonders for wide field images. The only downside is having to combine the sky with an untracked land image since the landscape will be blurred by the tracker's rotation.

It may be easier to toggle between the two images on my web page
http://davidsphotography.zenfolio.com/p451763782/h416f03c1#h416f03c1



Single capture
http://davidsphotography.zenfolio.com/img/s6/v149/p1097794497-5.jpg


DSS output of six stacked captures
http://davidsphotography.zenfolio.com/img/s6/v149/p1097795623-5.jpg



Jun 30, 2015 at 01:39 AM
JSecord
Offline
• •
Upload & Sell: Off
p.1 #19 · p.1 #19 · Star stacking - tougher than I thought? :)


for stacking to really work, you need to have each frame be properly exposed. I don't use DSS, I use a median stack in PS, but it's the same idea. What I do is shoot 10 frames, each iso10k or 12,800/ f2.8/ 10 seconds. The individual frames are pretty noisy (but not that bad with the d750), but after stacking there is basically no noise, and the stars are pinpoint (even when printing big). I combine the sky frames with long foreground exposures after


Jul 04, 2015 at 01:08 PM
dgdg
Offline
• • • • •
Upload & Sell: Off
p.1 #20 · p.1 #20 · Star stacking - tougher than I thought? :)


JSecord wrote:
for stacking to really work, you need to have each frame be properly exposed.


+1




Jul 05, 2015 at 08:58 AM
1
       2       end




FM Forums | Post-processing & Printing | Join Upload & Sell

1
       2       end
    
 

You are not logged in. Login or Register

Username       Or Reset password



This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.