Alley in Pisa, Italy

by Lars Kotthoff

submit your photo


Hall of Fame
View past winners from this year

Please participate in Meta
and help us grow.

Take the 2-minute tour ×
Photography Stack Exchange is a question and answer site for professional, enthusiast and amateur photographers. It's 100% free, no registration required.

Last weekend I was experimenting with capturing star trails. I tried taking a set of 9 photos of the stars and Milkway (ISO 1600, f/2.8, 30 seconds exposure). Here is resulting combined photo (stacked with StarStax):

blurry star trails image

The resulting image is very blurry. Can anyone tell me what to do differently next time so I get a better shot? My theories are:

  1. I should have avoided the Milkyway, as it is not going to come out crisp in a long exposure shot lasting this long.
  2. Wind may also have been a factor. I have a very cheap tripod, and it was rather windy that night.

Anything else that may attribute to this type of blur?

EDIT:

As requested, here is a single photo from the set:

single shot

Notice that I composed the shot with the Milkyway in it, which I believe is adding to the blurring effect in the final stacked imaged.

Also as requested, here is a 100% crop of the above image:

enter image description here

share|improve this question
3  
I think you answered your own question - cheap tripod + wind is not going to render small pinpoints very sharply. I have a moderately good tripod, and with wind also get poor results. –  MikeW Sep 3 '13 at 3:49
    
That appears to me not very blurry, just soft from denoising.Was it a zoom lens? If yes, then was the physical length of that zoom lens not at its shortest? If not shortest, then I would suspect the zoom mechanism has drifted a tiny amount with each shutter release inflicted vibration. I can't think of other reason why would those startrails go at odd angles, not the usual uniform circular pattern. –  Esa Paulasto Sep 3 '13 at 4:04
2  
A wide angle lens will allow stars both above and below the celestial equator to be visible at the same time. Those near the north pole will curve one way, those near the south celestial pole will curve the other way. –  Michael Clark Sep 3 '13 at 4:42
2  
Oh, but of course. Thanks, @MichaelClark, I live so near the North pole and it never occurred to me how it would look much different elsewhere :) –  Esa Paulasto Sep 3 '13 at 7:21
    
@FinerRecliner could you post another sample? A single exposure from the set of this stacked version? Any one of them. –  Esa Paulasto Sep 3 '13 at 19:52

5 Answers 5

I don't think this is a tripod stability problem. The star trails are all straight, not wiggly as they would be if the camera where moving during or between each exposure. The dimmer stars are easy to see as nine distinct dots. None of the dots indicate camera movement during any of the nine exposures, and none of the lines of the nine dots for each star are distorted as would be expected if camera movement were a problem.

I think this is a focus problem that is acerbated by the time gap between exposures. It may also involve blurring due to overexposing the brightest stars. The effect of slight defocusing or overexposure is that brighter stars also appear to be larger stars. Instead of drawing a smooth arc as one very long exposure would do, combining nine shorter exposures with significant gaps in between creates lines of variable width. The lines created by the brightest stars are widest at the points where the stars were centered during each exposure, and narrowest at the points in between those centers. Similar to this drawing:
enter image description here

These irregularly shaped lines may be causing our eyes/brain to see the trails as blurry. There also appear to be significant compression artifacts in the image that are arranged in an almost brick like pattern of alternating horizontal and vertical patterns.

Here are some things you can try:

  • Try reducing the overall exposure by reducing ISO and narrowing the aperture enough to reduce the net exposure by a couple of stops while also extending shutter speed to around 2 minutes or so. This will help you in two ways: the stars will be trails in each exposure (rather than dots) and the Milky Way will be much dimmer. Since it moves around the sky along with the stars you can't get a sharp Milky Way and star trails in the same exposure or stack of exposures.
  • Shoot on a cloudless night so that the movement of clouds isn't an issue.
  • Figure out what is going on with the weird compression in StarStax. I've seen images stacked using StarStax and never noticed that kind of compression before. Make sure your original images are saved as RAW files. Batch process them with a RAW convertor and export to StarStax as uncompressed TIFFs.
  • Adjust your White balance so that the dimmer stars are predominately white, not mostly blue. The human eye has the most trouble focusing on the blue end of the spectrum and this may also be contributing to the blurry perception.
share|improve this answer
    
The grid is 8x8 pixels of JPEG compression artefacts. So it has to be a stack of nine JPEG images where artefacts seem to get more visible "brick" form. –  Esa Paulasto Sep 3 '13 at 22:34
    
I see the issue you are describing when I zoom in very closely on the photo. I tried re-creating the image in StarStax with it's "gap filling" mode, which is supposed to resolve this issue. While the star trails look slightly better when zoomed in, the overall image still has the blurred effect just like in the original photo I posted. Here is the "gap filled" version: i.imgur.com/5lHl6xJ.jpg –  Finer Recliner Sep 5 '13 at 1:15
    
What, exactly, do you want the final image to look like? The Milky way will always be blurred if you are capturing star trails since it moves at the same rate as the other stars in the sky. Additionally, since you are using such a wide angle lens, the stars to the north and the stars to the south move around a different focal point (north & south celestial poles), so even using a motor drive will create blur near the celestial poles. Telling your stacking software to align the stars will also do the same thing, since the relative positions of the stars are changing as they move east to west. –  Michael Clark Sep 5 '13 at 4:17
    
After seeing the one exposure sample of the set, it now looks like out-of-focus issue as Michael Clark suggested. There's help for focusing to infinity in photo.stackexchange.com/questions/23972/… . –  Esa Paulasto Sep 5 '13 at 4:45
    
Sometimes what appears to be a focus issues is really an exposure issue. Blown highlights tend to blur on the edges... –  Michael Clark Sep 5 '13 at 4:49

Get a solid tripod. Use a solid tripod head.

Hang a weight from the head to increase the mass. A pail of water will do and you can vary the amount of water to increase/decrease the weight or tie the end of a heavy branch to do the same thing.

Use a remote shutter release.

Use a lens shade to keep stray light from the lens during the shot.

Good luck

share|improve this answer
1  
Along with the lens hood, putting the eye-cap (that bit of rubber on the camera strap) over the viewfinder can be helpful at reducing stray light, especially if there's a lot of light behind the camera (or if you're using a torch to see what you're doing). –  drfrogsplat Sep 3 '13 at 7:11
2  
Stray light via the viewfinder is rarely an issue when a properly aligned mirror is up during exposure. The benefit of the eye-cap is during metering, but most astrophotography is done with exposure set manually. –  Michael Clark Sep 3 '13 at 19:37

As others have said, a cheap tripod is a significant problem when doing long exposures. Look for more stable support to ensure good results when doing long exposures and when you may need to deal with wind-induced movement.

That said, I'm not convinced this image is actually "blurry." I agree that something about it doesn't look sharp, however look closely and there are areas that are sharp. It's easy to see clouds in the lower half of the image, and a closer look shows clouds on the upper half, too. Clouds are funny things: to look at one, you see this tightly-formed shape, but actually there's a haze ("clouds") throughout the much of the sky. I suspect the blurriness you see is just that: the haze from less-formed clouds. Try shooting on a cloudless night and I suspect you'll see much better results.

share|improve this answer

It appears you have done the images into JPEGs first. Then stacked all nine JPEG images one over another. This introduces an emphasis on the 8x8 pixels square of JPEG compression blocks, which then shows up on your startrails stack as a visible grid of straight lines over the entire image.

From your image I cropped a 4 by 4 grid of these artefacts, 8x8 pixels each, so the crop was 32 x 32. Then I resized it up to 10x magnification, resulting a 320 x 320 image to make it more visible here. I also marked (red square) one of those 8x8 blocks in the crop.

Perhaps this grid is what you meant with the blurry feeling? At least it definitely reduces the image quality. I suggest you use RAW format for shooting the original set, and then process the images as usual, but save them as TIFF 16-bit file format. The StarStax software can handle 16-bit TIFFs. This way you're not stacking JPEG compression over another.

enter image description here

share|improve this answer
    
I see the blocks you are describing. At first I agreed that it must be caused by the JPEG compression scheme. So I re-did the stacking, making sure to export the original images as uncompressed TIFFs before using them in StarStax. Then in StarStax, I doubled checked that I was exporting the stacked image as an uncompressed TIFF as well. The image blocks are still in the final image! It must be something that StarStax is doing... –  Finer Recliner Sep 5 '13 at 1:27

The fundamental problem here simply appears to be short exposures. If the sample of a "single frame" photo added in the recent edit is an indication of all the frames, you are quite simply not making a star trails frame.

To do star trails, you need exposures long enough to produce actual trails. The trails are not produced by stacking 50,000 short exposure shots such that each single points of light from a star "combine" to form a chain, and therefor a trail. The trails are produced by keeping the shutter open for minutes at least, and the passage of the stars across the sky during that duration will expose across a "trail" of pixels.

To get very long star trails, you can expose dozens, maybe a hundred shots that are several minutes long at least (or much longer, half hour even), and stack them.

The answer to your problem is quite simply that you are not exposing for anywhere nearly long enough to actually produce trails.

share|improve this answer
    
You can effect a synthetic long-exposure by stacking many shorter, sequential exposures. I don't see this as being an issue. –  Fake Name Sep 5 '13 at 5:20
    
I appreciate the advice. This is my second attempt at creating star trails. The first time I tried, I left the shutter open for about 15 minutes, but the image was very noisy and had lots of hot pixels. So this time, I tried reading some tips ahead of time, and the internet recommended stacking 30 second exposures to avoid noise and hot pixels: petapixel.com/2013/04/25/a-complete-guide-to-star-trailing, digital-photography-school.com/how-to-photograph-star-trails, –  Finer Recliner Sep 5 '13 at 13:53
    
@FakeName: Synthetic long exposure looks like what the OP posted...blurry. You are effectively chaining point lights, not stacking trails. You can, and should, still stack trails...but you actually need TRAILS in each frame to do so. You don't need to take a single five hour exposure...stacking multiple 5, 10, 15 minute exposures will do fine. –  jrista Sep 5 '13 at 15:49
1  
@Finer: A couple of things. First, noise is a simple fact of life when it comes to night sky photography. EVERY night sky shot has it, especially when viewed at 100%. There are a couple reasons why it doesn't appear that way when you view other peoples work: 1) They downsampled considerably for publication on the web 2) They stacked a few dozen shots, which (when you use a proper tool like DSS) can greatly reduce the amount of noise. Next, hot pixels are a no brainer to fix. DSS (Deep Sky Stacker) will usually take care of them for you. A small amount of color NR will too. –  jrista Sep 5 '13 at 15:51
    
Next, when you expose for longer, especially if you expose for several minutes, you can use a much lower ISO setting. A lower ISO setting will mitigate noise levels and hot pixels. Finally, your sensor is going to get hot regardless...taking 5000 30 second shots or 500 5 minute shots in continuous sequence are both going to result in a higher average sensor temperature. It will NOT cool off in the 1 second delay between frames. Generally speaking, if you are exposing for less than a minute, you aren't doing star trails photography. You are simply doing wide field night sky photography. –  jrista Sep 5 '13 at 15:51

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.