Star Trails

[Josh Brock]

I have been trying to shoot star trails lately and for some reason the shots arecoming out really noisy. I Have been shooting with a low ISO and a slow shutter speed.

I'm useing a ziess 50 F 1.2 on a cannon t2I.

Can some one share their settings for star trails or night imagery that they have taken?

Maybe my problem is in my meager editing skills?

Any advice would be appreciated.

Also I would love to see any star trails shot by any of you.

[Cayenne Redmonk]

For stars trails, maybe open the aperture to whatever your camera has between 4-6, use a shutter release or a timer to start the shot. Try 24,48,and 96 minute exposures for trails, ISO50-100, maybe 2-8 minute exposures for points to check the framing before committing to a long shot.

Cold kills batteries. I like film for star trails because I don't have to run data aggregation for hours while worrying about the battery going dead and losing the shot.

–G.B.–

[Josh Brock]

I have been using ISO 100 and a wired external trigger I my exposure times have varried from 15 – 45 mins.

and the longer the exposure the more noisy the image gets even at 100 ISO. Should I be shooting wide open? OR should I close up my apreture? Its so hard to expirement with this kind of photography just cause each image takes so long to capture.

[Eric Lundquist]

You may be seeing "hot pixels" instead of noise. If so, replicate the same settings you used on your star trails shot (ISO, aperture, shutter speed) and take a photo with the lens cap on. You can remove these hot spots in post processing as the the photo you took with the lens cap will be entirely black except where the sensor has the hot spots. Mask them out and you should have a less "noisy" image.

[Adam Criswell]

Try turning on Long Exposure Noise Reduction, this will automatically do what Eric suggested. It should dramatically reduce the number of hot pixels, although you may still get some.

[Josh Brock]

Ill try the hot spot reduction techniques. Hopefully that will do the trick.

[Eric Lundquist]

Some cameras have an anti-shake feature you that you can disable when taking long exposures on a stable base (ie. tripod)

[steven franchuk]

what you are doing works great with film cameras. For film I have taken pictures a couple of hours long and seen no noise. for digital the results are not as good due to hot pixels and sensor noise. The amount of noise present is dependent on temperature, iso setting (slower is better) and exposure time (again shorter is better). Using your cameras noise reduction feature will help but it won't fully resolve the issue.

To get the best results with digital use a series of short exposures at iso 100 at about 30seconds of exposure and then use software to stitch them together. auto noise reduction cannot be used in this methode. However if one dark frame is taken first that one dark frame can be manually subtracted from all the others in the series before stitching. The titching process can result in small gaps between star trails if the star moves enough between frames. Some software out there can automatically fill in the gaps. The other way to minimize the gap is to use wide angle lenses. The star won't appear to move as fast with a wide angle lens.

Note the noise reduction feature in your camera will cause it to take two pictures. One with the shutter closed (the noise frame) and one with the shutter open. After the second picture is taken the camera will subtract the noise from the shutter open image and only save the final result. this will double the exposure time and your battery may not last long enough to get the exposure your want.

[Eric Lundquist]

Good info Steven! As a followup, I recently found this thread over at Desk to Dirtbag after seeing one of their photos on the MLD facebook page. There's enough detail in the posts on how to achieve similar results with an intervelometer to take consecutive shots.

[steven franchuk]

For more information on noise, cameras, and photo software you might want to check out photo.net

[Corbin McFarlane]

I don't use a Canon so this may vary slightly, but here are some things I have picked up with my Olympus:
1. The best ISO is usually higher, for example on my E-M5 ISO ISO 800 is probably the best way to get noise-free star trails (maybe ISO 1600). DXOmark publishes this sort of data, though I learned by scanning a forum.
2. Use a shorter shutter speed, 1 minute works great though some cameras are limited to 30s. I set my camera to continuous shooting mode with 60s shutter, and use rubber bands to hold the shutter button down. Way cheaper, simpler and lighter than intervalometers, etc. There are also smart phone apps that can do this.
3. Stitch 60+ minutes of exposures in startrails.exe or other free star trail stitching software. This basically just keeps the brightest level recorded for each pixel.
4. There will be less noise in colder temperatures. Ideally it would be well below freezing; hot summer nights as in southern Arizona would create incredible noise. I saw a relationship for this somewhere, it was a 1 stop decrease in noise for every 15 or 30 degree temperature drop, I can't remember the exact number.

[steven franchuk]

"The best ISO is usually higher"

this is wrong! For minimum noise the always use the lowest iso setting. Low iso (reduced light sensitivity) and gives you the lowest noise to but requires longer exposure times to get the proper exposure. Hi iso will give you Short exposure times but a lot more noise.

You can see the noise difference in test photo's in DPreview.com of your camera (page 16):
http://www.dpreview.com/reviews/olympusem5

"There will be less noise in colder temperatures."

for all practical purposes you will see no significant difference until you get down to temperatures below -50F. At normal temperatures ranges you will see no significant difference.

Note: I have extensive photography experience and professional photo equipment.

[Corbin McFarlane]

"The best ISO is usually higher"

>this is wrong! For minimum noise the always use the lowest iso setting. Low iso (reduced light sensitivity) and gives you the lowest noise to but requires longer exposure times to get the proper exposure. Hi iso will give you Short exposure times but a lot more noise.

This is sort of correct, in that you should always use the lowest ISO for the desired aperture and shutter speed. If you later brighten the picture using software, then you would have been better off noise-wise using a higher ISO (provided greater shutter speed or aperature weren't possible). That is only true up to an ISO which varies by camera at which point brightening in software is preferable (ISO 1600 for mine).

Addtionally, longer exposures become noisier on a digital camera. Eventually there is a tipping point between using a higher ISO and using a shorter exposure length. I tend to assume 60s for my present camera, but that is mostly because that is the start of Bulb mode. I have tried 4, 8, and 16-minute exposures at ISO200 and they are clearly too noisy, but I don't know exactly where the tipping point is. I expect the tipping point will also vary with outside temparature (see below). In summary, minimizing noise with digital cameras is a trade off between minimizing ISO and minimizing exposure length, and is camera-specific. I would guess for a Canon T2i the 60s rule would also be a good way to determine when to raise the ISO as far as 800 or 1600, as with my E-M5.

When using exposure stacking software like startrails.exe, random noise will cancel out in favor of the brighter pixels, but hot pixels will accrue. Hot pixels are a product of longer exposures and higher temperature, thus shorter exposures and colder temperatures accrue less of them. It isn't possible to do black frame subtraction after every shot because the star trails become a dashed line; therefore, short(ish) shutter speeds, high(er) ISO's, and stacking are the way to get the least noise in a star trail image. The exact point for best image quality varies by camera.

"There will be less noise in colder temperatures."

>for all practical purposes you will see no significant difference until you get down to temperatures below -50F. At normal temperatures ranges you will see no significant difference.

This is wrong! There is a clear difference between shooting at 0 deg F and 60 deg F. Colder temperatures are especially beneficial for longer exposures. A cold temperature goes a long way toward reducing the noise-generating heat created by the sensor during the exposure. Have you tried? Alternatively, the internet is a good source of information.

If your pro gear is relevant to this discussion, the only possible way is because your camera suffers less read noise and hot pixels than the cheapo models. This makes your advice to the OP worse, rather than better.

[Author]

Posts