Help with removing this noise from star trail?

After a week-long vacation in the Colorado San Juans, I was excited to have some success with star trails. At my home on the east coast, it's terribly difficult to find somewhere suitable enough to do such a shot, so I was excited at the opportunity. Using a 5D + TSE 24mm II, I shot this at 5.6, 100iso, for about 115 minutes. This is one example of several. I DID NOT use the in-camera dark-frame noise reduction, and I'm hoping this failure can be recoverable.

I'm new at this, but I'm hopeful that there are software techniques for recovering this shot in post. Can anyone offer any suggestions in CS5 to help me recover my shots?

Byte size: ZERO | Content warning: NOT AN IMAGE

Byte size: ZERO | Content warning: NOT AN IMAGE

You can use Noiseware or something similar (Noise Ninja, etc.) to reduce the amount of noise, but the best thing to do is use a low ISO, and keep the camera cooled. With long exposures, the camera gets warm, and the warmer it gets the noisier it gets. The colored noise you have (reds, blues) are a result of heat in the camera. There's plenty of ways to keep the camera cool, one of which is a simple fan blowing on it. You might introduce a bit of vibration, but on a sturdy tripod you should be okay. Other options get a little bit more involved, like taking a styrofoam cooler and modifying it to fit around the camera and installing small fans.

You don't want to use in-camera noise reduction, it will add time to the exposure, and add more heat. In-camera noise reduction works at the same duration as your exposure, so if you have a 1 minute exposure, it takes a minute for the noise reduction to run. You've now extended your camera on-time to two minutes instead of one, which dds more heat, and more noise.


Run over to Cloudy Night astro forum, and ask there. There's a ton of great people there and a ton of useful info.

I've never seen noise that bad, and IMO running it through a noise reduction software will smear it so bad as to be unrecognizable.

No it won't. You might lose a small amount of fine detail, but with star trails and the framing of your images you won't see much loss. Noise like that is very common in long-exposure astro imaging, it just takes the right settings, the right practice, and a cool camera to keep it at a minimum. You can easily remove most of the noise in these two images, and at this size you won't hurt the image.

However, for printing and larger online images it's best to practice, use the right settings, and keep the camera cooled.


You don't have "image editing ok", but I took the liberty of running your images through Noiseware, to show you what results you'd get from your two images. You lose a tiny amount of finer detail, but overall the images remain as shot...

HTTP response: 404 | Byte size: ZERO | PHOTOBUCKET ERROR IMAGE

HTTP response: 404 | Byte size: ZERO | PHOTOBUCKET ERROR IMAGE

Katodog, THANK YOU. Sincerely, it's a great suggestion. I've downloaded Noiseware (was never impressed with Noise Ninja in the past) and I'm working on making these shots use-able.

Yes, there is no trouble at all modifying my files. In fact, it is much appreciated. I'll work on fixing the shot and then I'll post up something when I'm done.

Thanks again!

I'm actually having trouble duplicating your noise removal with Noiseware. Would you mind sharing what settings you are using to achieve this effect?

Sure, open up Noiseware and choose "Default" in the "Settings" menu. Now here comes the hard part...you have to click "OK".

HTTP response: 404 | Byte size: ZERO | PHOTOBUCKET ERROR IMAGE

HTTP response: 404 | MIME changed to 'image/gif' | Byte size: ZERO | PHOTOBUCKET ERROR IMAGE

You can still do a dark frame after the fact. Ideally it would be at the same temperature as when the exposure was taken, but the difference between one with the temperature being the same and not, will be perhaps 10% of the difference between a dark frame and no dark frame. Do the subtraction in something astronomical so that it's done with linear data (eg. startrails, DSS, etc) and then apply your noise filtering.

Steve.

Weird. I am not getting the same results at all. I'm using Noiseware Professional 4.2 in CS5. Whether I edit the small jpg's I posted, or the original RAW, the results are the same.

I'll have to play with the settings later - but I'm failing to even come close to the results you posted.

That's kinda weird, should be fairly close to what I did.