Night photography: Long exposure noise reduction on a warm desert night
Technical details
The image was lit by moonlight, with a small amount of distant sodium vapor streetlight visible on the car windows. Two exposures of 7.5 minutes at f/9.5 ISO 200 were layered together for a 15 minute cumulative exposure. This exposure is the same as 5 minutes at f/8 ISO 200, or 10 minutes at f/8 ISO 100 -- a very typical full moon exposure when you're away from the lights of the city. I stopped down to f/9.5 for depth of field, and because 15 minute star trails often hit a sweet spot for me.
Why take 2 shots?
One had light painting, the other one did not. This keeps the options open for either a moonlight only shot, or a light painted shot. By keeping the interval between shots to 1 second or less, the 2 images can quickly be combined for longer star trails. The image above includes the sky portion from both 7.5 minute exposures with no light painting (i.e., moonlight only).
Test exposures, and taking control of your star trails
Before making the Green 1971 Dodge Demon image, I made a test exposure of 10 seconds at f/8 ISO 6400. The test exposure allowed me to evaluate the histogram, composition, and focusing before committing to the shot. Test exposures also allow you to creatively position the star trails in your images. When you're facing West, the stars will go down. To the East, they'll go up. To the South they're almost horizontal. Point North, and they spiral around Polaris (the North Star). In this case, the camera position was facing to the southwest. By zooming in on my test exposure on the LCD, I was able to pre-visualize the interaction of the star trails with the trunk of the car.
When it's hot, turn on in-camera noise reduction
Shooting at night in the desert when it's 75 degrees with no wind is a real treat. Oftentimes sky noise and hot pixels are not readily apparent on your camera's LCD screen. Testing different exposure lengths in warm weather is necessary to find out when to use in-camera noise reduction with your digital SLR.
When it's cold out, current cameras such as the Canon 5D Mark II do not require noise reduction on night exposures in the 6-8 minute range. Typically I do not run in-camera noise reduction when the ambient temperature is 60 degrees or colder. In this case, the ambient temperature was 70-75 degrees -- warm temperatures can drastically increases noise in long exposures. I set the Canon TC80-N3 Timer Remote to capture 2 exposures of 7.5 minutes with a 1 second interval between shots. The 5d Mark II was set to run in-camera noise reduction. When shooting multiple consecutive images with a timer remote, the 5D Mark II, 5D, and 7D will wait until the entire sequence of shots is complete to run noise reduction.
As an experiment, I shot a few 7.5 minute exposures without noise reduction under the same conditions. These images had a large amount of green hot pixels in the sky. When it's warm out at night, either run noise reduction, or shoot shorter exposures. Interestingly, the workshop participants using the Nikon D700 seemed to have consistently cleaner skies under the same exposure and temperature conditions without using in-camera noise reduction. I didn't do any formal testing at the workshop, but the D700 seemed less susceptible to temperature induced long exposure noise for night photography. Perhaps there's a D700 owner and desert-hound out there that can confirm long exposure noise performance under warm temperatures?