When shooting long exposures, one of the biggest issues you can run into is that of noise, especially when you’re shooting at night, as there’s far more black in the image. Similar to when you crank up the ISO on a camera, the longer an exposure is, the more likely you are to pick up unwanted noise artifacts.
To help consumers figure out what camera system might be best for their work, landscape photographer Brendan Davey has created an impressive database that compares how much noise cameras produce at various exposures.
Davey is known for his nighttime landscapes and aurora photography, so it’s no surprise he’s well-versed in low-light conditions.
The database he’s created is made up of five different images captured by every one of the cameras, each of which is tested under strict conditions to ensure the most accurate comparison from one camera to the next. Specifically, Davey says each camera is shot at four different exposure times at ISO 3200 inside a dark room around 21 degrees centigrade (roughly 70 degrees Fahrenheit). All built-in noise reduction is turned off and each image is captured as a RAW file in Manual or Bulb mode.
Once the image is captured, Davey opens each one in Adobe Camera Raw, where all settings are put to ‘0’ with the exception of the white balance, which is set at 4000K. Considering most cameras now look incredibly clean at ISO 3200, Davey shows off the noise structure of each file by bumping up the exposures +2.5 and +5 stops — the former of which is meant as a control, while the latter is used for the various exposure times.
Currently, there are 41 cameras in the database. The list includes Nikon, Canon, Fujifilm, Olympus, Sony, and Pentax cameras from the past few years. Below is a small comparison chart showing off cameras from three of the included manufacturers:
We’ll leave it up to you to decide which of the above cameras produced the best results.
It’s worth noting that you shouldn’t take the visuals at face value. As Davey says, there are a handful of things to keep in mind that affect the accuracy and relevancy of the results:
- The age of the camera.
- If the camera supports fully disabling NR (some do not, and in some models that can affect the RAW file).
- Additional processing applied to the image regardless of the NR settings.
- The size and pixel density of the sensor.
- The accuracy of the ISO — in some cases this can be almost a full stop out either side of 3,200. (1,600–6,400 ISO).
- Manufacturing variation (tolerances).
To check out the database in its entirety, head on over to Davey’s Sensor Noise DB page. Although he’s done this for free and has shared the results with the world, he does have a donation option if you are in a position to contribute.
