What is the lowest light level that a camera can meter? This is vitally important to those who film day-to-night timelapse or build low-light-capable webcams.
The manufacturer specifications for metering ranges are fanciful at best, so with the assistance of a homemade integrating sphere, we are systematically testing as many cameras as we can obtain. The results are summarized in the following chart. You can learn more about the project and methodology by clicking the links below.
Camera Metering Reliability Chart
|Camera||Metering limit def. 1 [EV]
Metering failure occurs at the point where the camera compensates 2/3 stop for a 1 stop change in ambient brightness.
|Metering limit def. 2 [EV]
Metering failure occurs when the image is 1 stop darker than the image taken at an ambient light level of EV0.
|Manufacturer claimed value||Moon phase equivalent to def. 1 limit||Camera stops down lens
|If DSLR: better in live view?||If DSLR: benefits from <f/2.8 lenses?||Exp. comp to put histogram in middle:|
|Sony A7 III||-6.4||-8.1||EV-3 @ F2||18%||Yes, and “live view shooting: setting effect off” has no impact||—||—||0|
|Sony A7R III||-6.0||-7||EV-3 @ f2||28%||Yes, and “live view shooting: setting effect off” has no impact||—||—||0|
|Sony A7R II||-4.4||-5.6||EV-3 @ f2||65%||Yes, and “live view shooting: setting effect off” has no impact||—||—||0|
|Canon 1D mk IV||-4.3||-5.7||EV0||67%||No||Yes||Yes||2/3|
|Canon M5||-2.3||-3.8||EV1 @ ISO100||99%||No||—||—||1 1/6|
|Canon 5DsR||-0.9||-2.1||EV1 @ 50mm f1.4 ISO100||—||No||Yes||Yes||1|
|Nikon D800E||>0*||Approx 0||EV0 @ f1.4 ISO100 in matrix metering||—||No||Yes||No||1|
Metering tests were conducted with a Rokinon 50mm f/1.4 lens at f/2.8 for all cameras except Nikon DSLRs, which used a Nikon 105mm f/2.8 Micro at f/2.8. More details can be found by clicking the individual cameras.
*Outside test range.
1. Metering failure occurs at the point where the camera compensates 2/3 stop for a 1 stop change in ambient brightness.
2. Metering failure occurs when the image is 1 stop darker than the image taken at an ambient light level of EV0.
How These Calculations Were Made: An Integrating Sphere
Sean Goebel conceived and constructed this project, and I (Matthew Saville) will soon start testing new cameras as I review them. Sean built the integrating sphere, which is a sphere that is painted white on the inside and has LEDs for internal illumination. The camera lens pokes into the sphere. Carefully designed baffles cause the inside of the sphere to be nearly perfectly evenly illuminated.
The internal illumination level is adjusted to step from EV0 to EV-10 in 1-stop increments, and a photo is taken in Av mode at each light level. F/2.8 lenses (almost always a Rokinon 50mm f/1.4) are used. The resulting image brightness is plotted against the corresponding ambient light level, and an example plot is show below:
More of these plots can be viewed by clicking the cameras in the table above. Since there isn’t an obvious single point where metering fails, we defined two thresholds for metering failure as described above. For more in-depth information about how this information is produced, and how to implement it in the real world for better results as a photographer, see the links below.
Additional Reading | Integrating Sphere & Camera Metering Test Project
Main Project Page – Test Results
(YOU ARE HERE)