How this calculator works
A person blinks roughly fifteen to twenty times a minute, and each blink keeps the eyes closed for about a tenth to a fourth of a second. That means a person is mid-blink for a few percent of the time, more in harsh or dim light when the eyes narrow or water. Treating each person as independent, the chance that everyone has open eyes is the chance for one person raised to the power of the group size. Because that number falls exponentially, groups get much harder quickly: ten people at 95 percent each give only a 59.9 percent chance of a clean frame, so you need about two frames for one clean shot on average and four for a 95 percent chance.
Formula
chance one person is open = 1 - blink share. chance everyone is open = (1 - blink share)^people. photos for one clean frame on average = 1 / chance. photos for a 95 percent chance = ceil(ln(0.05) / ln(1 - chance)), because each frame must fail with probability 1 - chance.
Worked example
Ten people in good light each have a 5 percent chance of blinking, so the chance of a clean frame is 0.95 to the tenth power = 0.5987, or 59.87 percent. On average 1 / 0.5987 = 1.67 frames give one clean shot, and a 95 percent chance needs ceil(ln 0.05 / ln 0.4013) = 4 frames. In bad light the single-person figure falls to 0.90, so ten people give 0.3487 and need about 2.87 frames on average and 7 frames for 95 percent. A group of twenty in bad light has only a 12.16 percent chance per frame.
Advertisement
Frequently asked questions
Where do the 5 and 10 percent figures come from?
They are a model, not a measurement of your group. A person blinking about every four seconds for a quarter of a second is mid-blink about 6 percent of the time, and good light allows longer open-eye periods. Harsh light, wind and tired eyes raise the figure, so the custom option lets you use your own estimate.
How do I improve the odds?
Take a burst and choose the frame where everyone is open, or ask the group to close their eyes, count to three and open them on three. A moment of surprise or a count that ends at an unexpected number also breaks the blink reflex.
Does the chart assume people blink independently?
Yes, and real groups do not. People often blink in response to the same flash or sound, so the true chance of an all-open frame is a little lower than independent blinks predict. The estimate is still useful for planning a burst.
Why does the chance fall so fast as I add people?
Each extra person multiplies the chance by the single-person figure again, so the exponent grows with the group size. That is why four people are far more than twice as hard as two.