How this instrument works
People blink involuntarily and unpredictably, and a group photo only works if every single person's eyes happen to be open at the exact instant the shutter fires. This instrument models that as a probability problem: an average person blinks about 10 times a minute while posing for a photo (people naturally blink less when they know a camera is pointed at them), and each blink lasts roughly 250 milliseconds, so the chance any one person's eyes are open at a random instant is about 1 - (10/60) x 0.25, or roughly 95.8%.
For a group, every person's eyes have to be open simultaneously, and since blinking is treated as independent from person to person, the group's success probability is that single-person probability raised to the power of the number of people. A 4-person photo succeeds with probability 0.958^4, about 84%, per shot — noticeably lower than any individual's own 95.8% chance, because more people means more chances for someone to blink at the wrong moment.
Bad lighting makes this worse: slower shutter speeds and dimmer conditions are associated with a higher effective blink rate (this instrument models it as 15 blinks per minute instead of 10), lowering the per-person and per-group probability and pushing up the number of photos needed for the same confidence level. The number-of-photos figure comes from treating each shot as an independent trial and solving for how many trials are needed before the chance of at least one success reaches your target confidence.
- Enter Number of people in the photo — every additional person lowers the group's per-shot success probability.
- Toggle Bad lighting on if you're shooting in dim conditions or with a slow shutter, which raises the modeled blink rate.
- Set Desired confidence of getting a blink-free shot (%) — how sure you want to be that at least one photo in the series is blink-free.
- Read Chance the whole group's eyes are open for a single photo to see your per-shot odds.
- Read Photos needed for your target confidence for the minimum number of shots to take.
Worked example — a 4-person group photo under good lighting
One person alone under good lighting has P = 1 - (10/60) x 0.25 = 1 - 0.0417 = 0.9583, about a 95.8% chance their eyes are open in any given shot — needing just 2 photos to hit 99% confidence. Add three more people to the frame and the group's per-shot probability drops to P^4 = 0.9583^4 ≈ 0.8434, about 84%, since all four people's eyes have to be open at once.
To reach 99% confidence with that 4-person group, the formula works out to 3 photos: after 3 independent shots, the chance that none of them was blink-free drops to (1-0.8434)^3 ≈ 0.4%, comfortably under the 1% failure threshold that 99% confidence implies. A 6-person group under bad lighting targeting 95% confidence needs the same 3 shots here too, since the higher blink rate and lower per-shot probability are offset by the less demanding 95% target.
Questions
Why does adding more people to the photo lower the group's success probability so much?
Because everyone's eyes have to be open at the same instant, the group probability is each person's individual probability raised to the power of the group size — and any number less than 1 gets smaller, faster, the more times you multiply it by itself. Going from 1 person (95.8%) to 4 people (about 84%) to 10 people (under 65%) shows how quickly the odds stack against a large group, even though each individual person is still very likely to have their eyes open.
Where do the 10 blinks/minute and 250ms blink-duration numbers come from?
They're commonly cited averages for how people blink specifically while posing for a photo, where blink rate naturally drops from a resting average (often cited around 15-20 per minute in casual conversation) because people consciously try not to blink when they know a camera is about to fire. A single blink itself — the physical closing and reopening of the eyelid — typically lasts around 100-400 milliseconds, and this model uses 250ms as a representative midpoint.
Does bad lighting actually make people blink more?
It's less that dim light directly causes more blinking and more that bad lighting forces a slower shutter speed, which means the camera's effective 'exposure window' during which a blink could ruin the shot is open longer, and photographers often need more attempts in low light generally. This instrument models that combined effect as a higher effective blink rate (15/minute instead of 10) to capture the practical result: you'll typically need more shots in bad lighting to reach the same confidence level.
Is this exact, or just a useful rule-of-thumb estimate?
It's a simplified statistical model built on average blink-rate and blink-duration figures, not a measurement of any specific person's actual blink behavior — real people vary in how often and how long they blink, and factors like fatigue, allergies, or contact lenses can shift someone's rate well above or below the average this model assumes. Use the result as a practical planning guide for how many shots to budget, not a guarantee for any particular group or individual.
What confidence level should I actually aim for?
99% is a common target for an important shot (a wedding party, a formal group portrait) where retakes are expensive or impossible, while 90-95% is often plenty for a casual group photo where a slightly-closed eye is a minor, easily-forgiven flaw. Higher confidence always means more photos needed, so balance how much the shot matters against how many extra frames you're willing (or able) to take.