The blink free photos calculator shows how likely it is that nobody blinks in a group photo and how many photos you should take to be nearly certain of at least one good one. Enter the size of the group, the light and the level of confidence that you want.
How to use the blink free photos calculator
- Enter the number of people in the photo.
- Choose good or bad light.
- Enter the confidence that you want, for example 99 percent, and press Calculate.
The idea
The idea comes from Nic Svenson and Piers Barnes of the CSIRO, Australia's national science agency, who won the 2006 Ig Nobel Prize in mathematics for calculating the number of photographs you must take to nearly guarantee that nobody in a group photo has closed eyes [2]. A blink takes only a fraction of a second, but with enough people the chance that at least one person has their eyes shut in a given photo is large.
The chance for one person
The CSIRO item takes the average person having their photo taken to blink ten times a minute, with each blink lasting about 250 milliseconds [1]. Blink rates in real life vary: Wikipedia gives an average of around 17 a minute in a laboratory and only 4 to 5 a minute while reading, with blinks lasting 100 to 400 milliseconds [3]. The chance that a person's eyes are open at a given moment is then p = 1 − blinks ÷ 60 × 0.25. In good light we use 10 blinks a minute, which gives 23 ÷ 24, or 95.83%. In bad light we use 15 a minute, an assumption of our own that the sources do not state, which gives 15 ÷ 16, or 93.75%.
The chance for the group
If everyone blinks independently, the chance that nobody blinks is the single-person chance raised to the power of the number of people. The CSIRO item writes the number of photos as 1 ÷ (1 − xt) to the power of n, where n is the number of people, which is the reciprocal of that same power [1]. For 5 people in good light the chance is (23 ÷ 24) to the power of 5, which is 80.83%, our own arithmetic. The chance that at least one person blinks is 100% minus that figure, or 19.17%.
How many photos to take
Call the chance that a single photo has a blink 1 − P. The chance that every one of T photos has a blink is that number to the power of T, so the chance of at least one good photo is 1 − (1 − P) to the power of T. To reach a confidence C, we need T to be at least ln(1 − C) ÷ ln(1 − P), rounded up. This confidence form is our own working, a variation on the reciprocal rule in the CSIRO item. For 5 people and 99%, it gives 2.79, so you take 3 photos.
A bad-light example
For 5 people in bad light, this example is our own working. The chance of a blink-free photo is (15 ÷ 16) to the power of 5, which is 72.42%. The chance of a blink is 27.58%, and for 99% confidence we need ln(0.01) ÷ ln(0.2758) = 3.57, so 4 photos. A group of 5 therefore needs one more photo in poor light than in good light.
Rounding and the rule of thumb
If you round the single-person chance to 0.96, you get 0.96 to the power of 5 = 81.54% and 2.73 photos. We use the exact fraction 23 ÷ 24, which gives 80.83% and 2.79. The answer is 3 photos either way. The CSIRO item also offers a quick rule for groups of fewer than 20: divide the number of people by three in good light and by two if the light is bad [1]. For 5 people that suggests about 2 photos in good light, fewer than our 99% figure, because the rule does not name a confidence level.
Larger groups
The chance falls quickly as the group grows. For 6 people in good light it is 77.46%, and the number of photos for 90% confidence is 2. For 50 people it is about 11.9%, and for 200 it is about 0.02%, which needs about 22,900 photos for 99% confidence. These are our own results from the formula.
Choosing the confidence
The confidence is the chance that at least one of your photos has nobody blinking. It must be above 0 and below 100, because absolute certainty is impossible, as the chance of a blink never reaches zero. A confidence of 90% needs fewer photos than 99%, and 99.9% needs more. For a family photo 90% is usually enough, and for a team photo that you cannot retake you may want 99%.
Assumptions
The model assumes that every person blinks at the same average rate and that blinks are independent. In reality, people blink in response to a flash, a joke or a countdown, and some blink more often than others. Dry eyes, bright sun and tiredness also change the rate. Wikipedia notes that blink rates differ between people and with the task [3], so treat the result as a guide and not as a guarantee.
Tips for the photographer
Take photos in a quick burst, and ask people to relax their eyes and then open them just before the shot. Good light helps, because our bad-light setting assumes more blinks. The calculator tells you how many frames to expect to need, but it cannot make the group stand still.
Reading the results
The first line gives the chance that one photo has nobody blinking, and the chance that at least one person blinks. The second line gives the number of photos to take for the confidence that you chose, and the chance that this number actually gives. That chance is always at least the confidence you asked for, since the number of photos is rounded up. The third line shows the working for a single person, so that you can check it by hand.
Common mistakes
- Using the number of photos as a promise. It gives a chance, not a guarantee.
- Entering a confidence of 100. Certainty is impossible, so use a value below 100.
- Forgetting that the light setting matters. Bad light assumes more blinks.
- Assuming that a blink-free photo is also a good one. Faces may still be turned away.
What next
Use this often? Add Blink Free P to your Home screen.
Frequently asked questions
What is a blink free photos calculator?
It is a tool that estimates the chance that nobody blinks in a group photo and how many photos to take to get at least one blink-free shot.
How is the chance of a blink-free photo worked out?
Each person's chance of open eyes is 1 minus the blinks per minute divided by 60 times 0.25, and the group chance is that value to the power of the number of people.
How many photos should I take for a group of 5?
In good light 3 photos give at least 99% confidence of one blink-free shot, and in bad light 4 photos.
Why can the confidence not be 100 percent?
However many photos you take, there is always some small chance that every one has a blink, so certainty is impossible.
Does the light matter?
Yes. The tool assumes about 10 blinks a minute in good light and 15 in bad light, which lowers the chance of a blink-free photo. The 15 is our own assumption.
Sources
Was this tool helpful?
What went wrong?
We only receive the reason you pick, never what you typed into the tool.
Thanks for your feedback.