Sensor sizes

Learning goal

You will be able to recognize the different sensor sizes by name and compare them roughly with one another: full frame, APS-C, Micro Four Thirds and phone sensors. You will be able to explain, through the amount of light a large sensor gathers, why two cameras with the same number of pixels give different photos; you will know how this difference shows up especially in noise and background blur in low light, and you will notice that sensor names given in "inches" are not real measurements.

The question

Both boxes carry the same number: twenty-four megapixels. One is a compact that fits in your pocket, the other a mirrorless camera with a big body. The salesperson in the shop says that both are "twenty-four megapixels", that is, that they take photos at the same resolution. But when you photograph the same face with both in a restaurant one evening, the photos don't look the same at all: one is grainy and flat, the other clean, with a soft background.

Same megapixels, so why a different photo? The answer is not in the number printed on the box but in the size of the sensor. Megapixels tell you how many points the image is made of; sensor size tells you how large a surface those points are spread across, and therefore how much light it gathers. In this lesson we open up the difference between the two numbers.

Context

Remember the bucket comparison in Lesson 1.4: a wide-mouthed bucket fills with rain faster, and a wide surface gathers light more abundantly. This lesson ties that comparison to concrete measurements. When you buy a camera, when you try to understand why a photo came out grainy, or when you look for a way to melt the background in portraits, sensor size is always quietly at work.

The problem is that camera advertising rarely puts sensor size up front; what it puts up front is megapixels, because a big number is impressive. Phone sensors, meanwhile, are both very small and confusingly named. Someone who can read sensor size sees behind this advertising language and sets realistic expectations.

Simple explanation

What are the sensor sizes?

The easiest way to understand sensor sizes is to pick a reference point. This reference comes from the most common format of the film era: 36 millimeters by 24 millimeters. Digital sensors of this size are called full frame. Sensors smaller than full frame are referred to by how much smaller they are than this reference. That ratio is called the crop factor.

In APS-C sensors, a common size, the crop factor is about 1.5 or 1.6, depending on the manufacturer. This means the sensor's sides are about one and a half times shorter than those of full frame. Its surface area, however, is less than half that of full frame, because area changes with the square of the sides: sides 1.5 times shorter create a difference in area of a little over two times. In Micro Four Thirds sensors the crop factor is 2; the sides are half as long and the area is a quarter. The sensors in phones' main cameras are smaller still; most are about the size of a fingernail.

Phone and compact sensors are often referred to by names such as "1 inch" or "1/1.3 inch". These names are misleading. When digital sensors became widespread they took the place of the image tubes in old video cameras, and the naming was inherited from those tubes. The diagonal of a "1 inch" type sensor is actually about 16 millimeters, quite a bit shorter than an inch, which is 25.4 millimeters.

Why does a large sensor give a cleaner image?

When you shoot the same scene with the same settings, roughly the same amount of light falls on every square millimeter of the sensor. Because a large sensor has more square millimeters, it gathers far more light in total. When you print the photo at the same size or view it on the same screen, the image from the large sensor carries that extra light.

The noise that appears in an image as the light falls, that blotchy, grainy texture, comes largely from too little light having been gathered. Think of the bucket comparison: if rain is only drizzling, the water collecting in a small bucket is clearly affected by the roughness at its bottom; in a big bucket the same roughness isn't even noticed. Because a large sensor gathers more light in low light, it pushes noise back.

This difference doesn't show up in the same place on every skin tone. Noise appears first in the dark areas of the image: a face in shadow, dark hair, darker skin. Phones with small sensors smooth the image to hide this noise; especially on darker skin, this smoothing can wipe out the texture of the skin and turn the face into a plastic or flat surface. That is why, when you photograph different skin tones in low light, it matters to check the result closely.

How does sensor size affect the background?

That look in portraits where the background melts softly away while the subject stays sharp comes from a shallow depth of field. Depth of field is how deep, from front to back, the zone that looks sharp in a photograph is. You will see this topic in detail in the exposure module; for now it is enough to know how it relates to sensor size.

Imagine wanting to build the same frame with a small and a large sensor. Because the small sensor sees a narrower area, the same frame needs a shorter lens or a position further away. Under these conditions, at similar aperture settings, a photo taken with the large sensor has less depth of field, and the background blurs more easily. That is why phones have a "portrait mode": because the small sensor can't produce this blur naturally, the phone blurs the background with software.

Visual evidence

Imagine two portraits of the same person in a dim restaurant one evening, taken from the same angle with the same framing. One was taken with a twenty-four-megapixel compact camera. The face looks bright enough, but when you look closely, the cheeks and forehead are covered with a fine, colored grain. The neck and hair in shadow have flattened into obvious blotching. The tables, lamps and paintings on the wall behind are almost as sharp as the subject; the eye doesn't stop on the face but scatters into the background too.

The other was also taken at twenty-four megapixels, but with a full-frame camera. The skin texture on the face, the eyelashes and the small reflection of light in the eye are clean and sharp. Detail can still be made out in the neck in shadow. The tables and lamps behind have turned into soft, warm blobs; the subject separates from the background on its own. The two cameras have the same number of pixels, but the full-frame sensor is many times larger than the compact's sensor. The difference lies in the light the sensor gathers and in the depth of field.

Common mistakes

Summary

Two cameras with the same megapixels give different photos, because megapixels decide the number of points while sensor size decides how much light those points gather. A large sensor gathers more light in total in the same light; this means less noise, especially in low light, and less depth of field, which makes it easier to melt the background in portraits.

Mini assignment

In this assignment you get to know your own tool's sensor.

1. Find the technical specifications of your camera or phone, and note its sensor size in millimeters or by type name.
2. In a dimly lit room, photograph a face or an object without flash.
3. Zoom in on the photo on screen and find where noise, or a smoothed, flat texture, begins in the areas in shadow.
4. Photograph the same object by getting closer so that the background stays far away, and compare how much the background blurs.

Check yourself

• Did you find your tool's sensor size, and could you roughly say where it stands relative to full frame?
• Did you see in your own photo that noise starts first in the dark areas of the image?
• Could you say what, apart from the sensor, you can change to increase background blur?

Glossary

Mini quiz

1. Why can two cameras with the same megapixels give different photos?

• A) Their megapixel counts are actually always different
• B) Their sensors differ in size, and so in the amount of light they gather
• C) One of them shoots in black and white
• D) Their screens are different sizes

2. What is the area of a sensor with a crop factor of 2, compared with full frame?

• A) Half
• B) The same
• C) A quarter
• D) Twice as much

3. About how long is the diagonal of a "1 inch" type sensor really?

• A) 25.4 millimeters
• B) 36 millimeters
• C) 1 centimeter
• D) 16 millimeters

4. What problem is most common on a darker-skinned face photographed in low light with a small-sensor phone?

• A) The face coming out completely white
• B) Noise, or the smoothing used to hide it wiping out the skin texture
• C) The face coming out upside down
• D) The background being blurred too much

Answers

1. B — Megapixels tell you the number of points; sensor size decides how large a surface those points are spread across and how much light is gathered in total.
2. C — The crop factor is a ratio of the sides; when the sides are halved, the area shrinks with the square of the sides and becomes a quarter of full frame.
3. D — The name "inch" is left over from old video tubes; the diagonal of a 1 inch type sensor is actually about 16 millimeters.
4. B — Noise appears first in dark areas; phones smooth the image to hide it, and this can wipe out the skin texture on darker skin.

Summary card

• Megapixels = the number of points; sensor size = the light gathered.
• Full frame is 36 × 24 mm; APS-C has a crop factor of about 1.5–1.6, Micro Four Thirds 2.
• Area shrinks with the square of the ratio of the sides.
• "1 inch" is not a real measurement; the diagonal is about 16 mm.
• A large sensor: less noise in low light, an easier blurred background.

Bridge to the next lesson

You have seen that sensor size tells you more than megapixels. So what do megapixels change, and what don't they? At the next stop the question is put like this: why can't a 48-megapixel phone beat a 24-megapixel camera? There you will see that megapixels are not a measure of quality but a count of detail, and in which situations that count really helps.