Sensors in digital photography

I‘s been some time since I wanted to develop an article on Sensors in digital photography. This little rectangle of silicon that has replaced our good old rolls of silver film rolls in our cameras has the same function (recording an image), but nevertheless has special characteristics.

FILM FORMATS IN PHOTOGRAPHY

We can roughly classify the formats of film cameras into three main families :

• The micro formats (format 110, format 126, Disc)
• The small formats (film 135)
• Medium and large formats (film 120 et +)

The choice of formats in the era of film photography

Today for a photographer who wants to equip himself with a digital camera, the choice of format arises: full format 24 x 36 mm, or APS-C 16 x 24 mm. However, this reflection on the choice of format is not specific to digital. In the time of film, this problem of the choice of format was already posed.

From the two most common types of film (the 120 format and the 135 format, but there were many others, see the table below), it was possible to obtain images of surfaces and ratios different.

The films 120 were used on so-called “medium and large format” cameras. Depending on the device, you could obtain images in the format:

• 6 x 6 cm (square format used by Hasselblad and Rolleiflex),
• 4,5 x 6 cm (used by the Mamiya 645, and the Pentax 645)
• 6 x 7 cm (used by the Pentax 67, and the Mamiya RB67 Pro),
• 6 x 9 cm, 6 x 12 cm and even 6 x 17 cm (panoramic format such as Tamiyama).

135 films were used in so-called “small format” devices. Depending on the device, we could thus obtain an image of different formats:

• squares of 24 x 24 mm,
• horizontal rectangles of 24 x 32 mm (ratio 4:3),
• 24 x 34 mm (used on the very first Nikon cameras),
• 24 x 65 mm (used on the Hasselblad XPan) panoramic format.
• and even 18 x 24 mm vertical “half-frames”.
• 24 x 36 mm popularized by Leica. Oskar Barnack the inventor of the Leica was at the origin of this 24×36 mm format.

The reference in 24 x 36 mm format

This 24 x 36 mm format (either a ratio of 3/2, that’s to say the photo is 1.5 times longer than it is wide : 24 mm x 1.5 = 36 mm), originates from the film of 35 mm dimension used by the cinema. How the 3:2 aspect ratio of 35mm film lasted nearly 100 years as the dominant aspect ratio in photography. The 3:2 aspect ratio obviously has the closest proportions to the golden rectangle than any other aspect ratio in photography. It gives it the potential to more aesthetically present the composition of a photograph.

The golden rectangle

This 24 x 36 mm format (135 film) finally became so dominant at the time of film, that during the transition to digital this format remained a reference for the size of digital sensors under the name “full format”.

However, during the development of digital, the first SLRs were equipped with an APS-C sensor (small sensor) and this for a question of cost. At Nikon, one of the first manufacturers to equip its cameras with a “full format” sensor, the first model was the Nikon D3.

English speakers call it “ Full Format ”, translated as “ Plein Format ”, and should not be confused with the acronym for “ Full Frame ”. In the latter case, a “Full Frame” photo sensor is a sensor in which 100% of the available surface is dedicated to capturing light, which is incompatible with CMOS architecture. Thus, a 24 x 36 mm CMOS sensor is Full Format, but not Full Frame, a loss of resolution is due to the presence of the amplifier on the photosite. However, in practice the term full-frame is often used to designate a 24 x 36 mm sensor (full frame).

Example: The Nikon Z9 sensor has 52.37 million pixels including 45.7 million effective pixels. A part of the pixels (52.37 – 45.70) is dedicated to the system. The photos have a dimension of 8256 × 5504 pixels or 45,441,024 pixels

Table concerning the main analog film formats

Cartridge film 135

Film 120

Film reel 120

GENERAL INFORMATIONS ON SENSORS IN DIGITAL PHOTOGRAPHY

With the advent of digital photography, the film was replaced by a digital sensor. Digital sensors are, like the silver halide of silver films, photosensitive elements, that’s to say capable of reacting to light.

A brief history of digital sensors

Like silver films made up of light-sensitive silver crystals, digital sensors are made up of photosites, that’s to say small photoelectric cells that capture light for each pixel (picture element) which will constitute the digital file of your image.

Moreover, there are several sensor technologies :

• The CCD sensors (Charge Coupled Device), the first CCD sensor appeared in 1969.

• The CMOS sensors (Complementary Metal Oxide Semiconductors) appeared in the early 1990s. Due to technological progress, they replaced CCD sensors in many areas. Without wanting to go into complex technical aspects, CMOS has major advantages:
  • reduced power consumption
  • lower cost
  • higher reading speed.

Furthermore we often speak of a sensor of X thousands of pixels to talk about the definition of a sensor. This is a terminology error, we should rather use the term photosites which are the elements constituting a sensor to talk about its definition. These are the photosites that will capture the light, and produce the pixels (picture element) of your digital file after processing by your device’s processor.

Size and definition of digital sensors

The definition of a sensor and the size of a sensor are two different and independent elements not to be confused.
Example :

• the Nikon D4 had a 24 x 36 mm full-frame sensor and a definition of 16 million pixels (photosites)
• the Nikon D500 had a 16 x 24 mm APS-C sensor and a definition of 20 million pixels (photosites)

In this example, the largest sensor is the least defined. Here we come to the problem of the size of the photosites. At the same sensor size, if the number of photosites is important, the smaller their size.

Impact on the picture of the size and definition of the sensor :

• The definition of the sensor (the number of its photosites) influences several elements :
  • on the size of your images. A very defined sensor will therefore allow you to make larger and higher quality prints (enlargement) of your photographs.
  • a very defined sensor will generate more digital noise in high sensitivities. In fact, to increase the density of the photosites on the same sensor surface, they must be smaller. However, the smaller the photosites, the less sensitive they are to capture light.
• The size of the sensor, for its part, influences the rendering of the image, in particular the notion of depth of field.

Digital sensor of Nikon Z9

THE DIFFERENT SIZES OF SENSORS IN DIGITAL PHOTOGRAPHY

The full frame sensor (full-format)

It measures 24 x 36 mm (3:2 aspect ratio) to be directly equivalent to film 135 of the same size. It is rather found on high-end devices.
 

The APS-C sensor

Sensor smaller than full frame (3:2 aspect ratio), in proportions ranging from 1.5 to 1.6 times smaller diagonal.
15,7 x 23,6 mm – Nikon Sony Fuji, coefficient 1.5 compared to 24 x 36 mm
14,8 x 22,2 mm – Canon, coefficient 1.6 compared to 24 x 36 mm
This sensor format is very common on entry-level and mid-range devices.
 

The APS-H sensor

19 x 28.7 mm sensor (3: 2 aspect ratio), whose diagonal is 1.25 times (rounded up to 1.3) smaller than the full frame, specific to some Canon SLRs. To my knowledge, this format is no longer used today.
 

The Micro 4/3 sensor

This sensor format has been dubbed the “Micro Four Thirds System”. Sensor 13 x 17.3 mm – coef. 2 compared to 24 x 36 mm. The aspect ratio of the image changes from 3/2 to 4/3, and gives the impression of a squarer image.
Olympus and Panasonic jointly created the Micro Four Thirds system or Micro four thirds, a photographic system based on hybrid cameras, and announced on August 5, 2008. In 2014 Kodak then joined the two initial designers.

Other small sensors

For information, these very small sensors are present on compacts, certain bridges and smartphones.

Visual comparison of the size of the main sensors

Large formats in digital photography

• Phase-one XF System, IQ4 sensor of 150 mpx and 54,3 x 40 mm
• Hasselblad X2D, sensor of 100 mpx and 43,8 x 32,9 mm
• Fuji GFX 100S, sensor of 102 mpx and 43,8 x 32,9 mm 
• Pentax 645Z, sensor of 50 mpx and 43,8 x 32,8 mm 

These expensive devices are the equivalent of what was the large format (6 x 6 cm, 6 x 7 cm, 6 x 9 cm) at the time of silver film. They are rather used in the studio, in architecture or in industry. Note however that some of these cameras are not more expensive than some high-end 24 x 36 mm full frame. Then there is the problem of the cost of optics!

For the vast majority of photographers, the choice of format will be between full format (24 x 36 mm) and APS-C format. The choice of large digital formats (Phase-one, Hasselblad etc.) will remain the prerogative of very specialized professionals, or wealthy amateurs!

Let’s also not forget the smartphones which have made great progress in photography, and which have replaced most compact cameras.

Commercial designation of the lenses according to the size of the sensor

Please note, most of these designations relate to the sensor format, it is now necessary to take into account the type of case (reflex or hybrid), the optical focus being different (shorter on hybrids due to the absence of mirror) the manufacturers opted for a different bayonet. (ex: mount Z at Nikon)
Note that full frame lenses can be used on APS bodies, conversely APS format lenses cannot be used on full frame bodies, because they do not cover the entire sensor, resulting in very significant vignetting , see a round image!

SPECIFIC ASPECTS OF SENSORS IN DIGITAL PHOTOGRAPHY

Sensitivity of digital sensors

The ISO standard determines the light sensitivity of digital sensors.
In film photography, the choice of film determines the ISO value. Each reel of film has its own sensitivity and it is therefore impossible to change the sensitivity from one frame to another on the same reel. The only possibility that existed in silver photography, was to “push” the sensitivity with certain films, for example from 400 iso to 800 iso (this operation necessarily applied to the entire film) and then to report it to the laboratory during the chemical treatment .

In digital photography, the photographer indicates the Iso value directly on the camera body. This sensitivity can thus be changed from one photograph to another.
Most digital devices have introduced the concept of automatic Iso. You indicate to the box a minimum iso value and a maximum iso value, as well as a minimum speed. The adjustment of the device will no longer be done on two parameters (speed/diaphragm) but on three parameters (speed/diaphragm/sensitivity).

For digital sensors, the increase in sensitivity (Iso) is linked to its definition, and to the size of the photosites. The larger the photosites, the better they react to light and easily transform it into electric current. Here we come to a crucial point in digital photography, the appearance of “digital noise” during the rise in iso.

Digital noise of sensors in digital photography

We call digital noise any parasitic fluctuation or degradation that the image undergoes from the moment of its acquisition until its recording. Digital noise is a general concept for any type of digital image. This regardless of the type of sensor at the origin of its acquisition. Visually, there are generally two types of digital noise that accumulate:

• Chrominance noise, which is the colored component of noisy pixels. It is visible as spots of random colors.
• Luminance noise, which is the light component of noisy pixels. It is visible as darker or lighter spots. These spots give a grainy appearance to the image.

In film photography, a similar problem existed (although of different origin). The sensitivity of the silver films was linked to the size of the silver crystals (the equivalent of the photosites of our digital sensors). On very sensitive films (having larger silver crystals) the image often had a more grainy appearance. This aspect could vary according to the brands of film and the nature of the developers used.

Dynamics of sensors in digital photography

We can define the dynamic range of a digital sensor as its ability to record a maximum of information both in the darkest areas and in the brightest areas of an image, either situations with very contrasting light zones. So we hear some photographers talk about a “blocked” area or a “burnt” area when commenting on their photos.

Exploiting the dynamics of a sensor

To make the best use of the dynamics of a sensor, it is preferable :

• to use RAW shots. JPEG should be avoided as much as possible. Because the latter is coded on 8 bits while the RAW is coded on 12 to 16 bits. Which means there is a lot more information available.
• use low sensitivities below ISO 400 as much as possible. The lower the sensitivity, the more information will be present in your digital file.
• to use good optics. Very bright optics will prevent you from increasing sensitivity too much in certain situations of lack of light.

Post-processing raw files is perceived by many photographers as a constraint, but this step is very important. RAW file processing software has significant capabilities to recover information in “blocked” or “burnt” areas.
If you start from the principle “I only work in Jpeg”, is it wise to invest in a very high-end device of which you will not exploit all the possibilities and qualities?

Dynamics of the sensors of some APN models

 IL :  luminescence index EV : Exposure Value

The dynamic is expressed either in contrast ratio (10,000:1 for example), or in dB, or in EV (or IL in French). More dynamic range sensor is greater, the more nuances and therefore more information available in the image.

Each time you double the amount of light, you increase by 1 EV, and each time you halve the amount of light, you decrease by 1 EV.

Some reference values:

• 3 IL : night or very dimly lit room.
• 16 IL : full sun with hard shadows.
• The human eye has a dynamic of 24 EV (with however an adaptation time)

The value of the diaphragms is always indicated in the form of a ratio (standardized scale):
f/1 – f/1,4 – f/2 – f/2,8 – f/4 – f/5,6 – f/8 – f/11 – f/16 – f/22 – f/32
This ratio corresponds to the focal length of the lens divided by the aperture of the diaphragm.
Example :

• a 300 mm focal length lens with an aperture of 75 mm corresponds to an aperture value of 300/75 mm either f/4
• a 500 mm focal length lens with an aperture of 178 mm corresponds to an aperture value of 500/178 mm either f/2.8

When changing from a standard aperture opening value to a consecutive value, the aperture opening area is multiplied or halved. As a result, the amount of light passing through the lens is also multiplied or halved. (see the table below)

EV Correspondence & Speed/Aperture Couples for 100 Iso

Focal length of lenses and their equivalences

It is frequently said that a 50mm lens on a full frame sensor becomes a 75mm on an APS-C sensor, or a 100mm on a micro 4/3 sensor. In reality the focal length of the lens does not change, a 50 mm remains a 50 mm whatever the size of the sensor used. In practice your 50mm on an APS-C sensor records a smaller image, because it cannot use the full area of the image sent by this lens, the result is the same as if you crop the image provided by a full frame sensor in post processing. This is the “crop” of 1.5 at Nikon/Fuji/Sony, 1.6 at Canon, 2 at Olympus/Panasonic if you use full-frame optics on an APS-C or Micro 4/3 sensor.

DEPTH OF FIELD AND SIZE OF SENSORS IN DIGITAL PHOTOGRAPHY

Definition of depth of field

Depth of field refers to the area of sharpness in an image, both before and behind the plane of focus. So in this area, subjects are sharp, while outside this area, subjects are blurry. The depth of field will however depend on several parameters:

• the focus distance
• the diaphragm opening
• the focal length of the lens
• the device sensor size

This notion of depth of field is thus linked with the notion of “bokeh“, the background blurs.

Depth of field related to sensor size

With small digital sensors, we see that the depth of field is greater than with a large sensor. Here it is not a problem related to the digital sensor, but a purely optical problem. This problem is the same in film photography. In practice, with a small size sensor, it will be necessary to use shorter focal length optics to obtain an identical image frame (see table).

example : If you’re using 50mm focal length optics on a full-frame, you’ll need to use 33mm optics on an APS-C sensor to get the same image frame. However, the shorter the focal length, the greater the depth of field. You get more depth of field with a wide angle than with a telephoto lens.

It suffices to refer to the notion of standard lens, that’s to say a lens having a field of vision substantially identical to the human eye (approximately 47°), it is estimated that the following focal lengths correspond substantially to the vision of the human field:

• with full frame 24 x 36 mm : lens 45 to 50 mm
• with APS-C format : lens 28 to 30 mm
• with the 6 x 6 cm and 6 x 7 cm format: the standard focal length is between 80 and 90 mm

Depth of field related to lens aperture

However, the size of the sensor is not the only element to be taken into account. The brightness of the lens (its maximum aperture) also has a big impact on the depth of field. The more you open the diaphragm of your lens, the more you decrease the depth of field. This is what a photographer does when he wants to isolate his subject, or have larger and more harmonious background blurs (bokeh). This operation is easier to perform with a very bright lens (f/1.8, f/2.8 or f/4 compared to lenses opening at f/5.6 or f/6.3).

So in wildlife photography, it was easier for me to get nice background blur when I used a Nikon 500mm f/4 AFS-VR telephoto lens than with a zoom Nikon 100/400mm f/4.5 – 5.6-S which at 400 mm opens at f/5.6, even if the latter allows to obtain pretty bokeh.

In principle, most lenses include a depth of field scale. On DSLRs where light metering is done with the lens wide open, there is a depth of field test button. This test button closes the diaphragm to the real value measured by the camera, to assess the depth of field in the viewfinder. On hybrids, the sight is always with real aperture.

GLOSSARY

• IL : luminescence index
• EV : Exposure Value
• Pixel : picture element, size of a few µm, joined or not
• Pitch : interpixel distance, varies by sensor format
• Format : line, interline, field, frame, full frame 
• Photosites : photoelectric cell which transforms a light intensity into an electrical signal
• Sensitivity : the amount of light and its color, depending on the semiconductor, the type of doping, and the coating 
• Noise : any signal not participating in the electrical transcription of the desired image, spurious signal
• Transfer rate (in pix/s, or in bytes/s): depends on the resolution and the format, transfer speed of a pixel

Link to Nikon website (not sponsored)