Essential knowledge


Pixel count

Exposure calculation

Color perception

Background blur

DOF calculator


Edit photos


Editing flow


I am not an artist, my subject is not the dramatics of lighting, or the esthetics of well positioned props, I am not interested in the type of photography where the photographer directs the moment (like in fashion photography).

I like the type of photography where the photographer hunts the moment (like in photojournalism and event photography). For me the subject is the human, animal or object which has determined me to take the photograph.

My photographic style is candid.


Display: NEC MultiSync LCD 2690 WUXI2.

Display calibrator (spectrophotometer): Xrite ColorMunki Photo.

Display calibration software: ColorMunki (Argyll and DispCalGui form a free, multi-platform solution, but inaccurate).

Photocamera: Canon EOS 40D.

Lenses (for the EOS 40D):

  • Canon L 70-200 mm F4 IS USM (the main lens).

  • Tamron 90 mm F2.8 Macro.

  • Canon 85 mm F1.8 USM.

Lens filters:

  • Hoya HD Protector (always on lenses).

  • Hoya HD Circular Polarizer.

Essential knowledge

Here is a summary of the most important factors that you must understand and control.

Sensor size

The most important technical which differentiates the various types of photo cameras is the amount of light which is captured for a photo. This amount of light is influenced by the size of the sensor. The size of the sensor influences in turn the size of the lens (mainly the diameter, but also indirectly the length).

The bigger a sensor is, the more light it captures for a photo and therefore the less noise that photo has (and looks technically better). The "noise" is something like television static: pixels with random colors and brightnesses.

A compact / pocket camera has a sensor size of about 8 * 5 mm, most cameras with interchangeable lenses (called DSLR cameras) have a sensor size of about 23 * 18 mm, while the so called "full frame" cameras have a sensor size of about 36 * 24 mm. There are cameras with even bigger sensors, but those are tens of times more expensive than common DSLR cameras.

The main advantages of a photocamera with a large sensor (= DSLR) compared to a photocamera with a small sensor (= compact camera) are:

  • Higher maximum background blur; DSLRs can achieve a much higher background blur than compact cameras can. This means that the foreground and background planes of the photographed scene have a clearer separation, which looks better for any type of photography (other than landscape). (This is an artistic advantage.)

  • Lower noise level per photo (which means better technical quality). (This is a technical advantage.)

  • Bigger and brighter viewfinder, which helps you see the photographed scene better. (This is a technical advantage.)

Focal length

The focal length (measured in millimeters) is the distance from the lens to the plane at which the photo is formed.

The length of a lens is an indication of a parameter called focal length; this can be set by rotating a ring on the lens (not the manual focus ring).

In compact cameras, the focal length is called zoom (and you usually can't see exactly which value is set). Typical compact cameras have an equivalent "full frame" focal length of about 30 mm for zoom 1x (= no zoom), and about 120 mm for zoom 4x (these relations are proportional).

When you take a portrait, you should stay 2...4 meters (7...14 feet) away from your subject and then change the zoom / focal length to fill the frame with your subject's head and shoulders.

You can see distortions due to the focal length here and here. You can see there that a long focal length (/ high zoom) is preferable for portraits.


  • Increasing it has a zoom-in effect, that is, it exposes the camera sensor to a smaller part of the subject.

  • Increasing it amplifies the background blur, especially in the far away background.

  • Distorts depth perception. Short focal lengths amplify depth (creating a spheric projection of the reality), while long focal lengths dampen depth (creating a planar projection of the reality). It's considered that an equivalent "full frame" focal length of 50 mm creates no distortions relative to what the human eye sees.


The F-number is the focal length of the lens divided by the diameter of the central circular opening of the diaphragm.

The lens diameter is an indication of a parameter called F-number; this can be set from the camera. The F-number determines the amount of light which enters through the lens (a larger diameter means that more light enters through the lens), and also the depth of the clarity zone (that is, how close or far is the background blur from the point of focus).


  • Increasing it allows more light to go through the lens and reach the sensor. The amount of light which reaches the camera's sensor is inversely proportional with square of the F-number.

  • Increasing it reduces the depth of field.

  • Increasing it reduces the background blur.

Shutter speed

The shutter speed (measured in exposures / second) is the inverse of the exposure time. The exposure time is the time during which the camera sensor captures light to form a photo.


  • Increasing it reduces the amount of light which goes through the lens and reaches the sensor.

  • Increasing it reduces the effect of (subject or camera) motion on the clarity of the photo.


The ISO is a post-exposure scaling factor which is applied to the light that reaches the camera sensor.


  • Increasing it amplifies the signal (and the noise) present in a photo, thusly producing a properly lit photo, but one where the noise destroys more and more details.

Background blur

The background blur is the amount of blur of a subject which is not in the focus plane, but either in front or behind it.


  • Shrinks with the distance to the subject.

  • Shrinks with the F-number. Grows with the aperture.

  • Grows with the focal length. The growth is limited by an asymptote which increases with the focal length.

  • Grows with the distance to the background.

Depth of field

The depth of field is the distance range in which the subject of a photo appears in focus (= clear).

The depth of field and the background blur are very important for closeups and portraits because they can produce a clear separation between the subject and the background, separation which has a very artistic effect.


  • Grows with the distance to the subject.

  • Grows with the F-number. Shrinks with the aperture.

  • Is virtually not affected by the focal length, for the same subject magnification (= subject size in photo).


Shoot in raw format and edit your photos!


Aperture (measured in millimeters) = The diameter of the central circular (really more of an octagon) opening of the diaphragm. The aperture of a lens is equal with the focal length of the lens divided by the F-number. A fast / wide aperture means a large aperture. A fast / wide / large aperture means a small F-number.

Autofocus = The automatic focusing process achieved by the camera and the lens. The precision of the autofocus is not constant and this is one of the factors which, at least in certain conditions, may turn photos into mere snapshots.

The autofocus is performed at a lenses physical aperture (for the current focal length), not at the F-number that you've set, which is very important in low light for focus speed and accuracy. Some autofocus sensors become active only if the F-number is smaller than a given value (like F2.8 for the center cross-type sensor).

Bokeh = The quality of the out-of-focus blur.

Circle of confusion (COC) = The amount of blur above which a person finds that the objects from a photo do not look sharp anymore.

The COC is in fact highly variable because it depends on the type of the viewed photo (on screen, printed or developed), on the size of the photo (which is normally 10...30 centimeters), on the viewing distance (which is normally 20...60 centimeters), and on visual acuity of the viewer.

Color temperature (measured in kelvin) = A characteristic of light which indicates its tint toward red or blue.

The complex version of the color temperature is the spectral power distribution.

The color temperature is used to set the white balance.

Crop factor = The ratio of the diagonals of a reference camera sensor (36 * 24 mm, 43 mm diagonal, called full frame sensor) and a compared sensor; this number is normally higher than 1. Basically, a cropped sensor is a sensor which is smaller than the reference sensor, that is, it "sees" a smaller part of a subject; the margins of the frame are cut out from the photo.

The surface of a cropped sensor is inversely proportional with the square of the crop factor. For example, Canon EOS 40D has a crop factor of 1.6 and a surface 2.56 times smaller than the reference sensor.

Diaphragm = A circular mechanical device (in the camera) which controls the aperture of the lens.

Dynamic range (measured in stops) = The ratio between the brightest and darkest spots captured by a camera sensor, where detail is still visible in the resulting RAW image; this may be expressed on a logarithmic scale. For example, the dynamic range of a sensor can be 12 stops, which means a 2048 (= 2 ^ 12) ratio.

Exposure = The capturing of the light which goes through the lens, by the camera's sensor.

Exposure compensation = Manually set bias of the automatic exposure determined by the photocamera, used for exposure modes like "Aperture Priority". If the taken image is too dark then set the exposure compensation to a positive value, and if the taken image is too bright then set the exposure compensation to a negative value.

Exposure time (measured in seconds) = The time during which the camera sensor captures light.

Exposure value = The combination of the aperture, exposure time and ISO. These parameters are responsible for the physical exposure of photos, and can't be change later.

F-number = The focal length of the lens divided by the diameter of the central circular opening of the diaphragm. A fast lens means a small F-number.

The amount of light which (goes through the lens and) illuminates the camera's sensor is inversely proportional with square of the F-number. You may be tempted to think that the amount of light should be proportional with the surface of the opening through which light enters through the lens, but it's not so.

F-number progression (each step represents a stop of light): 1, 1.4, 2, 2.8, 4, 5.6, 8, 11, 16.

Focal length (measured in millimeters) = The distance from the lens to the plane at which the image is formed. Increasing the focal length of the lens has a zoom-in effect, that is, it exposes the camera sensor to a smaller part of the subject.

Focus = The state in which the subject of a photo has maximum clarity / sharpness.

Gamut = A subset of the entire range of colors perceivable by the average human eye.

When you hear that some display is able to show millions or billions of colors, that does not refer generally to the "amount" (= gamut) of colors which can be shown, but refers to the detail of the shown colors (which has little relevance because from some point the eye can't perceive the difference).

ISO = The sensitivity to the light, of the camera's sensor.

Magnification = The ratio between the size a subject projected on a camera's sensor and its real size. Macro lenses typically have a magnification of 1, that is, a subject 10 millimeters tall also has 10 millimeters on the sensor; in such a case, if the sensor is 25 millimeters tall, the subject takes 40% of the photo's height.

Magnification of subject = The percentage of a subject from the total photo height / width.

If a person is photographed with a camera sensor of a given size and a given focal length, and then with another camera with a larger sensor and the same focal length, the subject looks smaller in the second photo because the larger sensor literally has more sensor space around the smaller sensor.

In order to have the same subject magnification in the two photos, the second photo has to be taken either from a smaller distance or with a longer focal length; the proportion is equal with the crop factor between the two sensors (= the width / height / diagonal of the larger sensor divided with the width / height / diagonal of the smaller sensor).

Metamerism = The matching of colors of with different spectral power distributions.

Colors reflected by objects may look differently depending on the spectral power distribution of the light which illuminates them.

For example, if an object is illuminated with sun light, it shows certain colors. However, if the object is illuminated with incandescent light, it shows other colors; also, some of the colors which were distinct in the sun light, now look the same. These similar colors are called metamers.

Noise = Distortion of pixel / color information inherent to camera sensors (= the noise doesn't come from the outside). The main reason why bigger camera sensors yield cleaner images is that for the same exposure (and subject magnification) they capture more light, while their noise remains relatively the same.

Shutter = A mechanical device (in the camera) which controls the time of the exposure of the camera's sensor to the light which goes through the lens.

Shutter speed (measured in exposures / second) = The inverse of the exposure time. A fast shutter speed means a small exposure time.

Spectral Power Distribution (SPD) = A large set of data which describes the energy per unit area per unit wavelength of the light.

The spectral power distribution of the light is of critical importance in color perception because physical colors are not simple sets of RGB values (as their are in photocamera sensors and computer displays). In fact, a physical color is a complex set of energy data at all the wavelengths visible to the human eye, which the eye transforms into a more simple set of data recorded with its rode and cone cells.

For example, an orange does not emit orange light. Instead, it absorbs all the wavelengths of the illuminating light except for a reflected set of wavelengths with different energies which people perceive as orange.

The simplistic version of the spectral power distribution is the color temperature.

You can see examples here.

Stop = A doubling or halving of the amount of light which goes through the lens.

Stop down = A halving of the amount of light which goes through the lens. This is equivalent with a doubling of the shutter speed (because this means that the camera's sensor is exposed to less light), and with a decrease of the aperture of 1.4 times (it's actually square root of 2).

In general, cameras display their shutter speed and F-numbers in (approximations of) thirds (or halves) of units in order to make it easier for the photographer to manually control the exposure, that is, to know how much the camera would expose an image.

Stop up = A doubling of the amount of light which goes through the lens. This is equivalent with a halving of the shutter speed (because this means that the camera's sensor is exposed to more light), and with an increase of the aperture of 1.4 times (it's actually square root of 2).

Tonal break = A visible outline (in an image) caused by an insufficient tonal range along the outline. If the tonal range were higher, the outline would be a smooth gradation.

Tonal range = The number of tones which show detail throughout the dynamic range; this may be expressed on a logarithmic scale. For example, the tonal range of a camera sensor can be 14 bits, which means 16384 (= 2 ^ 14) intervals.

Unit area = an absolute measurement unit for areas, like a square millimeter.

White balance = The real color of objects, as seen by the human eye or photocameras, depends on the color temperature of the light which reaches them. White balance refers to the adjustment of colors so that the colors from a photo look similar to those seen by the eye in normal daylight conditions.

White balance is crucial for indoors photos, photos which have a color cast from the incandescent or fluorescent light bulbs.

Zoom = The ratio between two focal lengths. An increase of the focal length of a lens results in a decrease of its angle of view, which means that a smaller part of the subject is exposed to the camera sensor, which means that the subject is zoomed-in (that is, its size is increased on the sensor). A camera's crop factor doesn't affect the zoom.

Human eye

People see things differently than a photocamera captures images. The human eye continuously sends images to the brain; also, different cells from the eye send parts of the viewed images in different ways: some faster, some slower. The eye adapts to the environmental light.

People can see very short (single) flashes of light, even shorter than 1/200 seconds, because the light stimulates the cells from the eye and it takes some time for them to become unstimulated, that is, for the stimulus to dissipate. If the eye cells are stimulated with another (similar) flash before they become unstimulated, the eye perceives a continuous light.

There is also a difference between a perceived flash light and what the brain can interpret (in order to recognize a color, a shape or an object).

Background blur

The background blur is the amount of blur of a subject which is not in the focus plane.

The background blur:

  • Shrinks with the distance to the subject.

  • Shrinks with the F-number. Grows with the aperture.

  • Grows with the focal length. The growth is limited by an asymptote which increases with the focal length.

  • Grows with the distance to the background.

The more blurry the background is, the more the subject "pops" out from the image.

You can read here the technical details.

Depth of field

The depth of field (DOF, or depth of field of clarity) is the distance range in which the subject of a photo appears in focus.

The depth of field:

  • Grows with the distance to the subject.

  • Grows with the F-number. Shrinks with the aperture.

  • Is virtually not affected by the focal length.

The depth of field beyond the subject is always greater than the depth of field in front of the subject.

The closer the camera is to the subject (in focus) and the farther the background is from the subject, the higher the background blur is and therefore the more the subject "pops" out from the image.

Use a high aperture (= small F-number) and a long focal length in order to blur the background and create a beautiful closeup (like a portrait). Use a low aperture (= high F-number) in order bring the entire range from the foreground to the background into focus.

The depth of field and the background blur are very important for closeups because they can produce a beautifully blurred background.

Angle of view

The angle of view is the angular extent visible in a photographed scene. The angle of view decreases as the focal length of a lens and the crop factor of a camera increase.

A photocamera (like the human eye) sees a scene in a conic manner, that is, the observed objects are within a cone which starts as a point at the lens (/ eye) and ends at infinity as a huge area.

Because a photocamera sees a scene under an angle of view, objects of a given absolute size appear to have different relative sizes (on the camera's sensor) when they are located at different distances from the lens (/ eye). For example, if you look at a mountain from its base, it will look massive to you (that is, it will occupy a huge portion of your visual field), but if you look at it from far away, it will look tiny (that is, it will occupy a small portion of your visual field).

Sometimes, it's not just an entire object that appears to have a different size, but also parts of an object. For example, when photographing a building from a small distance, at ground level, the top of the building will appear smaller than its base. This happens because the distance to the top of the building is significantly larger than the distance to the base of the building, and as such the top will occupy a smaller area (than the base occupies) on the camera's sensor. If you would photograph the same building from far away, the difference between the distances would be too small to be noticeable.

Perspective distortion

Perspective distortion is a visible alteration of absolute measures.

Think at photos taken of movie stars (on the "Red Carpet"). Their feet look thin and their heads look big.

Such photos must be taken from very close to the subject, or risk having obstacles in between, and therefore in order to frame the entire subject they must be taken with lenses with a short focal length.

Now think at the position of the camera. It's usually at the level of the subject's head.

Because of this, the distance from the camera to the subject's head is significantly smaller than the distance from the camera to the subject's feet. Therefore, X centimeters measured over the subject's head take a specific number of pixels on the camera sensor, and the same X centimeters measured over the subject's feet take a significantly smaller number of pixels on the sensor; this is because objects which are father away appear smaller (than closer objects).

This is why the feet of the subject appear so thin when compared to their head.

In opposition to this, when taking a photo of the same subject with a lens with a long focal length from farther away (in order to preserve the size of the subject on the sensor), the differences between the distances become insignificant (, making the photos look flatter / 2D).

To summarize, taking photos with a lens with a short focal length exaggerates the differences between the distances to the various parts of the subject, while taking photos with a lens with a long focal length flattens / compresses the differences between the distances to the various parts of the subject (relative to the distance to the subject, which keeps increasing).

Taking photos with a lens with a short focal length is not good for portrait photography, and a long focal length may require too much of a distance between the photographer and the subject, which is why the preferred focal length for this is somewhere between 85 and 135 mm, depending on the crop factor of the sensor. In portrait photography, a very high background blur is also desired, which in turn requires either a wide aperture or a short distance to the subject (and at the same time a large distance to the background, and possibly small subjects in the background so that their details can easily disappear in the blur).

Although taking photos with a lens with a long focal length flattens the differences between the distances to the various parts of the subject, thusly showing all the objects from a photo at their real proportions, this is still called perspective distortion because it's unnatural for the human eye to see this flattening of distances.

You can see distortions due to the focal length here and here.

High ISO look-and-feel by light intensity

Photos taken at a high ISO in sunlight with a good intensity exhibit a lower noise level and look better than those taken in low light because:

  • The dynamic range of scene illuminated with intense sunlight is higher than the dynamic range of scene illuminated with low light. This is because the environmental light reflects on atmospheric particles and on everything it touches, so it creates a sort of a floor for the light which will be captured in a photograph. But if the light which falls directly on the photographed subject is high then the ratio between the intensity of this light and that of the floor light is high, hence the dynamic range is high. If the light which falls directly on the subject is not intense enough compared to the floor light, the dark tones are downed by the floor light.

  • The intense sunlight has a spectral power distribution which better matches that of the sensor's photosites, which means that the tonal resolution of the photos taken in such light is much better.

  • The noise from a bright scene exists mainly due to photon noise, not due to the electronic / read noise. Photon noise increases with the brightness but only as the square root of the number of photons absorbed by the sensor, so the signal-to-noise ratio of bright scenes is higher than that of dark scenes. Read noise is relatively stable and does not increase with brightness.

Light changes with focus

In some lenses, changing the focus range from infinity to the minimum focusing distance (MFD) decreases the amount of light that goes through the lens (even though all the other exposure parameters are the same). This is typical for macro lenses.

To experiment, set the camera and the lens in full manual mode. Focus-out completely, at infinity, point the lens at a white wall and take a photo. Now focus-in completely, at the minimum focusing distance of the lens and take a photo. If you compare the two photos you can see that the second photo is (much) darker.

For lenses whose barrel extends when focusing-in, it appears that the amount of light which goes through the lens decreases linearly with the length with which the barrel extends.

As an example of this, for the Tamron 90 mm (whose barrel extends when focusing-in) the second photo has about 2 stops less light than the first photo. When the barrel is half extended, the taken photo has about 1 stop less light than the first photo.

For the Canon 70-200 mm the second photo has about 1/3 stops less light than the first photo, that is, virtually no light loss.


A protection filter doesn't visibly decrease the amount of light that goes through the lens.

Use a polarizing filter to eliminate reflections from non-metallic objects, like vegetation. The lack of such a filter can, for example, ruin the photos you take to vegetation after a rain.

A circular polarizer filter decreases the amount of light that goes through the lens with 1...4 stops. For example, a Hoya HD Cir-Pl takes about 1 stop (regardless of its rotation angle).

Rotate the mobile part of the circular polarizer filter to achieve the effect that you want.

Use this filter with care because it may affect the colors in an undesired way.

Pixel count

Current photocameras with Bayer sensors, like the usual Canon and Nikon photocamera, claim to have resolutions of X pixels (like 10 mega pixels). However, this is in fact the number of photosites, not pixels.

A photosite is really a monochromatic sensor, normally for either red, green or blue colors. They are arranged in a matrix of RGGB patterns (that's two photosites for green).

The reason why photos taken with such camera sensors really have X pixels is because the software which converts the sensor data into an image interpolates each photosite to an RGB pixel, also using the data from the surrounding photosites. This means that while you expect the software to convert each RGGB pattern to an RGB pixel, it really doesn't do that; in reality, it upscales the image 4 times.

This upscaling is a good approximation of an image with the same number of pixels because it contains certain information (= luminance resolution) to which the human eye is more sensitive than it is to color.

Exposure calculation and camera sensor size

The exposure of a photocamera is calculated per unit area (= a square millimeter) of the camera sensor because a unit area receives the same amount of light no matter what the total sensor area is.

In photocameras, the sensor size and the diameter of a lens used for it are proportional, so that for each unit area of the sensor there is a corresponding unit area of the lens, which means that each unit area of the sensor receives the same amount of light.

However, this means that larger sensors capture a higher total amount of light than smaller sensors, and that the images formed on larger sensors are brighter than the images formed on smaller sensors. This doesn't mean that the photos taken with larger sensors are brighter, since such photos are not viewed at a similarly increased size (they are viewed on the same displays and printed on the same paper).

There is, however, a consequence of this phenomena: the higher amount of light captured by the larger sensor overwhelms the noise, which means that the photos taken with it have lower noise levels if they are taken with the same subject magnification / size (which is usually what photographers do).

Different approach

In current camera sensors, the size of a photosite (in other words, resolution) has no visible effect on the noise level of the entire photo (= it only affects the noise level of the individual photosites). The only thing that matters is the size of the sensor (well, obviously, besides the technology).

The reason why there are sensors for phones, for compact cameras, for DSLRs, for medium format, and... telescopes, is the size of the sensor, not the size of the photosite. The bigger the light capturing device (= lens + sensor) is, the more light is captured for the same exposure.

It's difficult for people to understand that a bigger sensor means that the photos taken with it contain more light for the same exposure. They ask themselves where did that light go because they don't see brighter photos. The answer is simple: the photos are scaled to the same physical size (but the scale factor is smaller, because the sensor size is scaled to the same display / paper size), so the light goes instead into annihilating the noise.

ISO and noise level in the shadows

A photo taken with a high ISO exhibits a lower noise level than a photo taken with a low ISO, in the shadows.

While this may sound impossible, you can see the evidence here.

The explanation is that while the camera sensor has a certain dynamic range, the rest of the camera circuitry (through which the photons and electrons pass in order to form an image) has a lower dynamic range, so the noise floor for different ISOs is different.

Color perception

A color is a combination of several factors:

  • The spectral power distribution of the light which illuminates the colored object.

  • The way in which the spectral power distribution of the illuminating light is changed when the light is reflected by the colored object. This change depends on the full spectral power distribution (not only on the relative spectral power distribution), that is, it depends on the intensity of the illuminating light.

  • Atmospheric reflections (usually caused by dust and smoke). These may cause a reduction of the contrast (where blacks turn to gray), or a compression of the black tones (up to the point where they become mudded / indistinguishable).

  • The way in which the spectral power distribution of the reflected light is recorded by a photocamera sensor or by the eye (and interpreted by the brain). This change depends on the full spectral power distribution (not only on the relative spectral power distribution), that is, it depends on the intensity of the illuminating light.

A color perceived by the human eye is a spectral power distribution function of an illuminating light which is altered when reflected by matter into another spectral power distribution function and then read by the human eye and altered into a much more limited spectral power distribution function (by the so called RGB receptors), and finally processed by the brain into something we perceive as color.

For natural color appearance, the spectral power distribution of the illuminating light must be as close as possible to that of the natural sunlight (actually it's a standard called D65).

An object exhibits different colors in different lighting conditions, even if only the intensity of the illuminating light is different.

For example, the color sensitivity of the human eye for red is high when the light is intense, but as the light gets dimmer, the sensitivity for red decreases in favor of blue, that is, all the colors are shifted toward blue.

This is why using light bulbs with the same color temperature as day light (6'500 Kelvin) gives a bluish tint to all objects, while light bulbs with a lower color temperature (like 4'000 Kelvin) gives a neutral tint; the average incandescent light bulb has a color temperature of about 2'700 kelvin, which gives a red tint to all objects.

If the light bulb with a color temperature of 6'500 Kelvin were to give a light as intense as the day sun, then it would give no tint to the illuminated objects. However, the light from the average light bulb is several thousand times less intense than the light from the day sun.

The dynamic range and tonal range of a photographed scene depend on the intensity of the illuminating light. If the light intensity is too low, both ranges are low and the photos took in such conditions look rather gray. If the light intensity is too high, the photos are too contrasty, with blown out (= without detail) shadows (= blacks) and highlights (= whites).

From a technical point of view, photos look their best when their dynamic range is at a medium value, and their tonal range is at the maximum value. This is achieved when the intensity of the illuminating light is at a medium value. Basically, both ranges have to cover the optimum intervals perceived by the human eye.

You can see here how the environment looks like when lit with light of various color temperatures.

For details, see color vision, D65.

Light intensity

Here is the usual light intensity for a few common cases:

  • Livingroom: 50...200 lux.

  • Heavy overcast day: 1'000 lux.

  • Bright daylight (not the Sun itself): 10'000...100'000 lux.

The perception by the human eye of the light intensity is not linear with the intensity, but with the type of the light source, with the environmental light intensity, with a power of the intensity, and with its color (or more accurately, with its spectral power distribution).

Due to the complexity of the factors involved, the actual correlation equation depends on the application. For a general application, Stevens' power law is used. You can read more at Telescope Optics.

Color temperature

The Kruithof curve correlates color temperature and intensity, from a visually pleasing point of view.

A light with a low color temperature gives a warm appearance (red, orange, or yellow), and a light with a high color temperature gives a cool appearance (blue or white). The sunlight at the noon of a bright summer day has a temperature of 5'500...6'500 Kelvin. Incandescent light bulbs have a color temperature of about 2'700 kelvin.

In photography, the color temperature is upside down due to psychological reasons. A high temperature is used to indicate a warm color (yellow or red, which have a low temperature in physics), while a low temperature indicates a cool color (blue, which has a high temperature in physics).

Spectral power distribution

Each spectral power distribution set distorts the colors of the illuminated objects compared to other SPDs, whether the light comes from a natural source such as sunshine, dawn, sunset, or electric sources such as incandescent and fluorescent light bulbs.

Effectively, each point of a photographed scene responds differently, depending on the properties of the light which illuminates it, in such a way that it makes it impossible to obtain its standard color from its perceived color by applying color corrections (like white balance) on the entire photo. "Standard color" refers to the color which is obtained by photographing a scene in standard / D65 light. This is why photos taken in bad lighting conditions are simply bad (from a technical point of view).

The spectral power distribution is the reason why white balancing photos doesn't always yield the desired results. Each object from a photographed scene responds differently to the illuminating light, so white balancing a part of a photo may destroy the white balance of another part of the photo. This happens usually when the illuminating light is artificial.

Color correction

Some lighting conditions produce good or acceptable colors, while others produce bad or horrible colors. In other words, some lighting conditions produce colors which can be corrected so as to be similar to those produced in standard D65 light, while others produce colors which can't be corrected no matter what.

For example, sodium light stimulates matter only on a very narrow wavelengths band, while sun light stimulates matter on a very wide wavelengths band.

In such lighting conditions, the colors recorded on the photocamera sensor can't be corrected with a white balance because they are either literally not reflected by the matter, or they are recorded by the sensor as a metamer rather than the standard color (that is, colors which would be distinct in a standard D65 light, are the same in sodium light).

You can get acceptable colors in some indoor ligthing conditions, but some ligthing conditions are just bad.

Color corrections can be done with a custom white balance (using a gray card), but such cards only record what gray looks like (which gives an incomplete description of the illuminating light).

Color corrections can be done more accurately with a Color Checker, because this records what a large set of colors looks like in a given light. The associated software then reverse engineers the standard D65 colors.

Unfortunately, a photocamera sensor doesn't record the spectral power distribution (= the powers for all the wavelengths) at each pixel, but only the powers for the RGB wavelengths. So, in some cases, color corrections can't produce acceptable colors.

Background blur

Here are some graphs which depict the background blur for several photo setups.

The X axis represents the distance from the point in focus (not from the camera) to the background, and is measured in meters. An X higher than 0 represents the background.

The Y axis represents the blur of the background (beyond the point in focus), and is measured in micrometers. The blur is 0 in the focus point (0 on the X axis), that is, the objects from that point are in perfect focus.

The higher the value on the Y axis, the higher the blur is.

The red line represents the COC for Canon EOS 40D: 19.

If the blur is smaller than the COC (of the camera) then the blur at that distance would be imperceptible in a photo.

We use the following formula (derived from here) for calculating the background blur: b = f ^ 2 / (s - f / 1000) / N * |x| / (s + x) , where "b" is the blur (in micrometers), "f" is the focal length (in millimeters), "m" is the subject magnification (on the camera's sensor), "N" is the F-number, "s" is the focus distance (in meters), "x" is the distance between the subject and the background (in meters). A negative "x" yields the foreground blur.

For the far background, the blur has an asymptote which can be approximated with: f ^ 2 / s / N . This asymptote is reached when the distance to the background is significantly larger than the focus distance.

We consider that all lenses take photos of a subject at the same magnification (= size in photo).

Magnification formula: m = f / (s - f) .

In order to preserve the magnification of lens 2 the same as the magnification of lens 1, we use the following formula to calculate at what distance from the subject should lens 2 be: s2 = s1 * f2 / f1 .

The format of the camera's sensor is not a direct factor of the equation. However, the magnification (and consequently the focus distance) and the COC depend on the format.

COC formula: coc = sd * va * vd / pd , where "sd" is the sensor's diagonal, "va" is visual acuity of the average human eye, "vd" is the viewing distance of the photo, "pd" is the photo's diagonal.

The average human eye can see (according to this), in normal light, up to 50 black-white line pairs, contained within 17.5 millimeters, from 1 meter away. This gives a visual acuity of 0.35 millimeters from 1 meter away (about 1 / 3000).

The vd / pd ratio is matter of preference. Personally, I view postcards from a distance equal with 2 to 3 times the photo's diagonal, an A4 photo from a distance equal with the photo's diagonal, while a full screen photo on my display from a distance equal with 1 to 3 times the display's diagonal.

A common COC formula (found here) is coc = sd / 1500. To obtain this formula, in DOF calculator you have to use a "Desired COC multiplier" of 2.

You can use the DOF calculator to calculate the depth of field.

Graphs are generated with WZGrapher. (Alternatively, use SpeQ Mathematics)

In this graph you can see:

  • The blue curve represents the background blur for a lens with a focal length of 400 mm and F4.0, focused at 40 meters.

  • The green curve represents the background blur for a lens with a focal length of 200 mm and F4.0, focused at 20 meters.

  • The yellow curve represents the background blur for a lens with a focal length of 135 mm and F2.0, focused at 13.5 meters.

  • The purple curve represents the background blur for a lens with a focal length of 85 mm and F1.4, focused at 8.5 meters.

  • The cyan curve represents the background blur for a lens with a focal length of 50 mm and F1.4, focused at 5.0 meters.

  • The brown curve represents the background blur for a lens with a focal length of 35 mm and F1.4, focused at 3.5 meters.

The formulas are:

400 ^ 2 / (40 - 400 / 1000) / 4 * |x| / (40 + x);
200 ^ 2 / (20 - 200 / 1000) / 4 * |x| / (20 + x);
135 ^ 2 / (13.5 - 135 / 1000) / 2 * |x| / (13.5 + x);
85 ^ 2 / (8.5 - 85 / 1000) / 1.4 * |x| / (8.5 + x);
50 ^ 2 / (5.0 - 50 / 1000) / 1.4 * |x| / (5.0 + x);
35 ^ 2 / (3.5 - 35 / 1000) / 1.4 * |x| / (3.5 + x);

You can see in the graph above that the 85 mm lens has the highest blur for the most part of the background.

When the focal length and the F-number are multiplied with the same factor, far away in the background the lenses have the same asymptote.

For a given F-number and subject magnification, the depth of field is barely affected by the focal length, the effect is getting smaller as the focal length increases.

For example, for Canon EOS 40D, with a lens set at F4, a 20 mm focal length focused at a distance of 2 meters has a 1.72 meters DOF, a 200 mm focal length focused at 20 meters has a 1.48 meters DOF, and a 2000 mm focal length focused at 200 meters has a 1.48 meters DOF.

However, because the asymptote of the background blur grows with the focal length, the far away background plane appears more clearly separated from the focused plane when a long focal length is used, and this gives a more pleasing aspect to portraits.

DOF calculator

DofCalc is web-browser based application which calculates the depth of field for a given set of photographic parameters.

DofCalc is an open source application developed in HTML and JavaScript, specifically designed for mobile devices (like PDAs).

You can download it on your computer by right-clicking here and choosing "Save link as". Then just click on that file to open it in your web-browser and use it.

First public release of DofCalc: version 1.0 on 09 June 2009.


Photography is about perception / emotion, which is influenced by geometry (= composition, distortions due to the focal length and angle of view, depth of field / background blur), which is influenced by light (= natural or artificial, mood giving), which is influenced by directing, which is influenced by the scene (= subject + background).

Post-processing is used to alter the light / color and geometry after the fact.

Things which make a good photo: the moment (= the event, the feeling, the emotion, the message, the uniqueness), action, light, color, people.

A good photo is made by the subject, lens, camera and editing software, not by the photographer. The photographer is there to simply capture the moment and bring forward its soul, not make it.

The most important thing about light is not its amount, but its properties, like: color (or more accurately: spectral power distribution), the lack of reflections either in the atmosphere or from the photographed objects, the limited light contrast (between shadows and highlights), the high color contrast.

Things which may destroy a good photo

Missed focus: focusing either in front or behind the subject.

Inappropriate depth of field: too shallow for macros or too deep for close-ups.

Framing only a part of the subject: for example, the subject's legs are outside the frame.

Excessive light contrast: striking contrast between shadows and highlights.

Use both the landscape and portrait formats

The usual format of a photo is landscape because this is how the camera sensor is positioned in the camera, as the camera is held in its most comfortable position. Turn the camera vertically to take photos in a portrait format. Individual people or flowers are easier to isolate this way, so they may look better.

Use the flash outdoors

A strong sun light causes a strong light contrast. You can eliminate the shadows made on people's faces by lightening the subject with a flash. This is called a fill flash.

Be aware of the light

Avoid taking photos facing sources of strong light, like the sun.

Avoid taking photos of people facing strong sources of light because they will squint.

When taking photos of people in strong sunlight, position yourself to have the sun in front of you, 45 degrees at your side. Otherwise, the strong light will create strong shadows.

Early or late in the day fills the environment with a golden light.

Very good conditions for outdoor portrait photography are when the sun is covered by a thin layer of clouds which diffuse the light and thus eliminate shadows on the ground. Shadows are the main cause for ruined outdoor portrait photography, although they are sometimes used for establishing a mood.

A low ceiling of white clouds may cast a diffuse light which brings out the (saturated) color of the landscape.

Hold the camera steady

If you can't put the camera on something steady, like a tripod, hold it with a hand on the side which has the shutter, and with the other hand support the lens from beneath.

Shutter speed - focal length

In order to take sharp photos while hand-holding the camera, without image stabilization, use a shutter speed equal with or higher than the focal length.

For example, if you're using a lens with a 100 mm focal length then use a shutter speed equal with or higher than 100. If you're using a lens with a 200 mm focal length then use a shutter speed equal with or higher than 200.

Get to the subject's eye level

This may mean squatting so that the camera would be at the subject's eye level, like when you are photographing a child. This makes the photo more personal.

Get close to the subject

For photos with a more personal touch, fill the entire frame with the subject.

The background matters

The background (either in front or behind the subject) is at least as important as the subject.

Don't center the subject

An off center position, in landscape format, for the subject is usually better than a centered one. This can be done later by cropping the photo more on one side.

Photographed subjects are not three-dimensional

Something which looks beautiful to your eyes doesn't necessarily look beautiful in a photo.

This is for various reason, but one of them is that a photograph doesn't show the subject three-dimensionally, as your eyes do. Therefore, you have to find an angle which gives depth to the subject.

For example, a flower looks beautiful to your eyes from front, but when you're taking a photo of a flower you should try an off-center angle.

Focus the entire subject

If the entire subject's depth appears in focus, the photo appears clearer / sharper. However, extending the depth of field too much makes the background look less attractive.

Generally, the closer you get to a subject, the higher the F-number must be in order to create a greater depth of field.

Long exposure times

When taking photos with a long exposure time (like 1/10 seconds), take them in continuous shooting mode. You'll have a greater chance to get one of them sharp.

Take photos of people when they are not posing

Taking photos of people who are not preparing to be photographed gives a natural look to the photos.

Protect the highlights

When you have to deal with a very high dynamic range (colloquially known as a contrasty situation), like an illuminated patch of trees whose trunks are in the shadow, make sure that you record the images in RAW format, and that the image contains properly exposed bright areas (you will get nearly black dark areas).

Then, in an editing software lift the shadows as much as you can. This will brighten the dark areas of the image and will give the image a look closer to what you have seen with your eyes, closer to what professional photographers publish, and very different from what you are used to see from an amateur camera.

If you try to protect the dark areas instead of the bright ones, therefore exposing the dark areas properly and getting nearly white bright areas, it's most likely that the processed image will have very distorted colors, especially in the bright areas.

In other words, it's far easier to recover the shadows (with editing software) than the highlights, so it's better to protect the highlights when the image is taken.

Photos can't be taken in all conditions

While pictures can be taken in any environmental conditions, to any subjects, photos can not because the conditions either don't convey anything of importance to the viewer, or the lighting is really bad.

Never share bad photos

Never share bad pictures, especially of people. Delete them immediately. Make sure people get good photos of themselves. After a while you'll gain a good reputation and people will feel much more at ease with you taking photos of them.

Clear the skin defects, like zits, blemishes and moles. They distract the viewer's attention to an exaggerated level, and hide the underlying beauty of the bones, muscles and skin.

Long focal length

Taking photos with a long focal length produces a sensation of special harmony. This happens because:

  • Perspective distortion makes objects with equal absolute measures (in meters), located at different distances in the frame, occupy a similar number of pixels in a photo. This way, all the objects from a photo are shown at their real proportions. Even though this is not natural to the eye, it looks beautiful.

  • The viewer is "taken" / "flown" to the subject, as if zooming live into the frame of the photo, especially if the foreground is also blurred and the subject occupies only a part of the frame.

Choosing a lens

When you choose a lens over another, there are several characteristics that you have to consider:

  • The focal length range allows you to zoom in or out, and to alter the perspective distortion of the subjects.

  • The minimum F-number allows you to maximize the light gathering capabilities, and the depth of field.

There is no lens which is better than others for taking portraits, it all depends on what distortion you are trying to get, and on what space is available from you to the subject. While a small minimum F-number allows you to get a beautiful background blur, it also decreases the depth of field so much that you get a blurry portrait.

Cleaning glass

Wipe the entire surface of the glass (from a lens or a filter) with a wet microfiber tissue (moistened with ethanol), in a circular manner, in order to remove dust; a dry tissue may scratch the glass. The tissue must be label for cleaning glass.

Do this as rare as possible because the ethanol will eventually strip the delicate coating off the glass. Generally, dust will not show on photos, so there is no real need to remove it.

Then, before the moisture evaporates (which does quickly), wipe the glass with a dry tissue until all the moisture is gone.

If you exhale on the glass, you'll see the moisture from your breath condense on the glass and expose the cleaning pattern; this moisture evaporates quickly.


Canon EOS cameras automatically lock the exposure when "Evaluative" metering and "One-Shot" auto-focus mode are used. Simply press the shutter button half-way and both the focus and the exposure will be locked. Good for shooting portraits: lock on subject, then recompose.

Look at photos on a professional display. A notebook display is usually limited in its capabilities to reproduce colors. A good choice is to print the photos.

Sometimes, colors look better if the photo is underexposed. I find this to be true even for an underexposure of more than 1 stop.

Edit photos

The most important thing that amateur photographers miss is the extra step that professional photographers are taking: photo editing (also called post-processing) with the explicit goal of making the photos consistent (with what the photographer wants).

The problem with the photos taken by the camera is that they are perceptually inconsistent. The camera records the reality with a number of hardware parameters which can't be change later.

Everything else done by the camera to output a photo is just post-processing, that is, the modification of photos after they were physically recorded, using software.

If you are taking photos using a semiautomatic photo mode, that is, a mode where the camera automatically computes at least one of the hardware parameters, the camera will miss (usually by a small amount) a lot of times to use the hardware parameters that a photographer would use if he would have the time to manually set them.

Moreover, reality is not consistent. Light (amount, spectrum and directionality) and subject colors vary all the time.

Reality is boring, grayish. When taking landscape photos, you're also facing the problem of atmospheric conditions, usually in the form of suspended dust particles which reduce the contrast of distant subjects (making them look grayish).

A significant part of reality is always lost in its path to be seen as a printed photo (or as one displayed on a computer screen). The factors which generate a loss of reality are: light, camera lens, camera sensor, camera hardware parameters, photo post-processing, printing photos (= paper, ink and quality), displaying photos on a computer screen, the light in which the photos are seen, the eyes of the viewers.

Editing photos is the only way to bring the technology of the camera-display pair on par with the human eye. A camera records a scene in a colorimetrically correct manner, but the camera-display pair is literally incapable of reproducing the same scene that the photographer has seen with his own eyes. A camera and a display exceed the human eye on some levels, but on things like dynamic and tonal range the human eye is still superior and will be so for quite a while.

Also, the eyes of the photographer were physiologically adapted to the photographed scene, but those of the viewers are adapted to the environment in which they see the photos.

In the end, the reality seen by the eye of the photographer is less important than what the viewers of the photos see. The only good photos are the ones which look good to their viewers.

People see the grayish reality everyday, so why not make it look better in photos?

Editing photos is like using the manual exposure mode from a camera. Edit your photos to fit your taste! Don't worry about details not looking right anymore (because, for example, you've increased the contrast), a photo looks good when viewed overall.


Beauty dish

A beauty dish is a parabolic light reflector (= indirect light source) where the light beams converge in a focus point, creating a balance of hard and soft light.

It has an optimal distance from the subject, usually about twice the dish diameter.

The light created is between that of a direct flash and a softbox. It is evenly spread and has good feathering at the edges, and it gives the image a wrapped, contrasting look.

The light is crisp, detailed, delicate, smooth, luminous.

It is great for portraits, beauty and fashion images.


Generally, the closer an umbrella / softbox is to the subject, the softer the light is, and the farther away it is, the more directional and harder the light is. Note that the larger the umbrella / softbox is, the farther it can be moved from the subject while maintaining the soft light.

Editing flow

It is essential to follow the same editing flow in order to get consistently good photos.

Make sure you put your name in the camera's settings, so that all the photos can have a known author.

Take photos in raw format! Taking photos in raw format and then adjusting them on a computer, compared to taking photos in JPEG, is as different as day from night.

In your photo editor, import the photos from the photocamera to a "Originals" directory, in a subdirectory named "YYYYMMDD", where "YYYYMMDD" is the date when the photo was taken. Name the files "YYYYMMDD_index". Use your own development preset, created specifically for all imports, which is your starting point for editing.

Look (several times) through the photos and delete the bad ones and the redundant ones; maybe edit the photos for the basic things. Keep maximum 100 photos (between 50 and 200, depending on the intended audience) from a day's work, or more in exceptional cases. For this, you can use IrfanView because it loads the preview JPEGs from the RAWs, which is very fast.

Culling aggressively your photos will help you develop an eye for artistic detail and you'll become more demanding of your work. This also minimizes both the space taken by the photos and the time required to edit them. At the same time, they are not either too many or too few photos to show to your friends.

Edit the photos: exposure, black level, fill light, white balance (= color temperature and tint), tone curve, sharpness, noise reduction, vibrancy (or saturation), crop, selective color correction, heal skin defects.

Tag the photos with keywords; also rate them. Here are some examples of keywords: "Animals", "Landscape", "Making of photoshoot", "Models", "Objects", "People", "Print", "Vegetation", "Website"; some of them may have sub-keywords.

Export the photos in JPEG format to a "Exported" directory. Do not resize them and use ProPhotoRGB (if your display doesn't have a large gamut, use sRGB) and quality 90...100. (IrfanView can show color-managed photos.)

Export some of the photos in JPEG format to a "Print" directory. Do not resize them and use sRGB (printers have a much smaller gamut than displays) and quality 100. Print the photos on A4 paper.

Export some of the photos in JPEG format to a "Website" directory. Resize them (within 900 * 900) and use sRGB (because it's standard for the web) and quality 80.


As a starting point, for outdoor daylight photos I use the default editor settings with the following changes:

  • Tone curve: Linear.

  • Blacks: -40.

  • Shadows: 10.

  • Sharpness: 50.

  • Luminance noise reduction: 20.

  • Lens correction: by profile.

  • Vignette: -5.

Display calibration

It is crucial to calibrate the display on which you'll see and edit your photos, in the environmental light that you are going to use, at the preferred display brightness.

Turn on your display with at least 30 minutes before you try to calibrate it.

The illumination from your room (at display level, non-incident, with the display off) should be 30...60 lux. You should also have a color temperature of about 4'000...4'500 kelvin, but this looks cold in a home, so you need to have a separate lighting fixture.

Change your display's settings by using its control menu: brightness = 80...160 cd / m2, color temperature = 6'500 kelvin, gamma = 2.2.

If possible, calibrate your display with a calibrating device. Hardware calibration will not make your display better, it just makes it display images according to a colorimetric standard, in the limits of its hardware capabilities. Set the calibration target to D65 (color temperature = 6'500 kelvin, gamma = 2.2).


Do the printing on a photo printer.

A home printer provides all the necessary image quality, when the printing is performed at the maximum quality and with consumables (= paper and ink) of the highest quality. You should disable all forms of photo processing within the printer software.

I prefer to use a folder with a ring binder as a photo album, even though I have to perforate each photo by hand, because:

  • The photos are not covered by anything when you view them.

  • Photos can be grouped in categories because photos made later can be easily inserted anywhere in between.

  • It's faster than sticking photos to an album's pages, and it doesn't waste the paper used for the album's pages.

  • The photos can be easily moved to a more appropriate album (like a thicker one).

Printing the photos with a border minimizes the effect of the perforations and of the fingers (used to hold the photos while flipping them).

The advantages of my home printer are:

  • Better color match to my display.

  • Immediate results (I don't have to go out to print).

The disadvantages of my home printer are:

  • The ink is sticky and that's a real fingerprint magnet. The photos have to be kept to dry in a covered place, so that dust would not stick to the fresh ink.

The advantages of a photo-developing machine are:

  • Many paper sizes.

  • The paper can be glossy or matte.

  • About 3 times cheaper than my home printer (for the same size).

  • Outputs photos 10 times faster than my printer.

  • The photos are safe to touch immediately.

The disadvantages of a photo-developing machine are:

  • Difficult to find one which outputs proper colors, and sometimes the results are not reproducible.

Because of the price, I prefer to output my photos on a photo-developing machine.

License | Contact