Exercise: Colour Cast and White Balance

PART ONE:

For this exercise, you were first advised to familiarise yourself with the exact procedure for altering the white balance on your camera.

Find the following outdoor lighting situations, each of which has a different colour temperature:

  • Sunlight
  • Open shade on a sunny day
  • Cloudy

For each of these, choose a scene, subject or person to photograph. Take 4 versions of the image, but change the white balance options on you camera for each image. Start with auto white balance, and change it to cloudy for example.

In your processing software, note the differences, and compare the results for each scene. Note that the auto white balance will have attempted to make the best result under the circumstances, but may still be slightly different from the rest of the three others.

Sunlight:

Auto White Balance

auto

Cloudy

cloudy

Daylight/Sunlight

daylight

Shade

shade

After looking at the images, the auto white balance image is the most truthful. The image taken with cloudy white balance is extremely similar to the auto white balance image. However, the green in the grass and the blue in the sky are slightly more saturated. I am surprised that the cloudy setting is similar to the auto white balance (Even though there were clouds in the sky), mainly because when I took this photograph, I was stood in the direct sun, it was a really hot day and the sun was really shining. I expected the auto white balance image to be more similar to the daylight/sunlight white balance, because I was stood in the direct sun, however, when I looked at the images on the computer, the daylight/sunlight white balance has a blue cast to the overall image. The shade white balance image has an orange cast to the overall image.

Open shade on a sunny day:

Auto White Balance

auto

Cloudy

cloudy

Daylight/Sunlight

daylight

Shade

shade

After looking at these images, they are very similar to the images above. The auto white balance image is the most truthful and the cloudy white balance is very similar to the auto white balance. The daylight/sunlight white balance has a blue cast to it, and the shade white balance has an orange cast to it.

Cloudy:

Auto White Balance

auto

Cloud

cloud

Daylight/Sunlight

daylight

Shade

shade

When looking at these images, the auto white balance again is the one most truthful. However, I must admit that I prefer the image taken with daylight/sunlight white balance. The auto white balance image has a slight blue cast, where as the ones taken with the shade and cloud white balance have an orange cast to them. The image taken with daylight/sunlight is just right in my opinion, it has the right amount of colour saturation, and the visual colour temperature isn’t too warm or too cold.

Conclusion:

I am surprised that out of all three of the exercises, the auto white balance setting is the setting which is most true to the actual setting/subject that was photographed. I was expecting at least one or two, to be more accurate when taken with another white balance setting, however, I can see similarities in some. I am pleased with all of the images, and I think that it depends on someone’s personal preferences as to what setting is the ‘best’. All of the images can be used, it just depends on whether or not you want an image with a visual cool colour temperature, or a warm colour temperature. I sometimes prefer images with a warm colour temperature, or some with a cool colour temperature, it just depends on the subject that has been photographed.

PART TWO:

For this part, you are asked to find and shoot a mixed lighting source scene. One of the most predictably mixed is an Indoor/Outdoor scene at dusk, in which the interior is lit by incandescent lighting (Orangeish) while the exterior, under a clear sky, is Blueish. Whether you decide to shoot from the inside looking out, or the outside looking in through a window or as door, it is important to get the timing right. The point in which the light levels between indoor and outdoor are approximately the same.

Shoot 3 versions of the following white balance settings for this scene.

  • Sunlight
  • Tungsten/Incandescent
  • Auto

Compare the results.

For this part of the exercise, I decided to photograph the window in my bedroom. I set up my small bed side lamp under the window, so I could show the changes in the colour of the lamp, and the light through the window, when I change white balance. (Ps. The top part of the window is not over exposed, there is a clear sticker on the window for privacy)

Auto White Balance

Auto White Balance

Cloudy White Balance

Cloudy White Balance

Daylight White Balance

Daylight White Balance

Shade White Balance

Shade White Balance

Tungsten White Balance

Tungsten White Balance

White Flourescent White Balance

White Fluorescent White Balance

After looking at the photographs, the auto white balance is the one which is most truthful. It is very similar to the photograph taken with Shade white balance. There are similarities between Cloudy white balance and Daylight white balance. I had to look closer at the cloudy and daylight because they are similar to auto and shade, however, when I look at the colour around the window on both images, there is a blue colour cast, more on the daylight image. Whereas the auto and shade do not. Tungsten white balance is significantly different to the others. This image has a blue colour cast over all the image. It has a ‘Cool’ colour temperature. The window light has become very ‘Blueish’ and the lamp has become almost white, whereas in the other images, it is orange. The white fluorescent white balance is similar to the tungsten image, however, it has only a slight blue colour cast, mainly in the window area. The lamp is still white compared to the others that are orange, although there is a slight orange colour cast to it.

If I had to choose a favourite, I would choose the image taken with Daylight white balance. The image isn’t over powered by an orange colour cast. However, I would try to correct the parts that are underexposed.

 

 

Advertisements

Project: White Balance and Overall Colour

An essential setting on your camera is for white balance, but why is this so?

Apart from the colours that objects and surfaces have, and apart from strongly coloured lights, we expect the ambient light anywhere to seem neutral, with no particular colour. In other words, we don’t even think about the colour of the overall lighting. This is because our eyes adapt very well and quickly, to changed in the colour of the light. There are changes within sunlight depending on whether the sun is high or near the horizon, and between sunlight and tungsten lighting in your home. But our eyes and brain process the view so well that we ‘Neutralise’ the differences. The camera’s sensor however, is literal. It records all the differences in colour, and will make no adaptation. For this reason, the camera needs to be set to an appropriate colour balance.

Digital camera’s have a system similar to the human eye. This is known as the Automatic white balance setting. The red, green and blue signals are analysed for any overall dominant colour present in all parts of the image. This dominant colour is removed by varying amounts according to the colour and its strength. If the analysis is too simple, a subject with a strong dominance in one colour, such as autumn leaves or an expanse of ocean, would be wrongly white balanced so, in may systems, the highlights and shadows are also analysed and used to temper any removal or dominant colour. However, some digital cameras often have great difficulty with auto white balance (AWB). They can on occasions, create unsightly blue, orange, or even green colour casts. Understanding digital white balance can help you avoid these colour casts, thereby improving your photos under a wider range of lighting conditions.

Sunlight is our visual standard for colour. In the middle of the day, it seems colourless. We usually call it ‘White’. This is our benchmark for ‘Normal, Neutral’ lighting. Sunlight ranges from white to red in a colourful sunset or sunrise. In the shade on a clear sunny day, the light can take a blue-ish cast of the sky. This progression of colours is called the ‘Colour Temperature Scale’.

Colour temperature describes the spectrum of light which is radiated from a “blackbody” with that surface temperature. A blackbody is an object which absorbs all incident light, neither reflecting it nor allowing it to pass through. An example of this is heating Iron. When you heat Iron, it will first become red, then yellow, then white. Also known as White hot. At even higher temperatures, some materials can turn blue, then melt or vaporise.

Colour temperature is a useful description of light for photographers, even if we never deal with true blackbodies. Fortunately, light sources such as daylight and tungsten bulbs closely mimic the distribution of light created by blackbodies, although others such as fluorescent and most commercial lighting depart from blackbodies significantly.

With photography, the most important colour temperature is that of the sun. On this scale, it can be given a value, in degrees of temperature. The temperature is measured in degrees Kelvin. Similar to Celsius, but starting at the lowest possible temperature called ‘Absolute Zero’.

Noon sunlight is generally accepted to be between

5400K – 5500K

When the sun becomes lower in the sky in the late afternoon, the colour temperature also lowers to approx. 4000K.

At sunset, depending on the sky conditions, it can be approx. 3000K.

The reflected light from a blue sky ( In other words, in the shade ) it can be higher than 6500K.

There may be some confusion, as it is natural to describe red colours as ‘Warm’ and blue colours as ‘Cool’, whereas, the colour temperatures are in fact the opposite. Be aware when you are talking about the visual effect and not the colour temperature.

Since photographers never use the term ‘colour temperature’ to refer to a true blackbody light source, the term is implied to be a “correlated colour temperature” with a similarly coloured blackbody. The following table is a rule-of-thumb guide to the correlated colour temperature of some common light sources:

Colour Temperature Light Source
1000-2000 K  Candlelight
2500-3500 K  Tungsten Bulb (household variety)
3000-4000 K  Sunrise/Sunset (clear sky)
4000-5000 K  Fluorescent Lamps
5000-5500 K  Electronic Flash
5000-6500 K  Daylight with Clear Sky (sun overhead)
6500-8000 K  Moderately Overcast Sky
9000-10000 K  Shade or Heavily Overcast Sky

 

There are ways of adjusting the colour balance on your camera.

  • Auto White Balance: Using auto white balance means letting your camera’s processor to sort out the whites for itself.
  • Pre-determined Setting: Using this setting means that you have to judge for yourself, what white balance is needed for a specific scene. You have the choice to choose between Incandescent, Cloudy, Shade, Fluorescent, Tungsten.
  • WB Preset: With this setting, you help your camera measure a white patch ( Example, White piece of card )

 

Auto White Balance
Custom
Kelvin
Tungsten
Fluorescent
Daylight
Flash
Cloudy
Shade

Below is an example of different White Balance settings, taken from the website Digital-Photography-School.com.

 

References:  

http://www.cambridgeincolour.com/tutorials/camera-sensors.htm

http://www.digital-photography-school.com/introduction-to-white-balance/

Digital Photography Master Class, Tom Ang, Dorling Kindersley Limited, London 2008. Page 169-169.

ISBN: 9781409333906

Exercise: Scene Dynamic Range

With some idea of what range of brightness your camera can cope with, measure the dynamic range of various scenes.

Take five differently lit scenes, and within each, find and measure the brightest and darkest areas. Make sure the picture situation varies!!!

Make sure that one of the scenes has a very high dynamic range, and another has a low dynamic range, in other words, appears flat.

Other than this, there are no limits to the scenes you choose, because an important part of this exercise is exploration and discovery.

WILL BE UPDATED SOON

Exercise: Your camera’s dynamic range

For this exercise, you were asked to find a scene with a high dynamic range. The best conditions for this are

  • Bright Sunlight
  • One bright reflecting surface
  • An area of deep shadow with a dark surface

A suggestion was to use a A3 size piece of white paper as your highlight. Take the front of your house in the sunlight, and open the front door in order to case a black, dark shadow area. Pin the white paper somewhere near the door, or in the direct sunlight.

Make sure the ISO sensitivity is at its lowest, turn any noise reduction on your camera off. Set up the exposure and shoot so that there is just no highlight clipping of the white card, or white area.

Next, measure and make notes of the brightness of the white area and two or three of the darkest shadow areas. There are three ways of doing this.

  • Set the metering mode to ‘Spot’ if that is available on your camera
  • If you are using a zoom lens, set it at wide for the photograph, zoom into the maximum to measure the small areas
  • Walk close to the areas you want measure so that they fill the frame

Make a note of the aperture/shutter speed combinations, and note exactly which area you measured.

Open the image in your processing software. Zoom into 100% magnification. Move to the white area. Using the software’s pixel value sampler ( It shows values from 0 to 255 in the three channels ), check that the value of the white is only a little less than 255 in each channel.

Now move to the shadows, and adjust the brightness until the details are visible. You should now look for an area in which the level of apparently real detail and noise are competing with each other, the shadow are in which you have difficulty telling which is which.

Now calculate the range between the two ends in F-stops. You will need to be familiar with the normal lens aperture notation (F-2, F-2.8, F-4, F-5.6, F-8) Etc. Marking One-stop differences.

For this exercise, I decided to go to a place called Snuff Mills. It is a large forest with a house situated towards the entrance area. I chose a really sunny day for this exercise as I wanted the white house to be ‘bright’. I knew that there would be dark shadow areas, especially in the trees behind the house, which would be perfect for this exercise.

I set the exposure so there was no highlight clipping on the white house, I also kept the ISO low. The settings I used were, ISO 100, F/6.3, 1/400. After I took the first photograph, I used the aperture priority setting which meant I could keep the ISO 100 and the F/6.3, but the shutter speed would change depending on where I zoomed into. I used the ‘Spot’ metering on my camera.

IMG_4635 - Copy

Below is a copy of the original image. I have noted the spots where I zoomed in, and what the shutter speed/aperture readings were. The darkest area was 1/15sec and the brightest area was 1/800sec.

fgfg

I then had to calculate my camera’s dynamic range. I have to work out the number of F-stops between the darkest area and the lightest area. 1/15sec to 1/800sec. I may be wrong, but I calculate it to be 6 F-stops difference.

Conclusion:

I must admit that I have struggled with this exercise. It took me several failed attempts to do this exercise, as I was getting all of the settings wrong on my camera, which obviously frustrated me, so therefore I stopped doing the exercise, but I knew I had to complete it. I did some more research and finally began to understand how to do the exercise, which then led me to going to Snuff Mills in order to complete it. However, when it came to ‘Measuring’ the dynamic range on my camera, I find this quite hard. I personally think I may have measured it wrong, as reading about dynamic ranges on camera’s, I think my measurement is quite low, especially when dynamic ranges on cameras should be roughly 9 stops. It may just be where I zoomed into and measured, I’m not sure. I am hoping that the next exercise will help me understand this more as I am still quite confused.

 

 

 

 

 

 

Project: Dynamic Range

In the last couple of exercises, we looked at what are generally considered the two most pressing technical issues in exposing, for a digital photograph, highlight clipping and noise. They are at opposite ends of the tonal scale, and they define what is known as the ‘Dynamic Range’ of a camera. In earlier days, the range between dark and bright was often called the contrast range, however, it is now known as dynamic range, as this is more accurate.

There are several ways of measuring the dynamic range, however, the most easiest for a photographer is in the form of F-stops. The dynamic range of a scene, is therefore the number of F-stops between the brightest highlight, and the darkest shadow. The dynamic range of a camera is  the number of F-stops it can capture in one exposed frame. If the dynamic range of your camera is greater than the scene you are about to photograph, all is well and there should be no problem capturing the visible detail. If, on the other hand, the camera’s range is less than that of the scene, something has to be lost. This is the cause of most exposure problems.

As the dynamic range is between darkest and brightest that can be captured, we need to find the end points. The easier of the two is the brightest. Using the highlight clipping warning, we are able to find the exposure which captures the brightest point. Noise determines the darkest tone that can be captured. If you look in the very dark shadows of an image, where most of the noise is, there is a point at which it is impossible to distinguish between noise and real detail. In order to see this clearly, it is necessary to lighten the image temporarily in Photoshop. There are however, disagreements over this, which explains why different dynamic ranges appear for the same camera, depending on who is deciding.

 

 

Exercise: Your Tolerance for Noise

For this exercise, you were asked to find a situation that fulfils the following criteria.

  • Daylight indoors ( For the amount of light – outdoor sunlight would be too bright to allow high ISO settings, while much darker would involve long time exposures)
  • A combination of sharp detail and texture less areas ( Such as a white wall ) with some of the texture less area in shadow.

Set the camera on a tripod.

Start by taking a series of identical photographs, changing the ISO setting from one to the next.

Use the aperture priority setting so that there will be no difference in depth of field. My camera was set to F.4 and the ISO ranged from 100-6400

I decided to use the morning light, and photograph my wardrobe door, as it is white, and situated right next to my bedroom window. I stuck a black chandelier sticker onto it, as this was detailed and would help with showing noise.

I have included the original image, and a cropped section where I have zoomed in, to show the noise.

ISO 100:

Clear Image

ISO 100 (1)   ISO 100 (2)

ISO 200:

Clear Image

ISO 200 (1)   ISO 200 (2)

ISO 400:

A clear image, but a small amount of noise has been introduced.

ISO 400 (1)    ISO 400 (2)

ISO 800:

Noticeable noise, especially towards the top of the image where there is some shadow. You can begin to see noise in the dark shadow areas also.

ISO 800 (1)   ISO 800 (2)

 

ISO 1600:

Noticeable noise all over the image

ISO 1600 (1)    ISO 1600 (2)

 

ISO 3200:

Noticeable noise all over the image. More noticeable towards the top and in the shadow area.

ISO 3200 (1)   ISO 3200 (2)

 

ISO 6400:

Noticeable noise all over the image.

ISO 6400 (1)   ISO 6400 (2)

 

I will put the cropped images next to the ISO 100 and then I can compare them easier.

                               ISO 100                                                                           ISO 200

ISO 100 (2)    ISO 200 (2)

                                 ISO 100                                                                       ISO 400

ISO 100 (2)    ISO 400 (2)

                                  ISO 100                                                                       ISO 800

ISO 100 (2)      ISO 800 (2)

                                 ISO 100                                                                        ISO 1600

ISO 100 (2)      ISO 1600 (2)

                               ISO 100                                                                           ISO 3200

ISO 100 (2)     ISO 3200 (2)

                              ISO 100                                                                          ISO 6400

ISO 100 (2)    ISO 6400 (2)

When looking at the images magnified, I can see that with ISO 100 and ISO 200, the images are clear. With ISO 400, graininess begins appearing in the top area of the image, however, it is still usable as the detail on the sticker itself if clear. When I got to ISO 800, there is noticeable, yet faint, noise all over the image. The detail on the sticker is becoming less sharp. ISO 1600, the noise is increasing. With ISO 3200 and ISO 6400, there is significant noise all over the image, but it is worse towards the top of the image, where there seems to be a shadow, as the noise is more noticeable there. The detail on the sticker is blurred and not sharp. These images would be unusable. In my opinion, the best ISO to use in order to have less noise in your image is between ISO 100 and ISO 400. Any ISO after 400, you will begin to see noise in your image, even if you use a fast shutter speed.

 

 

 

 

Project: Noise

Noise is an imaging artefact, an error of capture that has nothing to do with the actual scene being photographed. It resembles graininess in film, but its causes are different. It increases with higher sensitivity settings, just as graininess is stronger in high speed films. Caused by many factors, including heat, the nature of the electrical signal and activity in the sensor itself, the outcome of noise is always the same, irregular. Noise usually appears as speckles, which can be black, coloured, or bright, depending on the scene and the exposure. Noise is generally more evident in lower mid tones and shadows, and in areas of even tone.

There are two situations where you are likely to find noticeable noise. The first is when you chose a high sensitivity setting, and the second is when you make a long exposure. When we increase the sensitivity, by using a higher ISO, we are amplifying the signal to the sensor, this process tends to increase noise further by making it more visible.

Several things contribute to noise, but the major cause of it is when then there are not enough photons of light striking the receptors in the sensor, this then causes a sampling error.

Because noise tends to mask detail, it is generally regarded as a flaw that should be avoided or reduced, however, it does have value, provided we introduce it only when and to the extent that it’s needed. Noise can help give texture to high resolution images. The important point to remember about digital noise, is that you have to decide which structures in the image are real detail, and which are noise.

There are different kinds of texture in the garment, folds, vertical ribbing and mottling. Think about the mottling on clothing. Is this noise, or is this part of the pattern? it could be either. Distinguishing detail from noise is ultimately subjective.

 

References:

Digital Photography Master Class, Tom Ang, Dorling Kindersley Limited, London 2008. Page 18

ISBN: 9781409333906