I have used Raw as my preferred shooting mode for a few years and use Lightroom to apply WB settings post-shoot. This is perhaps a rather lazy approach but has suited me until now. However the 2nd assignment of the course requires all the images to be taken as large size jpeg's and I realise I have become perhaps less conscientious when shooting and more reliant on  post-processing. Hence I need to get into the habit of adjusting the WB setting on my camera pre-shoot as the wrong choice might spoil an otherwise  acceptable image. As Kelvin  temperature increases light changes from red to yellow to blue, a colour usually  associated with coolness . The blue light of the higher Kelvin temperatures creates redder tones, whilst the lower temperature of yellow light will  produce a bluer hue. So when describing the overall appearance of an image as cool or warm I am not talking about the actual Kelvin temperature but the  effect the light has on the subject as I see it. 


Exercise Colour cast and white balance

Cloudy day with a grey overcast sky

Auto WB : Colour temperature K5100 when using my camera's auto WB the resulting image has the same colour temperature as the Daylight WB.  My subject's skin looks pale and rather washed out with a blue tinge , not flattering at all ! The brickwork behind is a dull  orange. 
Daylight WB : Colour temperature K5100. it was a cloudy day  I was expecting this WB setting to improve on the camera's Auto WB. Again the skin tones have little natual looking colour. However I do feel my subject looks a bit  less washed out and the image is a slight improvement than that produced by the Auto WB .
 Cloudy WB : Colour temperature K5800 setting produced a much more   natural image with flattering skin tones, with still quite dull but brighter brickwork behind. An improvement on the Auto and Daylight WB settings.
Shade WB: Colour temperature 6600 this is a much warmer looking image I personally feel the skin tones ,especially around the hairline, are too orange. 

Open shade on a sunny day

The next two images below were taken within minutes of each other on a clear bright sunny day with my subject in the shade of an old tree trunk. The top image using a shade WB setting has enhanced  the tones of her skin and the surrounding area , the hue of her pink top has become deeper and much more saturated. In comparison a Daylight WB setting has caused the camera to record pale rather washed out tones , they are not  flattering to my subject , or record the scene as I saw it. Light is coloured, changing throughout the day, however whilst the human eye easily adjusts to these colour differences the sensor cannot. It simply records what is sees and is unable to make the same allowances the human brain is capable of .

I decided to use two more  subjects for my open shade images,an orange haired rag doll and a blue pottery owl. I was interested to see how the cool tones of the owl compared with the warmer tones of the doll depending on the WB chosen. 


The auto WB has not done a bad job for either image but I think the colours are enhanced more by the choice of shade or even the cloudy WB settings. They become deeper and richer in tone, some of the richness of the colour is lost using an auto or daylight WB.

Daylight WB
Shade WB 
Both images now have a much warmer appearance with deep tones.
Cloudy WB
Not bad but I actually prefer the deeper richer tones created by the Shade WB.

Sunny day

This being Britain I had to be patient and wait for a sunny day to take these images ! I used Old Ted (he's as old as me and I am not saying how old that is !!!) . He makes a very patient subject who never complains,I use him frequently for photo projects. With the exception of the Auto WB any of these settings would be acceptable because of the subject I chose.  Although it was sunny the Cloud and Shade WB settings enhance Ted's fur , making the tones deep and rich. The WB setting is a useful way of controlling the final image and not rely on the camera's Auto setting. 

Auto WB . Ted is too pale.
Daylight WB. 
Cloudy WB
Shade WB
Mixed light
Three images shot from outside within a few seconds of each other looking back indoors at dusk on a very clear evening.  
Viewed large the Tungsten WB setting has produced the most neutral toned natural looking  image . The walls indoors are whiter than those rendered by the  AWB and Daylight WB settings which make the light indoors very unnaturally orange toned . However the warm glow from indoors looks welcoming , using the wrong WB is not always such a disaster, it really depend on the wanted outcome !

Tungsten WB                  AWB                           Daylight 

Conclusion.
Setting the correct WB setting when shooting Jpeg is crucial to creating better looking images, however used imaginatively  it can also be used creatively. WB is easily remedied when shooting Raw if a mistake is made at the time of shooting. 
 
For this exercise I need to find 5 differently lit scenes measuring the light to find the darkest and lightest areas.

The image below was taken indoors with my subject looking out of a brightly lit window.
 The dynamic range was 6 stops ranging from1/30 up to 1/1600. 
 Image taken @  1/800     F3     ISO 100   50mm If I had wanted I could have chosen to expose @ 1/1600 retaining more detail and colour in the sky making my subject a silhouette. Its still  quite a dark image but I like that, I feel if I had made it brighter by increasing  the exposure time it would be less atmospheric. 
  

 

 For this exercise I needed to find a scene that included a bright surface and dark  shadows in bright sunlight. All noise reduction was turned off and the ISO was set at 100 as instructed. My camera's evaluative metering gave an exposure reading of 1/250 @ F6.3 , in order to prevent the highlights being clipped I needed to underexpose and adjust this to 1/400 @ F6.3.Using spot metering I then measured 5 areas within the frame all @ F6.3 ISO 100
The white area of the card  1/1250                                             
Fence at right  hand side 1/80                                    
 Area directly behind the card 1/25                                                         
Area around the bottle (difficult due to reflectance) 1/15               
 Floor area underneath the bench 1/5                                                  
This gives a  dynamic range of  8 stops       
 
For the next  part of the exercise I needed to use the pixel sampler in PS to check the value of the white card to ensure it was just slightly less than 255, which it was, and finally look at the shadow areas to check where digital noise and detail became indistinguishable, I found this bit rather confusing. I must admit I initially found this exercise difficult to comprehend but after completing it have come to the conclusion its purpose was to make me realise that in a difficult lighting situation that is beyond the dynamic range capabilities of my camera some sort of compromise needs to be made. If the scene is within the 8 stops range of my camera I can be reasonably sure I will be able to capture the whole range of tones without highlight clipping or digital noise. Beyond this range a decision would need to be made whether to under or overexpose, this then relies on an aesthetic judgment, and becomes a much more  personal choice. Michael Freemans book Perfect Exposure: The Professional Guide to Capturing Perfect Digital Photographs  has a section dedicated to this he comments " In exposure there is no wrong and there is no right----Like any artist you have to stand by your own  opinion. Not everyone will agree with what you do, but does that matter ?"  Pg115. Another skill he suggests is well  worth trying to develop is "the  simple ability to look at a scene , see blocks of similar brightness, know intuitively what that brightness is , and how that translates into stops "Chapter 4:Style Pg 16. With this sort of  knowledge the actual dynamic range of a scene becomes less relevant than the ability to recognise that it is beyond the camera's limits to capture the full tonal range and understanding how to deal with. This  exercise has been an eye-opener for me as I was totally ignorant about the dynamic range of my camera and its capabilities even though I consider myself quite proficient with the technicalities of photography- I always manually expose, usually with reasonable results. Since completing this exercise I have bought a Sekonic light meter (see reading section March 2012) as perhaps the most important factor to successful photography is learning how to use light  effectively and creatively and I hope this meter will help me in tricky light situations to make a successful decision. 

http://en.wikipedia.org/wiki/Dynamic_range
 
I had a look through my "archives" and found this image taken at dusk using an ISO of 1600 which demonstrates just how digital noise using my camera at higher ISO levels creates ugly speckles . Through past experience I cannot image a situation where I would choose to use above an ISO of 1000 , much less ISO 3200 !!!
Exercise: Your tolerance for noise
 

Project: Highlight clipping


My camera's  highlight clipping warning is always left on.I used some old teddies propped against a white textured wall for this exercise.


Exercise Highlight clipping
Using an aperture of F1.8 and ISO 200 the correct metered exposure was 1/40secs.The highlight clipping warning just appeared at 1/20secs.My five images were taken @ 1/20  1/15   1/25 1/30  and 1/40. At 1/15secs the overexposed white areas of the wall have become featureless and there is very visible loss of texture, however it has become, lighter, brighter, and much less grey looking .To retain the whiteness of the walls it would be necessary to overexpose but this is at the expense of the visual detail lost in the clipped highlight areas around the pink and white teddy's head. I can see that there is no definition between the whitish fur and background, it has been totally lost. The pink fur on the left hand side of the face has become so pale the colour is indistinguishable.

Picture
@ 1/15 secs Highlight clipping indicated by the red areas.
Not really a problem for the white wall behind.


Picture
But the close up of the bear demonstrates just how much detail is lost due to the highlights being clipped.

Picture
As the images were shot in Raw I can attempt to recover this lost information using the Recovery slider in Lightroom. The more I increased the recovery slider the stranger the image looked to me. The tones looked somehow un-natural, I felt the whole image was degraded.

Trying to find an acceptable compromise to save the detail lost in the clipped areas I moved the slider about a third of the way across, this is not a perfect solution however the pink tones have now become visible again and some detail has been saved, but avoiding badly blown highlights in the first place is the best practice.


Picture
 Compromise
 Recovery slider used to bring back some of the lost information.

What have I learned ?
The image as initially taken (see below) at first glance looks acceptable, I wanted the backlight to outline my subject  and was aware at the time the highlight clipping warning was flashing. As the tattooed  back was the intended main tone I was not too concerned if the sky detail became blown, it was not especially important for this image. However using what I  have learned from this exercise I looked closer at the clipped highlighted areas. The sky , hair and shoulder are all overexposed, but the area that suffers most is the skin area around the shoulder. Looking closely there is total loss of skin texture and detail , as well as a highly visible break where the highlights become featureless white. Additionally the fringed area at the edges of the clipping, before it becomes featureless white, is rather strange  to look at, its hard to describe but the texture looks mottled which in turn  creates a rather strange appearance to the surrounding tones. Prior to doing  the exercise although I was well aware to avoid highlight clipping in retrospect I was not really knowledgeble enough to consider and take into account just how detrimental this can be to the quality of an image. Luckily this image was shot in Raw format so I was able to process and recover some of the lost detail in Lightroom.  Assignment 2 has to be shot in jpeg format with no post-processing , I will need to careful to remember what I have discovered and ensure I evaluate what is important in the frame and try to ensure I capture the best  quality exposure I can. However even when shooting Raw using this knowledge will save valuable time by reducing any post- processing work needed, getting the best image in camera should be my aim.

Picture
Original Raw image

Picture
Clipped areas in red

Picture
Shoulder shows loss of texture and detail

Picture
Auto tone recovery @ 29%

Picture
Recovery slider @ 100%

Picture
Recovery slider @ 50%

Final image. 

 
Exercise:  Sensor linear capture 
Prior to starting this exercise I read of a fellow DPP student's problem understanding the expected outcome of the exercise on the OCA Flickr group. I am glad I was made  aware the coursework text is incorrect as I too would have thought I was  totally dim !  http://www.flickr.com/groups/ocarts/discuss/72157628955228629/ 

Digital sensors respond to light in a different way to film and also how our own eyes react. The sensor responds to light in a linear, proportional, way and then performs a series of in-camera actions. Unlike the human eye it is unable to distinguish detail in the shadows and highlighted areas , this information is lost.
This exercise  examines what an image looks like prior to the in-camera process and how the camera performs gamma correction curves to alter an image and make it appear more customary. I used a 16 bit Tiff as instructed (on my old PC using Elements not PS) and created a new darker image . 

Picture
Original image

Picture
 Using a curve adjustment to demonstrate what the image looked like at time of capture. The altered image is now much darker , the colours are flat and dull.

Picture
The histogram indicates a dark image and shows how it looks prior to in-camera processing.

Picture
The original file saved as a new "linear" version: showing how the camera sees. 

Picture
Using a curve adjustment to demonstrate the gamma correction performed in-camera.The altered image has now become brighter. 

Picture
The histogram demonstrates how in-camera processing alters tonal values. 

I found the image emulating the in-camera gamma correction process greatly increased digital noise, something I hate. 
How I understand  the exercise, and what it means in terms of creativity, is that the sensor  records lines of light (linear capture). I need to be able to pre-visualise a scene as the camera sees it , ignore  what I see, and use this knowledge to gain control over the final image. The  projects in the remainder of Part 2 look at ways of how this may be achieved.