ASSIGNMENT 2: Digital image qualities

The aim of this assignment was to produce around six images on a common theme in order to explore the different lighting conditions on the image quality produced. These images were to include -


1. A high dynamic range (high contrast) scene. This means that the image should contain a range of tones from the very brightest to the very darkest, without being under or over-exposed i.e. producing image 'noise' at the darkest end or highlight 'clipping' at the brightest. As noted in the course notes for Exercise 8, a high dynamic range image needs, ideally, 


                     - bright sunlight
                     - at least one brightly reflecting surface
                     - an area of deep shadow with a dark surface




High Contrast Image 1
As an example, I took this pre-sunrise shot on a recent holiday in Northern Spain. The un-adjusted Raw image is shown below. ISO was set to 100 (its lowest setting on the Canon 50D) and the noise reduction function set to OFF (its default setting).

The challenge here was to capture the very bright pre-dawn sky whilst showing some detail in the dark foreground. Using a tripod and aperture priority setting the camera settings were as follows. f/11, shutter speed 0.6 secs, ISO 100, AWB and exposure compensation of -1 to eliminate highlight clipping (as shown by the absence of any clipping warning on the camera LCD).


The image was subsequently processed using the Lightroom Raw converter and then Photoshop to give the following image. 

The adjustments made were as follows. 


In Lightroom - increase Contrast to +86 to make the image punchier, increase Recovery to +63 to help recover any highlight problem, reduce Blacks to zero to give more shadow detail, increase Clarity to +75 to add some sharpness, slight increase of Vibrance to +22 to hep the sky tone and, finally, Sharpening to +50.


In Photoshop, Image>Adjustments>Levels to make the range 12 to 249 with a mid-tone adjustment to 1.04 to slightly lighten them. Finally, I converted the image to sRGB Profile for web posting (as with all photographs for this assignment).


The overall effect has been to give the image more contrast and to bring out some of the detail on the hills. Were I to take the shot again, I think that I would have used a narrower aperture (f/16?) to improve the sharpness on the horizon.




High Contrast Image 2
Another example of a high contrast shot was one taken indoors of a stained glass window, the contrast being between the brightness of the light through the window panels and the dark wood of the window-frames. The photograph was taken in the northern Spanish town of Comillas in a house, El Capricho, designed by Antonio Gaudí.


The un-adjusted shot from the camera is as follows.

As I used an exposure compensation of -1 indoors the result is a dark image, but without highlight clipping. Other settings were f/11, shutter speed 1/60 secs, ISO 400 (handheld) at 30mm (with 18 - 270mm zoom lens). I used 'Daylight' WB.


The Raw image was adjusted in Lightroom and then Photoshop to give the following result.

The Raw converter adjustments were Contrast to +70, Recovery to +76, Fill Light to +83, Clarity to +66, Vibrance to +36 and Sharpening to +60. I also made a vertical lens distortion correction of - 31 and then cropped the image as shown.


In Photoshop, Image>Adjustment>Levels to slightly adjust the range to 4 - 248 with a mid-tone adjustment to 1.11.


The overall effect produces a much more even image. Perhaps using a higher ISO setting would have been better although I am not convinced.




2. A low-dynamic range (low-contrast, 'flat') scene. This would be a daylight shot but, perhaps, in murky conditions with the subject matter comprising mid-tones so that there are little, if any, of the darkest or brightest tones available to be captured.


Such a shot was taken around the old church in San Vicente de la Barquera in northern Spain. Although it was a sunny day, this side of the church was in shade. The photograph was hand-held (ISO 400), f/11, shutter speed 1/100sec at 32 mm (18-270mm lens). I used a WB setting of 'Shade' and an exposure compensation of -2/3. The histogram indicated that there were very few pixels at the dark end and none anywhere near the bright end.

The un processed image (above) appears very flat compared with the processed image below.

This was achieved by increasing the contrast and using fill-light. The Raw converter settings used were Fill Light + 52, Contrast + 78, Clarity + 61, Vibrance + 22 and Sharpening + 60. In Photoshop, Image>Adjustments>Levels a range of 10 - 183 with midtones at 0.83.


A higher ISO setting could have been used in camera but I was confident of being able to produce a good final image at ISO 400.


3. Low-light conditions which require a high ISO setting.


An example of this is the following image taken on the near-by canal which I offer, not because it results in a good final image, but because of the interesting issues that it raises. The un-processed image is as follows

As can be appreciated by the dark image, the sky was very bright. In order to show some detail in the sky and avoid total 'white-out', I used an ISO of 1250 and focused on the sky before framing the shot. Even with this high ISO, focusing elsewhere in the frame resulted in a complete loss of definition in the sky even after processing in Lightroom and Photoshop. The settings details were f/10, shutter speed 1/4000th sec at 11mm using an 11 - 16mm wide-angle lens, exposure compensation -1 and 'Cloudy' WB setting.


After processing, the resulting image certainly showed plenty of definition in the sky but there was considerable 'noise' in a great part of the frame. Interestingly, this 'noise' was not present when I had taken the same shot (with identical settings) after focusing on the underside of the bridge, rather than the sky. However, in the latter case, it was impossible to retrieve any detail in the sky. 

The final image was  achieved with the following Lightroom settings. Exposure +60, Recovery +41, Fill Light +94, Contrast +76, Clarity +56 and Sharpening +57.


In Photoshop, Image>Adjustment>Levels to give a range of 3 - 255 with a mid-tone reading of 1.11.


I think that this final image is an extreme example of the noise problem when using a high ISO setting. Perhaps the answer would be to take several images at different exposures and to then combine them using HDR techniques (which I have yet to try!).


4. Mixed lighting conditions (e.g. daylight + incandescent, fluorescent + incandescent)




Mixed Lighting Image 1
A dusk shot in Victoria Square, Birmingham combines evening sky with lighting both on the Council Offices and within the water feature.


The un-processed shot, using a tripod and 5 second exposure, is as follows.

Apart from ensuring that there was sufficient light in the sky to define all the buildings, I used a tripod to obtain a long exposure in order to give a smooth appearance to the water and provide shape to the water cascading from the fountain. Live View indicated that there were no great highlighting issues, apart from those resulting from the bright lights in the pool, which I judged to be acceptable.


The processed image is as shown below.

Apart from a little straightening of the image and removal of a couple of cranes showing above the Town Hall building on the left, I have made the image a little punchier by increasing the contrast and clarity. Despite using the Recovery slider, the histogram still showed highlight clipping but I presume that this was from the pool lights. 


The Lightroom adjustment settings were Recovery +52, Fill Light +6, Blacks 0, Contrast =$&, Clarity +63, Vibrance +20 and Sharpening +50. In Photoshop, Image>Adjustments>Levels to a range of 8 - 255.




Mixed Lighting Image 2
Another example of 'mixed lighting' is a photograph taken indoors at Bolsover Castle in Derbyshire. It is a stairway lit by both daylight from a window on the right (and one behind) and a fluorescent light on the ceiling (above left) which provides detail on the ceiling. The un-processed image is shown below.

The photograph was taken using a tripod, which enabled a longish (0.3 seconds) exposure at ISO 100, using a wide angle lens (11 - 16mm). This avoided the disadvantages of using a high ISO. Other camera settings were f/11 and Auto WB.


The processed image is as follows.

By increasing the exposure setting in Lightroom and using some Fill Light, the image is improved greatly (without the 'noise' that a high ISO may have introduced by this increase in the Exposure setting). The Lightroom adjusted settings were: Exposure +51, Fill Light +36, Blacks 0, Contrast +86, Vibrance +56 and Sharpening at +50. WB was adjusted to 4400.


Again, I think a key is the use of the tripod and, in the latter case, framing the shot so that the fluorescent light was not in the image.






Conclusion


I think that I need to do some more experimenting with high ISO in circumstances where no tripod is available. 

Exercise 5: Sensor Linear Capture

As required, I made an adjustment to a tiff image within Photoshop (CS5), using Curves to simulate a linear image. On examining the new, darker, linear image alongside the original tiff image, it was clear that the histograms were different although, apparently, not as stated in the course material. For the darker image, the tonal values were concentrated strongly to the left (not the right as stated) whereas the original tiff image has a more 'normal' distribution of values. I don't understand this discrepancy.


Applying another Curve to the darker, linear, image, enabled the image to return to that of the original.


The exercise simulates clearly the process carried out within camera but I must be missing something re the histograms.

ASSIGNMENT 1: Workflow

For this first Assignment, we were required to construct and use a 'sensible workflow' to produce a final selection of between 6 and 12 photographs from a much larger number.


I decided to use, as my theme, the new Museum of Liverpool and its setting at the Pier Head alongside the famous 'Three Graces' - the Royal Liver Building, the Cunard Building and the Port of Liverpool building. When the museum was opened in July of this year, its design and location were heavily criticised so I wanted to see it for myself and make my own judgement. 


This is certainly no Bilbao Guggenheim. The museum website, www.liverpoolmuseums.org.uk , states that the design concept was developed in conjunction with a Danish architect and that "Manchester based architect AEW were later commissioned to deliver the detailed design". Could this be an example of Manchester putting one over on its Merseyside rivals?


Externally, I endeavoured to take a number of photographs which juxtaposed the new museum with one, or more, of the 3 Graces and other waterside structures around the Albert Dock area in order to demonstrate the complete difference in style. I wanted a few 'establishing' shots to show the museum in its wider setting and then some detail shots with various angles and particular features.


Internally, it was quite a challenge to find suitable shots as there seemed to be very little of architectural merit. The exhibits were closely packed and it didn't help that there were a lot of people there when I visited.


I spent about an hour taking shots both externally and internally using my 50mm Canon prime lens. All shots were taken in Raw, in aperture priority mode, at f/11 and ISO 400. I did not use a tripod on this occasion.




 In all, I took 54 photographs. Having downloaded them from my card reader into Lightroom 3 as .dng files, I then proceeded to edit them as follows.


Step 1 - Technical Edit


Using the Loupe view within Lightroom, I rejected only two of the shots for purely technical reasons (out of focus). Some of the remaining shots were under/over exposed but I felt they could be retrieved during the Raw processing procedure, if chosen at a later stage.


Step 2 - the selects


Again using the Loupe view, I reviewed the remaining 52 photographs and discarded   
16 of them, generally because the composition was unsatisfactory or I had taken several different shots of the same aspect and decided on those that I preferred at this stage. The 36 remaining were flagged as 'picks'.




Step 3 - the first selects


Again using the Loupe view, I worked within the 'selects' and reduced the 36 down to a more manageable 17 'first selects'. This selection was done by trying to obtain the best (or potentially best) compositions and by choosing my preferred of almost identical images in a sequence.


Step 4 - group and review


Having arrived at the 17 'first selects', I reviewed my choice in Lightroom using both the Loupe view and Grid view, and was satisfied with my decisions.


Step 5 - the final triage


In order to arrive at the final selection of photographs to submit for this Assignment, I carried out a stage in the editing process which is a variation, and addition, to that suggested in the course materials. As discussed in an earlier blog entry in relation to Exercise 1 of this course, as I shoot in Raw, I would make my final choice only after I had processed the candidate shots, as the Raw processing itself can significantly affect this choice. This wouldn't be so much of the case if I shot in jpeg although, even here, certain processing decisions e.g. cropping, can make a big difference to the final image.


Therefore, each of my 17 'first selects' was processed, first in Lightroom and then edited in Photoshop CS5. In Lightroom, I cropped and straightened the images where necessary/appropriate and then moved to the 'Basic' section to make adjustments to exposure, contrast, recovery, fill-light, clarity etc. as required. After sharpening, the image was then edited in CS5 for levels adjustments, removing extraneous detail, converting to sRGB from Adobe RGB before saving as jpegs. 


Only after this processing was I able to compare the 17 jpegs and choose 9 photographs for my depiction of the Museum of Liverpool, as shown below.

 An establishing shot

 Front entrance

 Liverbirds and Cunard Building

Overlooking the Pier Head

 Corners

 Reflections of the Mersey

Liverbirds from within

 Atrium Dome

Foyer Carpet and Customer from above

Conclusions

The above work flow worked well for me, especially with the adaptation of processing the Raw 'first select' images before the final triage. An enjoyable assignment.

Exercise 4: Editing

For this exercise I decided to take some action photographs of our two grandsons playing in the courtyard of our apartment block. I took over 70 photographs, sometimes using the 'continuous shooting' setting, and then edited them down to my six favourites, from which I chose two as requested.






Selected Photographs

Nearly - selected 4 Photographs

Observations


Using the editing process suggested in the course material was very useful and really codified the way that I have started to edit my photographs. Using 'flags' and 'stars' within Lightroom to reduce the shots to a manageable number is well worth doing although I must admit to being a bit lazy about this in the past.


As mentioned in the blog for Exercise 1, it has been my practice to Raw - process my 'first select' photographs into jpegs before making the final 'triage' selection and I will continue with this practice as it fits in well with the method outlined in the course notes.

Exercise 1: Your own workflow - 1

Although it is suggested that this exercise be a portrait shoot, I decided to do something else whilst still maintaining the  principal purpose of understanding my workflow by "taking a number of images, all in a concentrated period of time, and which will need to be edited in order to select one or two". 


My chosen subject was 'canal boats', which is appropriate as I live in a canalside location.


As required, I planned my workflow in advance, as follows.


1. Choose a subject and location. This was done as above.


2. Set time period and number of shots required. I thought that 30 minutes would be appropriate for taking 20 shots.


3. Decide on equipment. The Canon 50D was a given ( as it was a digital camera) and I chose my Tamron 18mm to 270mm lens for maximum flexibility.


4. Choose mode/white balance/exposure. Aperture Priority mode seemed most appropriate and f/11 to give sufficient depth of field. Although setting the white balance isn't as critical when using Raw as it can be adjusted while processing, I chose the 'cloudy' setting.


5. Take photographs. Check histogram, sharpness and composition after each shot. Adjust exposure compensation as appropriate. After a couple of shots, I decided on an exposure compensation setting of - 1/3. I re-took two shots because I was unhappy with the composition. 


6. Back home, check shots in camera and delete any which are obviously sub-standard. I deleted the two shots identified during the shoot.


7. Download remaining shots onto computer, into an appropriate named folder in Lightroom. 22 photographs were downloaded from the memory card.


8. Review downloaded shots. First pass to be keep/discard. Then star rating for the 'keeps'. 6 of the 22 were discarded in this process, leaving 16 shots to rate. 9 of these were given 1 star and 7 two stars. Of these 7, 4 were upgraded to 3 star. 


9. Choose 2 best shots from highest rated. Being left with four photographs and having to choose two, I decided that I would process each in Lightroom and Photoshop before picking the final two. Having done this, the shots I have chosen are shown below.




The first, because of the colour and the composition.

The second because of the interesting shape of the brasswork and the reflection of the building and sky.

Assessment of Exercise


The workflow used wasn't much different from my normal method although writing it down beforehand did help to highlight certain points, like the need to continually assess shots during a shoot. Having a limited time-frame was also a challenge, although it helped to focus the mind!


One of the problems I have with checking, not particularly the histogram, but the sharpness of a shot in camera is that I wear reading glasses and need these to see the display clearly. Of course, I could put on my glasses each time but this is not ideal.


A learning point for me relates to the final judgement of the, in this case, 3-star photographs. It is clear to me that the final decision should be made once the photographs are finally processed via Lightroom/CS5 and not before. Apart from any cropping required, this is particularly relevant as I shoot in Raw and processing is essential to achieve the final result. 

Introduction

As an introduction to this course, I thought that it would be useful to set down what equipment/software that I have been using for the taking and processing of my photographs.


I have a Canon 50D digital camera with a few lenses - wide-angle, 100mm macro, 50mm prime and an 18mm to 270mm 'zoom' lens. Where possible, I shoot in camera Raw for maximum flexibility, although I usually carry a Fuji compact camera which takes only jpeg.


However, I have recently purchased a second-hand Lomo LC-A camera as a 'fun' camera, following attendance at a Lomo course in Brighton (birthday present!) run by Kevin Meredith (Lomo Kev). This was my first serious use of a film camera - and I liked it, particularly the effects from cross-processing the film. Once I get the LC-A serviced, I shall incorporate it into my normal photography. 


For software, I use a combination of Lightroom 3, Photoshop CS5 and Adobe Bridge. It is fair to say that I am a novice when it comes to processing software but I am using Linda.com as an online learning resource for Lightroom. All my digital shots are downloaded to Lightroom initially, using a memory card reader.  I have a Sony Vaio lap top to process and store my photographs (currently backing-up to Crash Plan!). To colour calibrate the monitor, I use Spyder 3 Express software. 


For printing, I use a Canon Pixma MP640 or an on-line printing service (usually PhotoBox).


Finally, I have a Flickr account ('Citysnapper') where I display selected shots.