Thursday, 30 September 2010

Digital Photographic Practice 1: Exercise 19 – Correction

The opening part of this exercise concentrates on removing dust shadows – something I have never had to do because of the effectiveness of the sensor-clean system on my camera. To complicate matters the current re-build of the OCA student website means I can’t download the example file. So instead I have gone with this shot from my own collection. The blemishes are in fact defocussed raindrops on the window I was shooting through. From examples elsewhere on the web – here for instance – I’m comfortable that this will pose the same challenges.

I used the Spot Removal Tool in Lightroom 3 to remove the very obvious grey blobs – here it is in action:

In general terms you select a diameter for the tool suitable for the blemish, and click on the centre of the blemish. the system automatically selects a nearby area to cover the blemish. This can be tweaked manually if required. This manual tweaking makes it very similar in behaviour to the clone stamp tool in Elements, although it is slightly less flexible because you can’t ‘paint ‘ with it for a large area correction. The end result is here:
Moving now to flare correction – I tend to reject obvious examples of flare – so I have very few in my collection, and again no ability to download a sample from the website. With this in mind a deliberately introduced flare in this photo, taken on holiday in Egypt. Note the peculiar shape and colours of the flare highlight at the bottom of the picture. This is a characteristic of the extreme wide-angle lens I used which has a very large front element which is difficult to shade from the sun – it is clearly visible in the viewfinder , and can normally be avoided at the point of capture. I think it is clear from the outset that selecting this ‘blemish’ and applying a curve adjustment will be ineffective as the colours have covered the detail in the grass.

It is however , simply corrected using the Lightroom Spot removal tool – using several overlapping spots to cover the lower blemish:

Perhaps a less dramatic example is this one – taken a couple of days earlier:

This one was cleaned up using the clone tool in Darken mode, as suggested, although the end results were very similar to use of the Lightroom Spot Tool.

Acceptability of this type of clean-up
I struggle to understand any real objection to this kind of clean-up, except perhaps in forensic or other photography where recording the precise detail of the scene is critical. For example, the corrected pyramid contains some fine detail in a small area of stonework that is different from reality. This has no practical importance unless the photo is specifically for record or analytical purposes. Otherwise the scene is indistinguishable from my memory of the day.
The suggestion that you should deliberately leave your mistake in as a more ‘accurate’ version seems totally illogical. On that basis, should we simply be required to retain all our mistakes? There is an interesting parallel with the use of a spell-checker. Would anyone seriously suggest that in the interests of truth that I should publish this blog with all the spelling mistakes and typos that my banana-fingered typing introduced at the first pass. Of course not, and I can see no reason for doing the same with a photo. The flare/dust spots were introduced by the photographic process, and (subject to the provisos above for the special cases mentioned) it seems perfectly legitimate to remove them.

Sunday, 19 September 2010

DPP1: Assignment 3: Tone Roses

Well, finally I finsihed my third assignment and have sent it to my tutor. If anyone wishes to have a peek at the photos, or comment, they're on Flickr, here:

Have not tried serious black and white photography before, and really enjoyed the challenge of this assignment. I'll post my own thoughts on the photos once I've had my tutors feedback, but here's my favourite shot from the set.

Wednesday, 1 September 2010

Digital Photographic Practice 1: Exercise 18 – Colours into tones 2

As it says in the notes the aim of this exercise was to use channel adjustments to achieve a specific effect. The 3rd suggested topic in the exercise  (a pic of a garden with the greens appearing light in tone) appears relatively simple to achieve so to be perverse I’ve tried the other two.
Emphasising aerial perspective
The challenge here is to emphasise aerial perspective in a landscape. I chose the following image – a desert landscape from the Sinai - as a starting point as it already has a strong sense of aerial perspective. It is a slight crop from the original to remove some rather unsightly road edging from the foreground. The background mountains clearly show the characteristic blue haze which adds to the perspective.
A default conversion, with all the colour channels set to zero looks as follows:
This is rather flat and un-interesting. Although there is some sense of distance, it is perhaps slightly less effective than the colour version, and there is little tonal difference between the foreground, the distant mountains and the sky.
In his book ‘The Photographers Eye’, Michael Freeman suggests that favouring the blue channel during conversion emphasises aerial perspective (Chapter 2), so that was the starting point for the final conversion.
The overall conversion figures are as follows. I lifted the blue channel +75 and the Aqua channel +95, then I reduced the overall brightness of the picture by shifting the curve (-32 lights, –52, dark, –5 shadows) and darkened the sand still further (Orange –14, Yellow +7). Finally I softened the picture very slightly (clarity –6).
This treatment has had the visual effect of increasing the distance to the first mountains, by losing the clear join between sand and rock, and reducing the visible detail. By reducing the contrast between the sky and the distant mountains still further it has also made them appear more distant, increasing the apparent separation between the closest mountains and the others.
Lightening skin tones
The challenge in this case is to lighten the ski tones in a portrait without significantly altering the rest of the image. M y chosen image for this exercise is a grab shot of my daughter in a roadside cafe. the right hand side of her face is somewhat lost in the shadows caused by the relatively strong lighting from the left of the photo.
The default conversion gives this result, which increases the impact of the shadow even more.
Examination of the picture shows that the skin tones are quite red/orange, so using the same principle as in the desert landscape, I started by lightening these colours to give the following (which has also had some noise reduction applied).
I felt at this stage that no further adjustment was required to meet the objective of the exercise – other than a slight softening by reducing clarity to –8. I cropped this final version slightly because I felt the ring was slightly distracting, and if I were printing it for display I would clone out the highlight on the top-edge and perhaps crop some more from the LHS.
A relatively simple set of adjustments has succeeded in lifting the portrait from the shadows in a way that would not be easily achievable in the colour version.