Sitting my niece on a stool, I set about photographing her
using different white balance settings. First, I used the auto setting and
composed the shot so that both inside and outside could be seen. I can
instantly see that this mixed light has confused my camera. The green hedge and
red chair look fairly close to their true colours, although just a touch too
blue. To the right of the shot, an
orange colour cast can be seen, most notably in the white toaster which appears
to be pale orange. When looking at my niece’s face, I can see a slight blue
cast to the left and orange to the right but this evens out in my mind and altogether
looks acceptable. The mixed lighting levels have also created problems as the side
of Evie’s hair is totally burnt out.
I then set the camera
to the daylight setting. In this shot, the outside colours look almost as they
should; the hedge looks green and Evie’s right arm and the back of her dress
look very close to the right colours. To the right of the shot, the area which
is lit by tungsten lights, everything appears slightly orange, which again, can
be seen most clearly in the white toaster. The daylight setting seems to have
coped with the mixed lighting levels better as Evie’s hair isn’t as burnt out here.
Finally I set the camera to the tungsten setting and as
expected, as a result of tungsten light being more orange than natural light,
the camera setting has added a blue colour cast. The right side of the image
facing the tungsten lights is a touch too blue but fairly close to correct
colours. To the left, the hedge outside looks very blue, far from its natural
colour and a large part of Evie’s face is totally burnt out.
I can see from this experiment that mixed lighting is very
difficult to deal with. I know that editing software has made it possible to
merge two images to compensate for such problems but because of the fluidity of
the colour of light, this would not be a simple task.
......................................
For the second part of this assignment, I have found two
different types of fluorescent lighting. The first is of an office, lit by
bright fluorescent strip lights. Much of the office is white so it is easy to
spot colour casts. I took the photograph with the camera set to automatic white
balance and then the fluorescent setting; the result is quite similar for both
images; the colour temperature is the same with only the tint being slightly
pinker.
In the lighter areas of the image, both shots look almost
white; however on looking into the darker white areas such as the pillar, the
pink and green tinges are quite clear. A purer white could be achieved by
sliding the tint controller to a point half way between these settings.
Another thing I notice when looking at these images is that
they have a strange dream like quality to them; the colours are not quite as
saturated as I remember them to be. I am unsure whether this is down to me
shooting the image through a window or because the fact that fluorescent
lighting does not emit a full spectrum of colours. I will explore this anomaly
further as I move onto study night lighting in my next project.
My next image was taken in a coffee bar under what I assume
to be the CFL type lamps described in the brief for this project. These types
of bulb look like tungsten lamps but it is clear from my results that they are
actually fluorescent.
In this pair of images, the image taken on auto is greener
than that on the fluorescent setting, the opposite of the previous experiment.
In the left hand image, the colours appear washed out and the image rather
dull. The image using the fluorescent setting, although slightly too pink, is
closer to the actual colours seen on the day and much more vibrant.
The fluorescent setting has had quite different results on the two images because there are different types of fluorescent lighting. As my camera only has one fluorescent setting for white balance, I will always have to be careful when shooting in indoor public areas. For most sitautions, white balance will have to be corrected where possible when editing.
No comments:
Post a Comment