Thursday 25 November 2010

ISO Camera Tests

With digital video cameras, such as the Canon 7D or 550D, an ISO (which refers to film speed, even though digital cameras do not take film. Click here for an explanation of ISO) has to be selected before filming can occur. ISO speeds range from 100 to 3200 (and in some cases higher) and each have their own advantages and disadvantages. The reason for this test is due to the fact that within Finchers films, the shots are visually very dark, with little to in some cases hardly any light being used. Combining this test with the natural light test (as this is also something which is common within his films, by only using natural light and artificial light that is available in the scene to light shots) will allow me to develop an idea into how I should light my final assignment project. In this first ISO test that I conducted, I experimented with how the Canon 550D handles in extremely low light conditions, which in this case was simply the light from a lit match.

I wanted to see how much of the subjects face and surrounding area was visible depending on which ISO setting the camera was set to, so to keep the test fair, the factors that were kept the same were the f stop amount, the lighting conditions and also the lens used. I decided upon the Canon 50mm 1.8 lens as this was the lens that I originally used for my natural lighting test and was extremely impressed with its results.



There is a mistake in this video however, which I need to confirm before I go any further. The last two shots do not have the correct titles on them detailing what settings were used. I
increased the ISO in a sequential scale, so following on from ISO 1600 was the ISO 3200 test, and not the 6400 test that is stated. The ISO 6400 test is the last shot seen in the video! This was a title error on my behalf.

The first shot features an ISO of 100. Even with the aperture set to 1.8, letting in the maximum amount of light to the lens the image was still very dark, with only the edges of the fingertips being lit.

The next ISO speed was set to 200, which isn't a dramatic increase from the first. Technically there should not be much of a difference between the two, which is definitely the case from the image shown below.

Next up was the jump from 200, to 400. This was the ISO speed that I chose to shoot my natural light test on, and with the aperture fully open to 1.8 I felt this was the perfect setting for those conditions. In the low lig
hting test however, the image was still quite dark, with only a fraction of the subjects face being lit compared to the previous ISO.

The ISO speed doubles again in the next shot, this time showing far more dramatic results. This time the subjects face was clearly more visible and better lit, whilst still retaining all information and not rendering any grain/noise to the image.

Again, the ISO doubles to a level of 1600. I was always told that shooting about 80
0 would render problems, usually in the form of noise, giving the shot a very grainy image. Even with such a dimly lit scene this seemed to be the case, as the background wall area to the left of the subject clearly has subtle noise issues (which can be clearly seen when the image is at full size).

The next ISO was 3200 (b
ut please ignore the fact that the image is labelled as 6400). I had my doubts with shooting at 1600, and when I shot at 3200 it was clear to me that this ISO would most definitely not be used for my final piece. The noise and grain is now present across the whole of the image, and the highlighted areas such as the flame and the edges of the subjects fingertips are clearly blown out.

The final ISO was set to 6400 (again, it does say 3200 on the image but that title is incorrec
t). I wasn't going to shoot on such a high ISO as I already knew that the noise levels would be considerably high, but I did anyway in order to conduct a fair test. The grain in this shot is far more noticeable, with the highlights being blown out even more, as well as some of the midtones within the colour range being quite bright.

CONCLUSION
Taking each ISO setting into consideration, I personally think it best to shoot on an ISO of around 400 (possibly 200 in some cases). The reason being is because of the amount of light that the camera receives when the aperture is fully open. I have to think carefully about the amount of light that will be on location during shooting, and if I were to shoot on a high ISO ranging from 800 - 6400 the image would be far too bright and blown out. This is easy to understand as the lit match only emits a small amount of light, yet it is still seriously affecting the shot/visual quality on ISO levels such as 1600 to 6400.

By shooting at around 400 ISO I will be able to keep the scene dark on location, but still capture a high amount of light information onto the sensor, which can easily be pulled in post if I feel it is too bright in pre-production.