Showing posts with label sensor. Show all posts
Showing posts with label sensor. Show all posts

Wednesday, 5 June 2013

Exercise: Higher and lower sensitivity

The objective of this exercise is to demonstrate the use of the third element available to the photographer for the control of exposure.  It is this triumvarate which needs to be used to determine the appropriate exposure for the subject.
Aperture and shutter speed have been dealt with in previous sections, here we are looking at the sensitivity setting which is commonly expressed as ISO.  

When shooting film, the sensitivity of the film was likewise expressed in ISO, though frequently referred to as the film speed, and typically one selected the film to be used for a particular shot depending upon the subject and the environmental conditions.  Once loaded, you were stuck with that film and that setting, however there were workarounds which could be used.  With 35mm film, I preferred to wind the exposed film back into the cassette, remembering to leave the tail out.  I would then choose an alternative film.  If I did not happen to have a different film type in my bag, then all was not lost as alternative techniques could be used.  It is possible to use a technique referred to as 'push' processing to increase the effective speed of films.  Although it was referred to in this way, and the ISO could be set to, say 400 or even 3,200, what this actually did was to apply additional development to the underexposed areas so enhancing what was recorded on the film.  However, conversely, highlights would now be blown out and prove impossible to print.  This technique was fine when there was a narrow dynamic range, but care needed to be taken when there were a lot of highlights.  I referred to an excellent book in order to refresh my memory; published by Ilford / Focal Press and written by Jack H Coote Hon FRPS, called Monochrome Darkroom Practice.

The medium speed films such as Ilford's FP4 was not as good as the faster films, rated at 400 ISO, such as Ilford's HP5 or Kodak's TRI-X when push-processed.  Using the Ilford developer, Microphen, it was possible to push these to 800, 1,600, and even 3,200 ISO.

Apart from the downside referred to previously, typically, 'push' processing increased the size of the grain giving a gritty appearance to the print, which worked extremely well for some subjects and was quite popular.

So what is the relevance of my meander into the analogue world.  Well it shows that use of ISO to enable shooting in poor light conditions has always been a consideration and it had its downsides.  This is no different to today where I use different ISO settings in order to provide an appropriate shutter speed for difficult light conditions, or simply to achieve the shutter speed I require when photographing fast moving subjects such as jet aircraft.  Increasing ISO brings with it increased 'noise' in the digital world, which like in the analogue world is not always desirable.  When shooting at high ISOs then conversion to B&W is frequently desirable as some of this noise can be visually eliminated or disguised and it starts to emulate analogue images.  Sensor development, on which I am not going to write here, has enabled elimination of this noise with stunning effect, so at 400, 640 and 800 ISO, there is frequently not much call for concern particularly when using the latest full-frame cameras.

Moving on to the images which I have taken to illustrate the use of higher and lower sensitivity.  The first image is one which illustrates the use of high sensitivity to cope with poor light conditions.

Westland Lysander - ISO 1600, f/5.6, 1/250 sec
The photo of the Lysander was taken shortly after 8pm early in May, and as it was overcast, the light level was very poor.  The only way to achieve a photograph without undue camera shake was to increase the ISO setting to ISO 1600 from the "normal" one which I would use, ISO 200, to photograph aircraft in flight.

Greenwich market - ISO 100, f/4.0, 1/15 sec
The shot above, taken in Greenwich indoor market, really stretched the limits for hand holding and camera shake is visible at 1/15 sec.  This isn't really acceptable, so the solution, again, is to increase the ISO setting, this time to ISO 800.  The result is shown below:

Greenwich Market - ISO 800, f/6.7, 1/30 sec
Though still not absolutely sharp, it is a substantial improvement on the previous shot, and I could have easily increased the ISO to 1600 or even 3200.  With sensor technology improving seemingly each year, high ISO speeds are now producing very acceptable results.

Greenwich Market - ISO 100, f/4.5, 1/8 sec
 A further photo taken in Greenwich Market.  The first one, above is taken at 1/8 sec and shows considerable camera shake, whereas the one below at ISO 800 is much improved and is almost acceptable.

Greenwich Market - ISO 800, f/4.5, 1/60 sec
Use of higher ISO in order to assist with shooting hand-held is not restricted to indoors, but when shooting in shade outdoors, for example as shown in the images below.

The first pair of images is that of the sign outside the "Up the Creek" comedy club in Greenwich.  This sign has always amused me so it was just too tempting not to include it here.
Left: ISO 100, f/11, 1/10 sec  Right:  ISO 400, f/11, 1/60 sec
The small images above look fine, however differences are very marked when viewed full size as shown in the pair below.
The thinking man
The left hand frame visibly suffers from camera shake, whereas that on the right, helped by shooting at ISO 400, is much more acceptable.

There is a price to pay when using higher ISO.  This is true using film, but technology has made huge steps forward in sensor development, so much so that in the latest generation of sensors, coupled with the onboard processors, high ISOs such as 3200 or 6400 produce images with little in the way of noise, resulting in very acceptable images.  Even my EOS 5DII, which is several years old, has acceptable high ISO performance.  Below is an attempt to illustrate this:
Ballerina, Covent Garden
Top Left:  ISO 200, Top Right:  ISO 640
Bottom Left:  ISO 1600, Bottom Right:  ISO 3200
At the size shown above, there is very little difference between the images, though it is possible to see less definition in the shot at ISO 3200.  Differences are more visible full size as shown below:
Top:  ISO 200, Bottom:  ISO 6400
In the bottom image it is possible to see "grain" or noise, particularly in the dark facial area, though noise is also visible in the light colours of the building wall.  This noise is not particularly intrusive and could be easily dealt with by smoothing in Lightroom.

Tuesday, 16 April 2013

TAOP Part 4: Light

Time to make a start on this part as I have very little time in which to complete part 4.  The introduction to this section in the course manual covers light as "the raw material of photography" and then leads on to "The intensity of light".  Here there is a general discussion of measuring the brightness of light and easily translating it onto a scale which can be used practically to provide the appropriate exposure when taking a photograph.  There is also a brief discussion of sensor sensitivity which is measured in ISO numbers as well as histograms.  These latter two subjects I will expand on further later in this blog.

Measuring Brightness

At this point I reached out for my little-used camera manual to remind myself what metering patterns are available.  Being rather lazy, I generally use what Canon refer to as 'evaluative' metering and this does seem to work well in most circumstances, though when working with speedlights, whether in the home studio or out in the field, I use a hand held meter.  As I have been using hand held meters for some time, I thought that I'd cover those off first.

Hand held meters

First of all, on of the major advantages of using a hand held meter is that it becomes possible to take incident light readings.  Why is this important and when should an incident light reading be taken?  An incident light reading measures the amount of light falling on the subject and this gives an accurate reading to be used in determining exposure.  Such a reading is independent of the texture and colour of the subject, which means that it is not necessary to make adjustments to take those into account.  It is appropriate here to touch on how meters are calibrated, regardless of whether they are internal to a camera or hand-held.  The basic principle which is adopted in metering the brightness of a scene is to reduce it to an 18% reflectance grey card.  This has considerable significance where the subject is predominantly at one or another end of the scale, ie either black or white.  A black surface will cause the meter to calculate an exposure some times up to 2 stops higher than required and a white scene, such as snow goes the other way and causes up to 2 stops  under exposure, giving rise to grey snow scenes.  An incident light reading is independent of these factors so more likely to give you an accurate exposure.

There are practical considerations, however, and it may not be possible to take an incident light reading in most circumstances, so I use it primarily for studio work, whether by flash or other light sources, and also portrait and close up work outdoors.  My first hand held meter was a Weston Master V, which is currently not working, however in doing this research, I discovered a company which repairs these meters, so I may well do so.  What did it look like, well, somewhat like this:


The white plastic attachment seen in this photo is called an "invercone" and it fitted over the area where the selenium cell receptor was located.  This is shown in the photo below.  Using the invercone was how incident light was measured.  Reflected light measurements were made without this attachment, pointing the receptor directly at the subject.


The hand held meter which I currently use is an Sekonic Flashmate L-308S, which is shown below:

The digital display and multiple functions, make this a very versatile tool despite being one of the cheaper models.  In the picture above, the white area is a "lumisphere" which is used for taking incident light readings in the same way as the Weston.  The main advantage of this type of lightmeter is its ability to take readings of flash output.

For metering small areas, referred to as spot-metering, I recall using at college a Pentax Spotmeter V which looked like this:



A rather strange device which incorporated a viewfinder which you used to find the "spot" you wished to meter.

In-camera metering

Reading the manual for my camera, the Canon EOS 5D mark II, I find the following on a page describing the Metering Mode:


Not a wonderful scan on my part, but it illustrates the point.  There are 4 metering modes, all with specific attributes, well described in the illustration.  I have a tendency to use the default which is the evaluative mode, but this is not necessarily the most appropriate in all circumstances.  Having said that, it all depends on how much you get to know your camera.  In theory, partial and spot metering will provide the best exposure values where you are looking to apply your own compensation, for the scene.  Evaluative does many calculations across a number of zones and it could well be applying compensation, so adjusting further may not make much sense.

As the course text suggests, different scenes require different treatment and an understanding of how the metering works in the specific camera in use so that appropriate compensation, if any is required, can be applied at the time of taking the image.