Successful Filtration

Part 1
Colour Temperature and “non-daylight” situations.

I should start this article by reiterating my key philosophy behind the use of filtration in landscape photography:

“If you can tell you’ve used a filter, then your use of it has failed.”

All filters should be used purely to allow the film to see the subject in the same way that the human eye does. In other words we are trying to recreate on film the scene as it appeared to us when we stood behind the camera.

There is a subtle difference, however, when we come to discuss the colour of light (as opposed to contrast management as discussed in Part 1 of this series of articles). This is because the most successful images always depend on the “quality” of the light illuminating them. I have often stated that “I do not photograph the landscape, but the light that illuminates it”. How often have you visited the location of a superb image you have seen, only to find that it doesn’t look as good as you imagined? When it comes to successful landscape photography, light is absolutely the key. In order to succeed you must gain an intimate knowledge of light, and the way your film will react to it.

”Daylight” balanced film.

At this point I should say that the problem of colour discussed here is most important to users of colour transparency film. Colour negative film has to be interpreted at the printing stage by the printer, and with no fixed reference point, the resulting print could come out almost any colour at all (and frequently does). With colour transparency, the final result is clearly visible to all and, as such, isn’t open to interpretation by the printer because a direct comparison between transparency and print can easily be made.

If you look on the box that your film came in, you will probably notice the words “Daylight Balanced” or something similar. In order to represent colour successfully, film manufacturers have agreed on a standard for the expected colour of light and most colour films are manufactured to respond successfully to this standard. This standard defines “photographic daylight” as:

“a mixture of direct sun and open blue sky illumination, as found in Washington DC, USA on a typical clear day between the hours of 10am and 4pm”.

Obviously we do not make most of our images in these precise conditions, so why don’t we need to use filters on nearly all our compositions? The simple answer to this is that it is often the non-standard conditions that attract us to making the image in the first place. For example, when photographing a beautiful sunset, it is the glorious reds, golds, pinks and yellows that we are trying to capture. The last thing we want to do is remove all this colour and go back to the standard “daylight”.

There are, however, certain conditions where a more subtle change is required if we want our images to reflect the way we saw the subject when making the exposure. This is because when we look at certain subjects with our eyes, our brains interpret the input and apply certain “corrections” based on our experience of different viewing conditions. For example, a sheet of white paper viewed indoors (under household light bulbs), outdoors in the sun and outdoors in the shade under a clear blue sky, always appears to us to be a sheet of white paper. To film it would appear as various shades of orange, white and blue respectively, because it is being viewed under light of varying colour.

In order to clearly understand what is happening hear, we need to apply some order to the situation. We do this by applying mathematical values to the “temperature” of light, and arriving at a term known as “Colour Temperature”.

Please bear with me through this bit – it will all make sense at the end (I hope).

Colour Temperature and the Kelvin Scale.

In photography we borrow certain terms from physics in order to quantify the “temperature” of any light source by using terms that relate to so-called “black-body” radiation. Suffice to say that any “hot” light source will emit light, and the colour of that light will depend on the temperature of the source itself. For example an iron bar heated to 1000 degrees will emit light of a constant colour anywhere, anytime. In this way it is possible to give specific values to the colour of various light sources in a temperature scale that uses units called Degrees Kelvin. Degrees Kelvin can be considered as the same as degrees Celsius or Centigrade except that the scale starts at a level called “Absolute Zero” which is equivalent to minus 273 degrees Celsius. So - 0 degrees C = 273 degrees K.

The Colour Temperature scale covers a far larger range than the colours that an iron bar could be physically be heated to. Some key points on the scale are given in the following table.

One note of caution. The hottest temperatures on the scale relate to the BLUE end of the scale whereas the cooler end of the scale contains the RED colours. This is the opposite to what we might expect as we usually refer to Red colours as being “warm” and BLUE colours as being cool. Just to add to this confusion, the warmest (yellowest) filters are often referred to generically as “warm up” filters, but they actually reduce the colour temperature (move it towards red).

Filter systems for colour photography.

The most widely used system for describing filters that effect the colour of light in stills photography is the “Kodak Wratten Series” of filters. Nearly all other filter manufacturers will state that their filters are equivalent to a filter designated on the Wratten scale. The full scale, the relevant “Mired” values (more later) and the associated filter factors are given here (Note: the filter factor is the increase in exposure in stops, required to allow for the amount of light cut out by the filter). The Wratten series of filters can be split in to two main types:

• Colour Conversion Filters – which make large changes
• Colour Correction Filters – which make subtle changes

(There is another type called “Colour Compensating” filters which I will cover in a future article)

Filtration and “Mireds”.

If we were only concerned with measuring the colour temperature of a particular light source then the Kelvin scale would be sufficient for most of our requirements. Unfortunately the key to this whole article is that we want to CHANGE the colour temperature by applying a filter and this causes other issues.

The fact is that any single filter will affect the colour of a light source differently depending on what colour it is in the first place. This sounds confusing, but if you think about it, if we were to view a yellow light source through a filter of the same colour we would detect virtually no change. If instead we looked at a blue light source then the change would be significant. In practice, for example, a Wratten 81B filter reduces daylight by 715 degrees Kelvin, but lowers the colour temperature of tungsten light by only 65 degrees Kelvin.

Fortunately there is a solution. This is achieved by dividing the Kelvin rating of a light source INTO 1 million to give its “MIRED” value. Similarly by dividing the mired value into a million converts it back to degrees Kelvin.

The MIRED value (pronounced “my-red”) is an abbreviation of MIcro REciprocal Degrees, the key being the word reciprocal. The scale actually runs the opposite way to the Kelvin scale and blueish light sources have a low mired value whilst reddish light sources have a high mired value.

Just as light sources can have Mired values assigned to them, so can filters. The Mired values associated with filters are known as Mired “shift”. This is because the filters at the yellow end of the scale have positive values and “shift” the colour of light higher up the Mired scale. Conversely filters that are bluish in colour have negative Mired values and “shift” the colour of light down to a lower Mired value.

This may seem a bit pointless, until you realise that any given filter will always “shift” the colour of light by the same Mired value, regardless of the colour of the light source. As well as this, when using multiple filters, it is possible to add their Mired values together to calculate the overall “shift” they will cause. In addition we can say that :

Mired value of Light Source + Mired value of filter(s) = Mired value of result.

Film colour balance using the above.

As mentioned above, “Daylight” balanced colour film is expecting the light source to be:

“a mixture of direct sun and open blue sky illumination, as found in Washington DC, USA on a typical clear day between the hours of 10am and 4pm”

Now that we understand “Colour Temperature” and “Mireds” we can more accurately state that it expects the light source to be:

Colour Temperature = 5500 degrees K or 182 Mireds.

There are some special films designed specifically for photography under artificial lighting which are known as “Tungsten” balanced films. The more common of these which are available from a number of manufacturers are balanced to a light source of:

Colour Temperature = 3200 degrees K or 313 Mireds

The more observant of you may notice that the difference in Mireds between these two types of film is 313 – 182 Mireds which equals 131 Mireds. You will also see from the Wratten series scale that the 85B Yellow filter and the 80A blue filter have an associated Mired shift of 131 Mireds. It can be seen therefore that these filters are designed to correct for the use of Tungsten film in Daylight conditions, or the opposite – Daylight film under Tungsten lighting.

Using my filters in the “real” world.

So how does all this mathematical discussion really impact on our photography?

As mentioned above, there are, certain conditions where a more subtle change is required if we want our images to reflect the way we saw the subject when making the exposure. Also we need to be aware that our brains automatically compensate for small shifts in colour, such that we cannot expect to easily see when some form of correction is required. This comes down to experience, and as noted in the article on ND grads, there is no substitute for accurate note taking and the subsequent assessment of results.

Here are two examples of situations that may require us to apply filtration to correct for the fact that the colour of light does not match what the film expects.

1. Photographing in open shade under a clear blue sky, particularly when the tones of the subject are very pale. In situations such as these, the subject is lit by a light source whose colour temperature is far higher than the film expects (anything up to 20000 degrees K). In these circumstances, the processed film will contain a distinct blue colour cast, which can be corrected at the taking stage by using one of the 81 or 85 warming series of filters. This effect can be particularly noticeable if you are photographing snow in open shade. We have all seen the way that snow can appear blue in images made under these circumstances. We MUST remember though, that we may desire the blue cast for our snow scenes, as it helps to convey the feeling of cold to the viewer.

2. Although film manufacturers make every attempt to achieve perfect colour rendition in all circumstances, some films are unable to achieve this when it comes to particularly long exposures. This is known as “Reciprocity Failure”. In order to help overcome this, manufacturers publish how to deal with long exposures for individual film types. For example, Fujichrome Velvia 50 (my film of choice) suffers from a tendency to include a green cast for exposures longer than 4 seconds (the longer the exposure, the worse the problem). Strictly speaking it is necessary to apply pure Magenta filtration to counteract this property (colour compensating filters), but in reality a filter such as the 85C will help.

Hopefully the text, tables and examples in this article will help to clarify the situation regarding filters, the colour of light and the films’ response to it. I should finish by emphasising a few points :

• The decision to use a particular filter must be a considered one. Any of the Wratten filters that we choose to use will effect the WHOLE of our composition, not just the problem parts. Warm up filters used in images that contain blue sky can have a disastrous effect, making the sky appear muddy and dull. It is possible to have graduated 81 series filters specially made (e.g. to warm up the landscape but not effect the sky) but these are very expensive and difficult to obtain.
• Mixed light sources are very difficult to deal with. Photographers who specialise in architectural interiors for example will be all to familiar with the infinite colours that artificial lights produce. Add daylight to the equation and the balancing trick is especially difficult to overcome (this will almost always involve the use of a meter to precisely measure the colour temperature of the various light sources and may involve filtering the light sources themselves!)
• Practice makes perfect; try using the Wratten filters and note down their successes and failures in your results. Used correctly they can greatly enhance your ability to capture your subject in the way that it appeared from behind the camera.
• Remember that successful images are often all about emotion, mood and ambience. The colour of the light source is usually a key contributor to obtaining these successful images and sharing those emotions with the viewer. Don’t be in a rush to filter it away!!

No Filtration	Your ability to see these images successfuly depends largely on your monitor's ability to display them accurately (and my success at scanning them!). Even so I think the excercise is worthwhile as you can at least appreciate the relative effects. Note how horrible the blue sky looks through the stonger filters
Lee Filters 81A	Lee Filters 81B	Lee Filters 81C
Lee Filters 81D	Lee Filters 81EF	Lee Filters 85C

Footnote: I am in the happy position of owning and using a Canon EOS10D digital camera. The issues discussed in the above article are dealt with in an entirely different way in the digital age. It is possible to manually tell this camera that the light source is any temperature between 2800 and 10000 degrees K, thereby achieving the same result as the filters achieve with a film camera. This is a real advantage for digital photographers, especially as it is possible to shoot in RAW format and then adjust the Colour temperature at a later date i.e. to retrospectively apply the filtration!!!!!!