Camera icon Invisible Light

Home Photos About IR Photography Books Links Buy Prints

Thermal Imaging in Far Infrared

Infrared photography is often confused with thermal imaging. The difference is basically one of wavelength of electromagnetic radiation since the infrared light with which my photographs are taken lies just beyond visible red light: this is why it is called near infrared. Near infrared radiation is not related to the temperature of the object being photographed - unless the object is extremely hot - since the object is not actually the source of the radiation. It is just reflecting it.

The film 'sees' the object because the sun (or some other light source) shines infrared light on it and it is reflected or absorbed by the object. You could say that this reflecting or absorbing of infrared helps to determine the object's 'colour' in a four-dimensional colour space made up of blue, green, red and infrared. (There is more information on this in the colour mapping pages.)

If you make an object hot enough it will start to radiate heat and light.

An electric fire is usually hot enough for us to see reddish light as well as the infrared (and heat) it radiates. An electric iron (or flatiron for ironing clothes) is not hot enough to radiate visible light but it does radiate near infrared light.

In the first two editions of Walter Clark's book 'Photography by Infrared' (see our bibliography) there is an image taken using infrared film which shows a bust illuminated only by the light of two such irons. This photo is reproduced on the right. The irons themselves are in the shot and appear to be glowing.
Bust illuminated with flat-irons

Clark explains that infrared photography can see hot bodies at temperatures from about 250 C up to the point where they begin to glow visibly at about 500-600 C, depending on how dark it is and how much your eyes have adjusted to the darkness. (W de W Abney, writing the Journal of Photography in 1881, speculated that it might be possible to photograph a kettle of boiling water by its own near infrared 'light' but Clark thinks this 'appears to be quite out of the question'. Ye cannie break the laws of physics Captain!)

So here is the source of the confusion: you can photograph in infrared by the 'light' of a hot iron and yet the infrared illumination is not, in itself, heat anymore than red light is heat.

Light is often said to have a colour temperature. What this means is that the colour of the light is the colour of light radiated by a so-called black body (an idealised radiating object) which is at that temperature. This can easily be seen to apply to a photographic lamp or even the sun, but it can be applied to any source of light? Colour temperature is measured in Kelvins and the higher the colour temperature the bluer the light. Zero Kelvin is Absolute Zero and the freezing point of water is 273 Kelvins: a Kelvin degree is the same as a Celsius or Centigrade degree. In practice the actual temperature is not the same as the colour temperature. Here are the colour temperatures of some common light sources as quoted by Clark:

  • Open sky - approx 20,000 K
  • Overcast sky - 6,500 K
  • Direct sunlight - 5,400 K
  • Carbon arc light - 3,780 K
  • Photoflood bulb - 3,400 K
  • 100 Watt tungsten bulb - 2,865 K
  • Candle flame - 1,930 K

In colour photography, especially for slides, it is important to match the colour temperature of the light source with the colour temperature the film expects to see. Otherwise the colours will be incorrect. This seems counter-intuitive at first, because we ourselves do not usually have a problem when moving from one light source to another. However, our brains adjust for much of the colour difference in a way that films do not; so much so that it is very difficult to judge a colour correctly unless the lighting and ambient colouration are carefully controlled.

The two most common colour films are colour balanced (as this is known) for daylight or for tungsten lighting. Daylight shot with tungsten film will appear too blue and artificial tungsten light shot with daylight film will appear too red. You can correct this problem with negative film to some extent when you scan or print it so the issue is not so critical (but spectral highlights will probably undergo a colour-shift when you do this so getting the colour balance right to start with is still a good idea). Digital cameras also need to be set for the appropriate colour temperature. Fluorescent lights do not emit light in the same way as the sun or a light bulb ... or even a flash gun. There are noticeable gaps in their colour spectrum even though they can appear to us to be white. This usually leads to photos shot under fluorescent lighting appearing to have a green cast.

Unfortunately, in order to take an image of something by means of the 'light' it emits, the sensor has to be 'colder' than the radiation from the object that is being used to 'see' it. This can be the object itself glowing or it can be light produced by glowing objects such as the sun or a lamp. For conventional photography this is no real problem because even the light from a candle flame is at nearly 2000 Kelvin whereas room temperature is only around 290 Kelvin. When we get down to thermal imaging, where we want to 'see' objects at around room temperature using their own heat radiation the problem is very real. In general, the sensor has to be cooled to well below freezing. Only then can we start to image with real heat: Thermal Imaging.

You can see some examples of this type, courtesy of NASA's Jet Propulsion Laboratory, on this next page.

You will have noticed that there are heat sensors that can detect the heat from a body while operating at room (or ambient) temperature. This kind of device is used in movement sensors for security alarms and detects a change in far-infrared rather that its absolute level. They can be triggered by any change in temperature, even a breeze blwing in front of the sensor can do this. It is possible to produce imaging from this kind of sensor by using a shutter to continuously 'flicker' the radiation reaching the sensor array. A UK company, IRISYS, have developed a thermal imager based on this technology. This is their web page (opens in a new window).

The practical applications of this kind of imaging include the remotest of remote sensing. Mark McKelvey, one of the team involved in the JPL project I referred to above, has more recently been working on infrared spectrometry for the Mars Rover project. It is possible to identify minerals by studying their infrared reflectance: this is analagous to the 'colour' of the rocks but at much longer wavelengths than visible light. NASA's page is here (opens in new window) and there is more detail on a Cornell University page about the Mini-Thermal Emission Spectrometer, or Mini-TES, here (opens in a new window).

More background on infrared photography
Technical background NASA IR Video Camera IR 'Colour' Mapping
Imaging Abstracts Noctovision  
All contents of this site are Copyright © 1998-2004 Andy Finney unless otherwise stated and must not be copied, reproduced or mirrored without written permission.