Saturday, 19 September 2015

Birds and Light - The Terminology of Light

At a basic human level we all have some appreciation for Light and probably have more than a few words to describe its qualities.  Here linked for example are 36 different adjectives describing light!  We probably never give this terminology much critical thought.  All these words have a subjective meaning and we probably use these terms interchangeably in normal discourse.  But light can be described far more scientifically.  When one first begins to research the scientific terminology of light the seemingly overwhelming array of terms can be quite off-putting,  Thankfully we don't need to understand all these terms for the purposes of observation and photography.

In order to measure anything we first need to understand what it is we are trying to measure.  Light is clearly not a simple concept and so when we talk about measuring light we have to consider the overall context.  There are a lot of different terms and units of measure for light and these are often used incorrectly or out of context.  This blog is no exception and I must learn to try and use the terminology with more care.

Light is measured by the camera's or perhaps a handheld light meter.  Light is also measured by the camera sensor.  In the case of the light meter what is being measured is irradiance in units of watts/m2.  Each individual photosite on a camera sensor actually captures individual photons of light and uses them to accumulate a charge which is proportional to the amount of light received.  But photosites have a limited charge capacity and therefore a limited dynamic range.  A light meter typically measures light across a much broader range.

Light is part of a much larger electromagnetic spectrum and when we are talking about light in terms of that totality we are referring to the science of radiometry.  At this, the broadest definition of light we have terms like radiant energy (the total energy of electromagnetic radiation), radiant energy density (that energy per unit volume) and radiant flux (which is the rate of radiant energy per unit time).  Because light intensity falls off as we move away from the source we need to measure intensity using another form of measurement called radiant intensity or radiance.  Lastly we have irradiance, which is the radiant flux received by a surface per unit area.  This last term is actually all we really need for our purposes but it may be useful to understand that we have all these other complex terms and it is quite easy to use a term out of context.

All electromagnetic energy can be characterised by wavelengths along an electromagnetic spectrum.  If we need to measure light of a fixed wavelength we must use a different terminology - spectroradiometry.  Then it is just a matter of applying the word spectral in each of the contexts above.  So we have spectral flux, spectral intensity, spectral radiance and spectral irradiance etc., all describing the same things as listed above, but in a narrower spectral context.

This is all well and good but in reality we don't typically work within the full electromagnetic spectrum.  Our eyes are sensitive to only a small fraction of the entire spectrum.  So we don't normally talk about light measurement based on radiometry.  When it comes to human vision it is based on a more specialist field of study called photometry.

Of course, as humans we all perceive the world in slightly different ways.  There is no uniformity when it comes to biology and there are a large range of variables at play during any given observation.  So we use normal distributions or functions rather than fixed values in the world of photometry.  The luminosity function for example describes the average spectral sensitivity of human perception of brightness.  This is all made possible thanks to the work of the International Commission on Illumination CIE.

In the experiment below I used a lux meter to observe light under a foliage canopy.  A lux meter works in exactly the same way as any other light meter but the readings are subject to a correction for human vision using the standard daylight or photopic luminosity function.  Normally these meters are used for workplace or occupational monitoring and are adjusted for indoor use (a range 0 - 15,000 lux approx.).  So they are not ideal for outdoor light metering, where bright sunlight can achieve up to 100,000 lux.  I used the meter on a fairly dull day where the ambient illuminance was at most a mere 6,000 lux.

The Camera Versus The Human Eye
In a recent posing HERE I explored some of the key differences between the camera's light sensitivity and that of the human eye.   Camera sensors have a limited spectral sensitivity which differs significantly from that of the human eye.  Camera manufacturers create their own functions and algorithms to bring digital images more in line with what the human eye is capable of seeing.

Brightness and Luminosity Tools
Firstly, time for a quick breather!  Okay so far I have established that light is typically described in very subjective terms but a very specialist scientific lexicon also exists to describe and accurately measure light in all its complexity.  Radiometry is the study of the overall electromagnetic spectrum while photometry concerns only that narrow portion which we call the visible spectrum.  Thanks to the work of the CIE we have a fairly clear picture of how a typical human perceives light and colour and digital camera manufacturers have used this understanding in the design of their imaging systems and processing algorithms.

I now turn to a couple of oft misunderstood lighting terms used in image processing.  Simple image brightness tools increase and decrease the brightness level or value of each pixel linearly across all three colour channels.  So the adjustment is not spectrally dependent.  They are a fairly blunt instrument that don't take into consideration the particular characteristics of human vision.  Luminosity on the other hand does take into account human perception of brightness.  Once again, the luminosity function describes the average spectral sensitivity of human perception of brightness.  For every one blue cone and one red cone in the fovea of the human retina there are two green cones.  This in part means that we perceive green as being brighter than the other primary colours.  The luminosity slider used in Adobe Photoshop, MS Paint etc. is not a simple linear slider but takes account of our spectral light sensitivities and makes a weighted brightness correction across the three colour channels.  This explains why when we observe each of the colour channels of an image in monochrome they all appear distinctly different from the final compiled image. This is all explained in better detail by this Cambridge in Colour posting on Luminosity & Colour.

The distribution of tones across each colour channel must take account of the particular spectral characteristics of human vision.  This means that the greyscale appearance of each colour channel will look slightly different from the combined greyscale image, as illustrated above.  Colour channel layers can be created in Adobe Elements as I have done using this procedure linked HERE.

Incidentally, just to confuse things the term luminosity is used in another context in astronomy to denote the total amount of light emitted by a star per unit time.

No comments:

Post a Comment