topics: |  vision | biology | optometry |

Light intensities

The human visual system operates over a remarkably broad range of light
levels (Table 3–1). At one extreme, we are able to detect a star on a dark,
moonless night, while at the other, we can see a jet flying in the bright
midday sky. This constitutes an adaptational range on the order of 10 log
units (Boynton, 1979). [10 x]

[the pupil changing in size from 3mm to 9mm only increases the light input by
9x, so] changes in pupillary area account for only a small portion of
adaptation, approximately 1 log unit out of the 10-log-unit range.

Visible light intensities (candelas / m2): p.26

Tissue can get damaged:
	sun: 1010
	[intensity > 107 can damage tissue]

photopic vision (10² to 10⁶) : optial acuity ~ 10³
	100W filament: 10⁶
	paper under sunlight: 10⁵
	this book page under indoor lighting : 10²

mesopic vision (10^-2 to 10):
		moonlit paper: 10^-1

scotopic vision :  starlit paper: 10^-3
		  threshold light: 10^-6

1 candela = 1 lumen/steradian [energy / solid angle]

illuminance: lumens per square meter (lux) [or lumens/sq-ft = foot-candle]

Photoreceptors

Photoreceptors are specialized sensory receptors containing a photosensitive
pigment that absorbs light quanta, converting this radiant energy into
electrical activity. This is the first step in vision.  Both rods and cones
are slightly depolarized relative to a typical neuron. Rather than
manifesting a resting membrane potential of -70 mV, the potential is
approximately -50 mV. When exposed to light, photoreceptors
hyperpolarize—their potential goes from -50 mV to a value closer to -70 mV
(Tomita, 1970). You may find this surprising because stimulation is typically
thought to cause depolarization rather than hyperpolarization.

The degree of photoreceptor hyperpolarization is related to the intensity of
the stimulus, with an intense stimulus causing greater hyperpolarization than
a less intense stimulus. This is one reason that the potentials produced by
photoreceptors are referred to as graded potentials.

A summary of the steps that lead to the hyperpolarization of a rod are
outlined in Fig. 12–9 (Lamb, 1986; Pugh and Cobbs, 1986; Stryer, 1986). In
the dark, sodium ions (Na ) flow into the rod outer segment through pores
(channels)—the so-called


    Scanning electron image of rods (larger segments) and cones (smaller)
    in the tiger salamander. In humans, the rod and cone outer segments are
    more similar in size.  p.28

Rod Photopigment

The photopigment rhodopsin is contained within the discs of the rod's outer
segment.2 A disc contains approximately 10,000 molecules of
rhodopsin. Because each rod has approximately 1000 discs and an eye contains
120 million rods, there are approximately 1015 molecules of rhodopsin per eye
(Boynton, 1979). Each molecule of rhodopsin is capable of absorbing one
photon of light, and the absorption of one photon is sufficient to activate a
rod. The large number of rhodopsin molecules provides the eye with a
tremendous ability to capture light and contributes to our exquisite
sensitivity under nighttime lighting conditions.

A molecule of rhodopsin becomes bleached (i.e., transparent) when it absorbs
light. The absorption of only one quantum of light is required to bleach a
molecule of rhodopsin (Hecht et al., 1942).... The half-life for rhodopsin
regeneration is 5 minutes.

the rhodopsin absorption spectrum shows the probability of absorption
(indicated as relative absorption on the ordinate) as a function of
wavelength, in more detail (Fig. 3–4C). Quanta of 507 nm have the highest
probability of absorption. This is due to quantum mechanics: the rhodopsin
molecule and a quantum of 507 nm "fit together" well, thus increasing the
probability of absorption.

Other wavelengths are absorbed, but with less probability.  e.g. one may
assume that the rhodopsin absorption curve gives a probability of 0.20 that a
quantum of 507 nm will be absorbed and a probability of 0.10 that a quantum
of 680 nm will be absorbed.

Once a quantum of light is absorbed, all information regarding its
wavelength is lost, a principle referred to as univariance.

One may dark-adapt an individual by asking him or her to sit in a dark room
for 45 minutes, thereby maximizing the regeneration of the rhodopsin.
Subsequently, the minimum amount of energy required for the person to
detect stimuli of various wavelengths is determined. The minimum amount of
energy required for detection of a stimulus is referred to as the threshold
for that stimulus.

Cone Photopigments

In a typical human eye, there are three fundamental cone photopigments,
cyanolabe, chlorolabe, and erythrolabe, which show maximal absorption at
approximately 426, 530, and 557 nm, respectively (Fig. 3–6A).4 Each cone
contains only one photopigment. p.33

It is common to speak of three different classes of cones, each containing a
different photopigment. The cyanolabe-containing cones are referred to as short
wavelength-sensitive cones (SWS- or S-cones), the chlorolabe-containing cones as
middle wavelength-sensitive cones (MWS- or M-cones), and the cones containing
erythrolabe as long wavelength-sensitive cones (LWS- or L-cones).

The cone photopigments recover from bleaching at a faster rate than does
rhodopsin. It takes 1.5 minutes for 50% of cones to regenerate their
photopigment.

Stages that lead to rod hyperpolarization:

11-cis retinal ---[photon]-->
	11-trans retinal --> activates Transducin (a protein)
		--> activates Phosphodiesterase (PDE)(a protein)
		--> Breaks up cGMP into GMP
		--> Na+ pores close
		--> Rod hyperpolarizes

Purkinje Shift

As lighting conditions change from scotopic to photopic, the wavelength to
which we are most sensitive increases from 507 to 555 nm. This is the basis
for the Purkinje shift, the relative increase in the brightness of longer
wavelength stimuli as lighting conditions change from scotopic to photopic.

Retinal distribution of photoreceptors

The human retina contains approximately 120 million rods and 6 million cones.
As illustrated in Fig. 3–8, rods are most densely packed at approximately 20
degrees from the fovea, where they reach a peak density of approximately
150,000 rods/mm2. There are no rods in the fovea, which results in the
inability to see a small, dim object, such as a star, when it is foveally
fixated under scotopic conditions. Looking slightly to the side of a faint
star, causing its image to fall outside of the fovea onto the surrounding
rods, increases its visibility. Whereas the number of retinal cones may
remain stable as the eye ages, the number of rods decreases (Curcio et al.,
2000). p.37

Unlike rods, M- and L-cones are most concentrated in the foveal center, where
their density ranges from approximately 115,000 to 225,000 cones/mm2 (Putnam
et al., 2005). Although the density of cones is substantially reduced outside
the fovea, they are present throughout the retina. About only 5% of the total
number of retinal cones are located in the fovea. A similar percentage of
retinal ganglion cells are located in the fovea (Azzopardi and Cowey, 1996).

The ratio of L- to M-cones varies from person to person and has been found to
range from 1:1 to 16:1 in individuals with normal trichromatic vision (Roorda
and Williams, 1999; Hofer et al., 2005).

S-cones show a different retinal distribution than other cones (Calkins,
2001).  Not only are they considerably less numerous than either M- or
L-cones, constituting approximately 5% to 10% of the cone population, they
are not found in the central 0.3 to 0.4 degrees of the human fovea (Curcio et
al., 1991; Roorda et al.  2001). Peak density, approximately 2000 cells/mm2
is at approximately 0.5 degrees from the foveal center.

The high density of cones in the human fovea provides the basis for excellent
visual acuity and much of the richness of our visual experiences. Occupying
only a small percentage of the area of the retina (0.01%), the fovea encodes
a disproportionately large amount of information (Azzopardi and Cowey, 1993).

Light adaptation

When you step outside on a sunny day, the amount of light reflected into your
eye and focused onto your retina increases by a factor of several
thousand. In spite of this tremendous change in light levels, the appearance
of objects remains the same (e.g., your classmate's hair appears black
whether viewed indoors or outdoors on a sunny day). Within a very brief
period—so fleeting that you are unaware of it—your visual system adapts to
the changes in illumination levels, a process referred to as light
adaptation.

Within very dark light ranges, there is no change in adaptation... The
background is practically black, and internal neural noise produces as much
socalled dark light as does the background itself11 (Barlow, 1956; Fechner,
1860).

For very dim light, sensitivity reflects quantal fluctuations in the
background (DeVries, 1943; Rose, 1948). The background is so dim that
fluctuations inherent in the light source that produces it play a primary
role in determining threshold. This is frequently expressed as the
DeVries–Rose law, which predicts that Delta-I is equal to (IB)1/2.

The third section, which covers a 4-log-unit range, has a slope of
approximately 1, revealing that Weber's law12 is followed (Aguilar and
Stiles, 1954; Barlow, 1965; Walraven and Valeton, 1984). As the background
brightness is increased, the increment intensity must be increased such that
the ratio of the increment intensity (Delta-I) to the background intensity (IB)
remains constant. This constant ratio, Delta-I/IB, is referred to as Weber's
fraction or Weber's constant.

The Weber fraction for scotopic vision is approximately 0.14 (Cornsweet,
1970).  If the background intensity is 100 units, the increment must have an
intensity of 14 units (Delta-I = 14) to be detected. If the background is
increased to 1000 units, the increment intensity must increase to 140 units
(Delta-I = 140) to maintain a Weber fraction of 0.14 and thus remain
visible. Although the relative sensitivity of the visual system (0.14) does
not change as the illumination increases, there is a reduction in the
absolute sensitivity (the threshold goes from 14 to 140 units). This tradeoff
between relative and absolute sensitivity is referred to as sensitivity
regulation.

While visual resolution and contrast sensitivity are much higher under
photopic viewing, the ability to detect a stimulus is much superior under
scotopic conditions.  For a 500-nm stimulus to be detected under photopic
conditions, it must be approximately 3 log units (1000 times) more intense
than is required for detection under scotopic conditions.

This is because of spatial summation of sotopic (rod) connections is much
broader - so the ganglion cell see the summation of a much larger set -
hundreds of rods - about 10 minutes of arc -
than is the case for cone cells.


	layered	structure in the primate retina


--- author
Steven H. Schwartz, O.D., Ph.D. studied optics at the School of Optometry,
University of California, Berkeley and has taught optics in both the basic
curriculum and special Board review courses. He has co-chaired the
Examination Development Commission, National Board of Examiners in
Optometry. Dr. Schwartz presently teaches at the SUNY State College of
Optometry.

Send your jottings to Book Excerptise

to contribute some excerpts from your favourite book to book excerptise. send us a plain text file with page-numbered extracts from your favourite book. You can preface your extracts with a short review.
email to (bookexcerptise [at] gmail [dot] com).