The Modulation Transfer Function of the Human Visual System at Selected Spectral Frequencies

Rochester Institute of Technology

RIT Scholar Works

Theses

Thesis/Dissertation Collections

1-1-1969

The Modulation Transfer Function of the Human

Visual System at Selected Spectral Frequencies

George L. Ayers

Follow this and additional works at:

http://scholarworks.rit.edu/theses

This Thesis is brought to you for free and open access by the Thesis/Dissertation Collections at RIT Scholar Works. It has been accepted for inclusion

in Theses by an authorized administrator of RIT Scholar Works. For more information, please contact

[email protected].

Recommended Citation

I

The

Moduiation

Transfer

Function

of

the

Human

Visual

System

at

Selected

Spectral

Frequencies

by

George

L.

Ayers

Photographic

Science

4

Rochester

Institute

of

Technology

ABSTRACT

2

INTRODUCTION

3

COMPARISON

OF

THRESHOLD

CONTRAST

TO

MODULATION

TRANSFER

FUNCTION

10

OBJECTIVES

12

FLOW

CHART

DIAGRAM

13

EXPERIMENTAL

PROCEDURE

14

APPARATUS

SCHEMATIC

15

APPARATUS

PHOTOGRAPHS

16

RESULTS

AND

GRAPHS

19

CONCLUSION

28

APPENDIX

2 9

ACKNOWLEDGEMENTS

3ft

REFERENCES

36

ABSTRACT

The

Modulation

Transfer

Function

of

the

human

visual system

has

been

determined

with a sin

usoidal

test

target,

by

measuring

the

threshold

contrast as a

function

of spatial

frequency,

for

monochromatic

light

of constant average

luminace.

Data

were obtained

for

a

broad

band

range and wavelengths of

538

and

617

nm.

It

is

shown

that

a maximum

MTF

occurs at

frequencies

between

2 0

-3 0

cycles/mm on

the

retina and

that

this

sensitivity

decreases

for

both

lower

and

higher

frequencies.

Small

but

discrete

differences

appear

in

the

curves

for

broad-band

radiation and

the

green and red

filtration

at

the

photopic

luminances.

INTRODUCTION

The

usefulness of

describing

the

human

visual system

by

the

principle of modulation

transfer

function

was realized

by

Westheimer

in

1959.

He

termed

this

measure of

the

eyes

ability

to

distinguish

fine

detail

modulated

in

a sinusoidal

fashion

the

"Contrast

Sensitivity"

of

the

eye.

Fringes

were created

directly

on

the

retina of

the

eye

by

an apparatus

based

on

Young's

double

slit

interference

method.

This

gave a system

free

from

defocussing

of

the

optics or aberrations.

The

light

source

however,

was not coherent and problems arose

in

the

interpretation

of

data

in

the

photopic regions.

This

investigation

yielded

data

based

on quantitative

description

with

direct

measurements of a

variable

frequency

sinusoidal pattern at

threshold

luminances.

4

A

similar paper

(Campbell

and

Green,

1965)

updated

methodology

and used a split

beam

of a neon-helium

laser

(632.8

nm)

to

produce

an

image

directly

on

the

retina without

involving

the

optical

system of

the

eye.

This

also gave a measurement of

the

RETINA-BRAIN

resolving

power of

interference

patterns modulated

sinusoidally

.

Recognition

of

this

pattern as an actual

grating

was

found

to

be

difficult.

Campbell

and

Green

describe

it

as

that

of

"a

scintilating

area

brighter

than

the

surrounding

field

and also

desaturated

in

colour compared

to

the

rest of

the

field.

It

is

presumably

due

to

beating

between

the

periodic

fringes

and

the

regular pattern of

the

retinal mosaic"

.

This

effect occurred _at

high

spacial

frequencies.

In

the

course of a sub-experiment

these

investigators

were

concerned with

the

resolving

power of

the

visual system and

the

effect wavelength would

induce

on

the

shape of

the

contrast

sensitivity

curve.

Filters

chosen

in

conjunction with

the

spectral emission of

the

oscilliscope phosphor

(used

to

create

the

sine-wave

patterns)

gave a

blue

and orange

test

field

(480

and

600

nm)

.

Inspection

of

the

resulting

curves showed no

differences

in

the

shape of

the

two

curves.

5

Campbell

and

Gubisch

(1966)

using

a

double

pass

technique

studied

the

effect of pupil aperture and

fundal

qualities on

the

line

spread

function.

Calculations

transforming

the

line

or point

spread

functions

into

modulation

transfer

data

require

two

major

assumptions

concerning

the

optical system.

These

assumptions

however,

are

only

partially

satisfied

by

the

eye.

The

first

is

ISOPLANASIA.

Isoplanasia

is

fulfilled

at

the

fundal

areas

but

at

off-axis portions

loses

this

because

of retinal curvature and

to

some extent on small

foveal

impressions

or pits.

This

precipates

marked errors

in

line

spread

distributions

since

only

the

average

values can

be

calculated.

These

errors can

be

large

when

M.T.F.

is

calculated

from

image

distributions

of small

targets.

Secondly

in

the

transformation

from

spread

function

to

M.T.F.

symmetry

in

the

light

distribution

must

be

maintained.

However,

for

astigmatized or

decentred

optical systems

it

would

be

un

realistic

to

make

this

supposition.

"On

the

other

hand,

the

light

distribution

of straight

lines

21

or grids

may

be

derived

with a good all-over accuracy".

Further

reading

suggests

that

the

majority

of experiments

14

16

17

used either a

broad

band

analysis

(white

light)

' '

or a

4

5

23

26

region

corresponding

to

maximum eye sensitivity. ' ' '

of

how

the

visual system responds

to

different

spectral

- .

4,

24,

25

frequencies.

'

24

Schlaer,

Smith

and

Chase

(1942)

determined

the

visual

acuity

at

two

restricted spectral regions

(670

and

490

nm)

in

order

to

gain maximal separation of

the

rod and cone

functions

.

The

first

test

target

used was a

broken

circle

(C)

with

the

break

equal

to

the

width of

the

line

forming

the

circle and

one-fifth

its

outside

diamrter

and

the

second

field

was a

grating

having

opaque and

transparent

bars

of equal widths.

It

was

found

that

the

red

light

data

which represents

relatively

pure cone

vision

fell

on a single continuous curve.

Blue

light

data

on

the

other

hand

fell

into

two

distinct

groups.

Values

below

a certain

luminance

value

(0.03)

photons)

represented rod vision and

those

immediately

above combined

the

rod and cone vision

to

produce an

effect of

superadditivity

of

the

mixture.

This

area produced a

higher

response

in

terms

of visual acuity.

These

investigators

emphasized

three

major

limiting

factors

of visual acuity.

The

primary

factor

is

pupil size.

Since

the

variance of pupil

diameter

causes marked variance

in

visual

acuity

especially

in

the

longer

wavelength radiation careful consideration

must

be

given

to

this

factor.

Schlaer,

Smith

and

Chase

maintain

that

a pupil

diameter

of

3mm

is

sufficient

to

tansmit

the

zero

order spectrum of

the

test

grating

plus at

least

one

first

order

spectrum and also give maximal

acuity

data.

Secondly,

inhomo-geneity

of

the

light

source creates chromatic

fringes

in

the

image.

Finally

it

is

theorized

that

under certain

conditions,

the

diameter

of

the

foveal

cones

may

constitute a

limiting

factor

ii

the

geometric retinal

image

is

in

fact

smaller

than

the

actual

25

In

1965

Van

Nes

and

Bouman

studied

the

effect of wave

length

and

luminance

of

the

visual modulation

transfer.

Data

was obtained

for

three

values of monochomatic

light

(450,

525,

650

nm)

of constant average

luminance.

Threshold

modulation

was

defined

to

be

the

average of

the

"grating

just

visible"

and

"no

grating

visible"

luminances.

The

results obtained

were

for

the

right eye of one subject who was

3

diopters

myoptic.

The

target

itself

was viewed

by

a

chemically

accomodated eye

through

an artificial pupil and external optical. system

comprised of

five

elements.

Van

Nes

and

Bouman

found

"there

is

little

differences

at

photopic

luminances

in

the

modulation

transfer

function

curves

for

the

red,

green and

blue

wavelengths.

However,

it

is

left

to

the

reader

to

interpret

any

differences

in

the

data

curve

functions.

It

would seem reason

able

in

most

instances

to

make

decisions

about and predict

differ

ences

in

curves

based

on

the

fact

that

they

are

drawn

separately

yet quite

closely

together.

The

authors

have

employed no

statistical analysis

in

organizing

the

data

to

clarify

or

repudiate conclusions

they

have

extracted

from

data

obtained

experimentally.

Quantitative

methods were employed

to

obtain

the

data

which was

subsequently

subjectively

analyzed

for

the

presentation.

Again

in

this

region of experimentation

there

has

been

no

estimation of "between" and "within" subject variation.

Certainly

there

must

be

some observer error

involved

and a

calculated statistical analysis would present

this

information

clearly

to

the

experimenter.

data

in

a statistical manner

in

order

to

clarify

and substantiate

the

conclusions

drawn.

It

is

felt

that

an analysis of variance

technique

coupled with observer replication would yield a

valuable estimate

to

both

curve

functions

and within subject

variability.

This

technique

would also allow

investigation

of

between

observer variability.

Many

of

the

previous

investigators

involved

atropinization

of

the

eye with

the

addition of an artificial pupil and external

i 4.

3,

4,

5,

k8,

24,

25,

27

_. . . . ,

optical systems.

It

must

be

recognized

however,

that

the

human

visual system which

has

been

chemically

accommodated and with external apertures

does

not

necessarily

conform

in

any

known

manner with

the

eye under actual

viewing

conditions.

Depalma

and

Lowry

recognized

this

fact

'

and

theorized

that

since pupil

diameter

versus

luminance

functions

have

been

extensively

investigated

'

a stringent control

of mean

luminance

value

in

the

system would afford a similar

control of pupil

(aperture)

diameter.

This

control

does

not

take

into

account several major

pupillary

reactions which must

be

15

considered

if

no chemical

influence

is

involved.

The

pupil reacts

to

near vision

causing

a contraction of

the

pupil which

is

with

associated in some

mannerA

the

convergence of

the

eyes ana accommoda

tion

of

the

lens.

Spontaneous

thoughts

or emotions elicited

by

stimulii either

by

the

apparatus

in

the

experiment

itself

or

consenu-ally

cause

pupillary

dilatation.

Spontaneous

or reactive changes

in

the

level

of consciousness or

neurologically

speaking

changes

in

the

degree

of cortico-thalamo-hypothalamic

activity

affect

the

pupillary

reflex

dilation.

It

increases

with emotional stimulii

and

decreases

with

fatigue

by

increasing

or

decreasing

the

the

iris

.

Pupillary

movements can also

be

caused

by

an

intentional

short

closing

of

the

eyelids.

This

Lid-Closure

reflex consists of a

short contraction and

then

redilataticn.

Finally,

there

exists

apparently

spontaneous oscillation of

the

pupil under

steady

conditions of

illumination

which

have

no

direct

liaison

with

any

other reactions.

When

the

eyes are exposed

to

long-lasting

light

stimulation

the

pupils

contract,

then

redilate

partially

and

begin

tc

oscillate.

In

dim

illumination

as

for

threshold

readings

the

pupils

become

quite

large

and quiet

as

they

would

be

in

darkness.

As

the

intensity

of

the

stimulation

is

increased

the

mean

diameter

of

the

pupil

becomes

smaller and

the

rate of oscillation

increases

accordingly.

It

would appear

then

that

there

are

two

alternatives

for

eliminating

the

interference

of

pupillary

oscillation.

The

first

would

involve

a pre-experimental

investigation

to

test

the

hypothesis

that

there

is

no

insignificant

variation of pupil

diameter

for

the

observer under

the

experimental conditions.

This

would

involve

sophisticated and expensive apparatus and

cannot

be

employed

in

this

particular

investigation.

The

second method

involves

the

dilation

of

the

pupil

chemically.

This

can

be

achieved with

only

a minimal affect on

accommodation

by

using

a mydriatic

drug.

Each

observer would

have

to

be

individually

tested

as

to

the

exact effect

the

drug

would propagate.

This

method,

although

using

an artificial pupil

for

exact control of pupil response would eliminate external

pieces of optical apparatus.

It

may

be

argued

that

we now

have

created an unnatural

conditions.

However,

control over pupil size must

be

carefully

maintained either

by

an appropriate apparatus

included

in

the

instrumentation

to

maintain a

running

control or

by

removing

the

psychological and pfeiological

incongrueties

from

the

system.

In

this

respect a more accurate

study

of

the

viewing

conditions can

be

estimated.

There

will

be

no need

to

speculate

what

the

effects will

be

if

external optics are added

to

the

system

tc

maintain

focus

for

the

chemically

dilated

pupil.

The

fact

that

any

optics

between

the

object space and

image

space

must

degrade

the

sinusoidal

light

distribution

is

a major

factor

which cannot

be

disregarded

as

has

been

done

in

many

investigations

It

is

felt

that

the

latter

course of action would yield

the

most consistent results

but

it

involves

constant medical attention

during

the

process.

For

this

reason past

data

gathered

by

other

/-no 0*3

experiments ! ' must

be

relied on

to

supply

adequate

COMPARISON

OF

THRESHOLD

CONTRAST

TO

MODULATION

TRANSFER

FUNCTION

The

modulation

transfer

function

of a

transilluminated

system

is

defined

as

the

contrast ratio at

the

modulation of

the

output

to

that

of

the

modulation of

the

input

of a

sinu-soidally

distributed

target.

This

approach

is

widely

applied

in

the

photographic

industry

to

the

response of a

receiving

layer

(the

emulsion)

or

the

light

dispersion

characteristics

of an optical system.

Provided

there

is

linearity

in

the

system

any

number of componqjb parts can

be

deduced

from

sufficient

knowledge

of

the

aggregate.

Another

fundamental

advantage of

using

a sine-wave

target

is

that

the

modulation remains sinusoidal

even when

imaged

by

a

poorly

focussed

or

imperfect

optical system.

In

this

type

of physical arrangement a measurement

is

made

of

the

output magnitude

for

a constant

input

magnitude of

variable

frequency.

In

this

context

it

is

difficult

to

correlate

the

purely

physical system

to

a psychophysical visual system.

Although

the

input

values are physical measurements of

the

light

modulation of sine-wave

test

target,

the

output values can

by

no means

be

described

with

this

extreme

accuracy

since

it

is

a

quantity

which

does

not permit observation

in

the

physical sense.

During

this

investigation

the

response variable

for

the

output will

be

the

reciprical of

threshold

contrast or modulation

transfer

function.

In

order

to

assume

the

linearity

needed

for

mathematical manipulation of

the

function

certain conditions

need

be

clarified.

One

major supposition

is

that

the

internal

noise

level

of

the

visual system

is

a constant value.

It

will

high

spacial

frequency

targets

is

limited

by

an

internal

noise

source

that

the

noise

level

is

independent

of

the

different

frequencies

and

the

output contrast of

threshold

measurement

is

constant.

Therefore,

the

contrast

loss

in

passing

through

the

apparatus and visual system

is

the

reciprical of

the

THRESHOLD

CONTRAST

at

the

input

end.

Consequently

the

reciprical

of contrast

threshold

is

defined

by

the

same parameters as

modulation

transfer

function

and can

be

substituted

for

it

as

the

function

plotted against spacial

frequency.

The

concept

that

(THRESHOLD

CONTRAST)

is

a good

first

order approximation of modulation

trnasfer

has

been

further

substantiated

in

a communication with

Mr.

J.J.

DePalma

(Kodak

OBJECTIVES

PART

I

To

test

the

hypothesis

that

the

modulation

transfer

function

(reciprical

of contrast

threshold)

as generated

by

a

sinusoidal

target

of

diminishing

spacial

frequency

on

the

BASIC

FLOW DIAGRAM

E3S)

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

Each

Fixation

Point

of

F

has

R*CTl

Broad

Band

Visual

Radiation

Each

Fixation

Point

of

F2BX

has RxC

Analysis

of

Variance

Conclusion

1

1

1

I

1

1

1

1

1

1

1

1

1

Each Fixation

Point

of

F

has

RxfjTi

PART 1

PART 2

[image:15.572.9.559.17.716.2]

EXPERIMENTAL

METHOD

1.

Microdensitometer

traces

are made of

the

sinusoidal

test

target

at each spacial

frequency

to

determine

the

maximum

transmittance

Tmax

and minimum

transmittance

Tmin.

The

method will

follow

closely

that

of

Depalma,

J.J.

and

Lowry,

E.M.

(1962)

which made use of

the

following

quantities,

B

max

B

.

mm

max

T

.

mm

B

o

V

o

V

r

M

P

O

B

VL

= _maximum

luminance

in

test

_{object at}

threshold

= _minimum

luminance

in

test

_{object at}

threshold

= _maximum

transmittance

in

test

target

-minimum

transmittance

in

test

target

=

luminance

_of

test

target

= _spacial

frequency

of

test

target

cycles/mm

= _spacial

frequency

of retina

=

magnification

= _object

distance

(mm)

= _effective

image

distance

(mm)

=

filtration

(nm)

=

SCHEMATIC

OF

MTF

INSTRUMENTATION

(-

40

in.

h

14

in.

E

M,

Beam

Splitting

Mirror

M2

Front Surface

Mirror

H

Integrating

Box

L

Tungsten

Light

Source (300

watts)

0,

02

Opal

Glass

Diffusing

Screens

F,

Test

Filters

F2

Neutral

Density

Filters

T

Sinusoidal

Test Targets

with

viewing

aperture

included

Figure

2

(15)

The

test

targets

are

inserted

in

the

front

filter

assembly

of

the

integrating

cone

(fig

2)

designed

so as

to

be

evenly

illuminated

in

the

200

candle power range.

The

binocular

observation

distance

is

fixed

at

2

inches

from

the

viewing

post or

14

inches

from

the

sinusoid

target.

Instead

h

of

2

fixed

position reostat controlled

light

sources,

a single

300

watt,

moveable

tungsten

filament

illuminates

the

target.

System

components consist of

3

front

surface mirrors mounted

at

45

degree

angles

(M~),

opal glass screens

(0,,

0)

and a

set of

filters

(F,

,

F

)

.

The

opal glass

forms

a

uniformly

illuminated

background,

eliminating

the

filament

image

on

the

transilluminated

target

(T)

and

the

veiling

luminance

surround.

Filters

F,

consist of a

Wratten

No

58.,

Green

and

Wratten

No.

25

red as well as a

broad

band

analysis while .F2 consists

of a series of graduated neutral

density

filters

used as needed.

The

target

frequencies vary from

a coarse

3/8

c/m

to

6.0

c/mm.

The

target

size will

be

1

m.

in

length

by

1/2

cm. width and

the

vieling

luminance

will provide a surround of

2.54

cm square.

The

surround will

be

of

the

same spectral content as

the

target.

The

observer views

the

target

and

fixation

point

through

the

vieling

luminance,

an opal glass

diffusing

screen,

filters,

and

a

beam

splitting

mirror

(M,

)

.

The

purpose of

the

vieling

luminance

is

to

provide a control of

the

surround content at a

fixed

level

for

each

test

target

and

to

provide a means

by

which

the

contrast can

be

varied.

The

only

optics

in

the

way

of

the

object space as seen

by

the

observer

is

the

partially

transmitting

mirror

(M,

)

.

This

obstacle can

be

mathematically

removed

from

the

system once

its

MTf

value

is

determined.

In

making

a

threshold

observation,

the

carriage of

the

moveable

filament

is

placed so as

to

give

the

test

target

the

maximum contrast and visibility.

The

rooms

lights

are

turned

off

and

the

subject

is

required

to

wait

60

seconds

before

the

test

target

is

illuminated.

When

the

apparatus

is

switched on

the

subject views a

test

target

with

both

eyes with

the

surround

and

vieling

luminance

at a minimum

brightness.

The

carriage

is

then

slowly

moved

forward

to

increase

the

vieling

luminance

and

reducing

the

target

to

threshold

detectibility

.

The

filters

are

the

removed.

When

a

threshold

setting

has

been

accomplished,

photometer

readings are made of

the

vieling

luminance

BV?

and

the

cone

luminance

B

.

o

The

modulation

transfer

is

described

as

.__,

B

T

+

B

T

.

+2

BVL

|

MFF

= _o

max o mm J

VL

.

B

T

-B

T

.

(

'

o max o mm

In

investigating

the

effect of

the

MTF,

it

is

most

usefully

expressed as a

function

of spatial

frequency

on

the

retina,

V

.

The

spatial

frequency

on

the

retina

is

computed

from

the

spatial

frequency

in

the

test

object

by

the

following

relationship:

V

=

V

_:

1/M

=

V

.

Po/1-,-r

_/tv,

r o o

U/PI

(B)

In

calculating

V

for

each

test

object and

for

each

viewing

situation,

the

usually

accepted value of

17

mm. was used as

the

effective

image

distance

PI.

This

is

the

distance

from

the

second

nodal point of

the

eye

lens

to

the

retina.

RESULTS

All

results are

for

binocular

vision of

four

observers,

each

having

individual

eye characteristics.

Figures

3-6

give

the

response of each observer

to

the

test

conditions

for

broad-band

radiation as well as green

(538

nm

)

and red

(617

nm)

filtration,

The

horizontal

axis

has

been

mathematically

remodelled

to

read

in

retinal spatial

frequmcy

rather

than

that

of

the

target

frequency.

The

three

parameters all produce a maximum

in

sensitivity

or

minimal

threshold

at about

20

-3 0

cycles/mm

(

Vm

)

.

For

photopic

luminances,

thresholds

for

spatial

frequencies

smaller or greater₃

than

V

are

higher.

m 3

Figures

7-9

show

the

response of each observer

for

each

filtration

level.

All

curves

take

the

same general shape

but

are

displaced

according

to

the

specific observer

characteristics .

OO

9Ql_

8Q__

60 F

lip

= a -- = = m

i(INN

Wk=

AUGUST

Tmza-._. _

RRRvmE

^

,

<

&

~r~

<

so '(*_Nn

..

V.

iJ' _ <^" A ?o ,<> . p (n 1it r

)

!,

n

*\

1Q_ 9-=

8-1

7_l 6_F =

|

E

_

E

=

1

_mE

^5

1

=

11

111

i^ ~Er^j-I- '-s

[-=ttH

_v

8

~.

\

^a~^ < > 1 <

\

3

"A;

i n on /ip> cn z.r\ _{7C 8!) o}r\

ju uu i>i^ ₁

.'.\J 1JO

\

:

*<

>f'/kl \A .

1 R b

L

u t

!<- 1 -

H

_ETir

J/^

j.

IV 1 aL. RIOr

IrFl

r i-/a

/.'V.1A>

EcaJu

ni_Jliih_{! Ul An}r>i/ f nin CP_tic

-J-4n II i _average

1

g

Ul eJ

10

p|

=3= ;_ ==T= =

i

j

i

---i

j |

-.--=-.. . mm i = 8 =E-L 7

1

6 ===~ ~

1

)^

=

MORETON

-

11 AUGUS

1

III!

4 "

s

*

IE

>

l

E

. V

y / s

a rt ,

s

A,'

IV1 _V.

'/

\

,

SL

m >

v

r

A

t

\

\

< i

-iT-^ 1 "^ 1

V

]

"

\

L '

e

R

? fc>v^

s s s_V

5

s,

p

\

(j

\

< i

JJ=L

m

-|

11

a=

i

~

R

1 1

=

i

m\

llllllllllll

8-== =_E

6_EEE;

liill

^_

*** *.

i

5_=L_: z-Sal^

3

IB or n Ii

Sr

zt? JCj 4 o Jo o Ij 7u 0 V 1'J\l 1 1 u Izu ijo

SI

>A

T 1/ J

-1

ti

(,

i h l\

(

Y

R b

1

n M A

p

n ji

G 1/

7

M

1 b

*

..

4

TI nr

h

_qUl e

4

u -~ -_

Ei

McEl 1 cr.

R

0

12 AUGUS

8 E 7 = 6 E =Z

-Y

m^ 5 4

i

-!< ============

s

S

JiS'

/

a N

/

\

/ s

/

v\

/

(i -. U

j 1

:>

_\ <

/

N

? -/- ,

>.

1

*/

t1

/

Cr

f

r

f

/

e-/

1

J

f

r\ " \il -HfJ v

O

10 9_!ama^

HI

|

1

1

i!

aa

-|

m

i

=-I =-I

a

||

II

>

1

I"

|

I

E

N,_ - '

^di

8_:

E

=

E

E~

^s*-

...

=====

^

;==]==i=^^a

5-J:

u-^i =r+

^

s^

3

l n Of- ifi ii D /-I t,n T

K

a_o _ !__. _

Jo 9u PLc

1 0

1

20

1 JO

( 1

>r

fi 1 i/

,i

1 !l

K

h

1

r !/

c

N\Yo P

IC^ _<\

G H T

4 C

_{f t}

-th

1

;

/

s

UQ.

F g

ure

5

9ft_

_

===!==1=1111===

^OUTHV

1115

VELL-

9

AUGUST

o

30

G^f

-3

\ \

t

,

t

JT6r

j /

SL

_/ ^

^----=h

"

j '

c

R

<st

V

\\

4l

'\

10

_ -:

iiiiiiiiiiiiiil

^~_^ii====~=^==* ^=

jpii

liiiiiiiiiiiiiiiiiii

=Ey ==m|3^=

m=F=p

= = = =

==~ ======_3J===

=3E3^

7_~[E:pm =

IfEE

~ -m17

=======111=====

H

_EEE==iEa=--=.-.m=_-.-.

IlllllUll

111111II11

6_== = = = =

= == === ==== =

^^E

i~ . .. ..

L J

-jsam>^ fRE

1 1

5 i

At

\aV-v

V **>

^_c

^_. \ p^

B V

3

?

>.

IO _{zJ JO 40 DO} i.r\ 7 ~\ m ori₎ i?_itFU nu lzv_ ₁Jv_J

<naTIACa FfiHonrEtscr - _rettst

<V

:

MYiaaS:

R

GHT

-5-50 + 0.50AXIS 90

LEFT -4.00

+

-OOAXIS

80

FU

jure

6

oo

. , i

i

t

i

i

i

"TT^l

i : --9Q .

"

^

1

M

" '"

f"

i.-: ~

!

(=

E

-h

^=^^|E

s_

E;

---================= 70 1 ===jE4^==^=E^m-=E=m-=m-~-j1-+4= _ ^_ ---ea

1^=^

j^^^n

nm

'

3ROAD-BAND

ANALYSIS

r*

-f--^l '

2

_MORETON

2

* *

^m5

____=_

__

m^Jmm. ___.

40 -j^-jS^

"Si^5>

'~^~

3=j-f

SOUTH

WEM^

1

../

^

^_

~S

"

:

t n/n

^eei>^^e:--HeN^

fW4

?

V.S%

z

*v

3

/

4

$sV

ZjZ'J.

V

\\

*

I

"

^

O.

*

a

kE

O

S

s

X

N

-3 V>- r-

V

?o r ' >. _,

r^s^L

... \ ,

k n.

h*-i ~

EE^EEEEEEEEEEEEEEEEEEEEEEE^EEEEEEE

5

K_ ______

_*

E . _,

^5E *

~v

_{r '_ L,}

^S V

'

V

^S

%

t

> <v *.

i

-V ^

X

3_.

l _ .

zs:

-EV

" '

a

₁

^

X

-

vE-^

\

io -

\

\

%

_

iiplII||jI[|||jiIIIIIIlB

=====l|=!p==~|~|~|

^^^JE^=p=rimE^p-_m-^

lpiiii|i|liii|iiiliBii

5

TrrmP^frrffl^T^JW

^ffliffl^^H

-*s

4^-=!=

-~E=

1

l-sth-1 _{| n} _

~

_ >*.

V

.

. c

""' A;

\

^

V

3 . \

J BO vO VOO 1IO 120 1JO ->,

I-lOO

__^_____^

oft

| |

..[_i

1

!

1

j

1_

j

80

)

'

i i , ' , i

--o

____e_-~Fr

i^

6Q_

GREEN FILTER

-No.

58

^== =||=============

^E

*

i^^^^=--

-'-^^^^^L=U^===^^^^

Am * V.

4 *

/ * _

3Q ,

/

/

__,__m _ .. ___

^^_

_____ . p 1 i r

1 f^"^

^L _^^

_U_

***

r

\

^i ^

*

/

*

^*^v ' '

Ej

\ v v ""*

a\

l

* _>V ?

J-

_-

^s_

"

/

^

*_s *

-I

\

^

^f

O

^L

^V-^

. t. ?

f+ ₁

J- __*

___\_-1 * _.

oi _ ^_

\

\ *>.

* L ' _\

*

^^fc

V

\ *

^_

\

* k

\

*.

^V

\ *

^V.

fc '

%

*S

L_

*

^V

*

^

' >w

io

V

y

"B

^sJ

liiiiii=iiiiiiii=iliiiiiiiiiiiiiiiiiiiiiiiiii

Jll_E-lHl_riis9l

6_- ~ ____ ~ ~ ~ __

=J^

~~

\-y--Z. =fc_t= ~

j|^_3__;

*

%*%

~^\1

-m

rrrrrrH

nTTrH4fHtt~H==f

-h

LH_HJmKqa

3 ""

1G

w ju

4o

M-OO 70 8

3

9D

ICC

EHO^

120

ICO

1

,\

pfr'Ti

_

___e3J

*_

gMllpF'

i

r-^'-

"-r

psjgijipiiMiiijliijI

. ' < , =

I>Q_ -EEfEp

^==^=

7D

=====

60 i -i

-. ~t: >

1 ' _i - KtU rILl cK No. Id

\ -pp_ ₁

E=

:_|______

T

~

KLEIN

:

-rj'zrp

MO RE TON

tH

TTi

I**l"

nTFHEH

as\ _i *____

._^_ _4._b -30_ *~

"fr

* ^^^ __. *

, y Xs**- *

*

_r __i^__

*

f

w *

^*^_,

?

~~

'_

^^^k.

' _L

' _>f ~*' '

^

^

?o

\^^

V

?,

ejI

V

-< _s, '

"/' "

"^1

^1~^

^.

_ ^^ ^

/

L _ ^17

1

\

k ^

^K l_

. _\ ^.

V "\.

I 1

*"

_1^L

I _

'

_

\k

J '__ *

^

t

V.

1

_ ** x

*i

^ i >v

i;

\l

N^

-_==__=

:__p_p

z= =r

~ ~~

~-z

~~ ~ ~

z$ zz ~ zz~^~ ~~ zS

IIIIIllijllllIIlllIIIIIlEiiiiEiiiim^MEIEIIiilEiililllllliiii

<-E

EEE-E

5EEEE

EEE

rE-E--Ei(EitsE-E--~--p-E

EEE-E-E

E

Ml

ErffmEH

I

rETrNwrHa#ml^

I

I

I

I

EE

EE

t _:

a"i

a

1J *} okj hw J^ o\J tj ov- t\j i uu IIU \ 2.\j IJO

S'AiIAl

EEE_____

-Rt

isia

EguieS

i __

II 1 1 1

1 1 1 1 1 1 1 1 1 1 1 1

1

1 1

1

1

TF

-CONCLUSIONS

It

has

been

shown

that,

using

the

Contrast

Threshold

criterion,

the

Modulation

Transfer

is

a

function

of

spatial

frequency

and spectral content and

has

a max

imum

sensitivity

at about

20-30

cycles/mm.

From

this

maximum

MTF

,

the

sensitivity

to

the

sinusoidal

test

target

decreases

for

spatial

frequencies

of

both

higher

and

lower

content.

This

relationship

was maintained

by

all wavelengths

tested.

This

maximum response surrounded

by

a

decreasing

response suggests

the

interaction

of

two

distinct

mechanisms.

This

could

be

explained

by

assuming

a

low-pass

component associated with

the

optics of

the

eye

lens,

ocular

media,

pupil

diameter,

and

the

retinal

mosiac

diffusion.

By

experimentation

Flamant

has

deter

mined

that

the

high-pass

component of

the

visual system appears

to

be

intimately

related

to

the

inhibitory

processes

taking

place

from

the

retinal mosaic

to

the

brain.

This

is

composed of a

complexity

of

interrelated

mechanisms;

neural,

chemical,

electrical,

and psychological,

There

is

a small

but

distinct

degradation

of

the

observer

MTF

as

the

spectral

band

is

shifted

from

the

bare

filament

to

green

to

red.

This

can

partially

be

explained

by

realizing

that

the

retina

is

a

thin

sheet of

interconnected

nerve

cells,

including

a mosaic of photoreceptive rod

and cone cells.

The

cones

function

in

daylight

or photopic

conditions and

have

a peak

sensitivity

to

radiation of about

560

nm while at

617

they

have

a much

lower

response.

The

retinal surface around

the

fovea

is

composed almost

entirely

of cones and so

the

response of

the

eye

in

this

area

is

completely

dependent

on

the

cones.

This,

combined

with

the

possible

interaction

of

pupillary

diffraction

APPENDIX

OBSERVERS

1.

Miss

Jo

Ann

Klein

Age

2 0

Myopic

Right

-1.75

+

2.50

Axis

10

Left

-

1.75

2.

Mr.

Wayne

Moreton

Age

22

3.

Miss

Louise

McElroy

Age

19

Myopic

Right

-4.00

Left

-2.75

4.

Miss

Janet

Southwell

Age

19

Myopic

Right--

5.50

₊

0.50

Axis

90

Left

4.00

+

1.00

Axis

8.0

TARGET

FREQUENCIES

3/8

3A

1

1/8

1

1/2

2

1/4

3

3

3A

4

1/2

6

Vo

RETINAL

FREQUENCIES

7.9

15.9

24.8

31.8

47-7

63.6

79.5

95.^

127.2

Vr

PUPIL

SIZE

1.5

Y

10.8

=5.15

Trolands

r

According

to

Lowenstein,

0.,

and

Lowenfeild,

I.E.

,

5.15

Trolands

would give a pupil

diameter

of

3.4

mm,

This

would of course

be

only

the

average pupil

diam

FILTRATION

FILTER

DOMINANT

WAVELENGTH

LUMINOUS

TRANS

MITTANCE

WRATTEN

58

538.

2nm

19.8

%

WRATTEN

25

617.7

nm

22.5

%

WRATTEN

4 9B

457.7

nm

0.11

%

The

blue

filter

was not used since

it

caused

defocussing

problems with

the

optics of

the

eye.

SAMPLE

CALCULATION

SHEETS

-SEE

figures

10,

11.

Only

two

runs

for

each observer could

be

made

in

the

time

limits

placed upon completion.

Each

run consisted of

120

pieces of

information

and

lasted

for

two

to

two

and one

half

hours.

For

this

reason

the

analysis of variance

technique

was not employed.

The

instrumentation

and observer error varied with

the

wavelengths of

light

used and with

the

sinusoidal

target.

An

averaging

technique

was employed

in

order

to

plot

the

final

curves.

Figure

12

shows an average

MTF

for

broad

band,

green and red

filtration

for

all

the

observers.

If

enough measurements

had

been

carried out

this

would

be

the

standard

MTF

response curves

for

the

average visual system.

An

interesting

side experiment undertaken substantiates

the

fact

that

a given color

looks

brighter

if

its

surroundings are

its

complimentary

colour.

The

Wratten

No

58

and

No

25

filters

were

again matched

in

luminous

transmittance

and

then

the

red

filter

was

placed

in

the

target

assembly

and

the

No

58

was placed

in

the

beam

of

the

vieling

luminance.

This

gave a

target

of green and red

bars

surrounded

by

a green

field.

When

the

red

bars

disappeared

threshold

to

be

directly

related with

the

general

importance

of

border

effects

in

perception.

Additional

experimentation

in

this

particular area

may

lead

to

an enhancement of

high

altitude aerial

instrumentation

visibility.

Time

did

not permit

the

use of

the

MTF

data

reduction

for

the

beam

splitting

mirror.

Additional

work should

be

done

to

eliminate

SAMPLE

VISUAL

MTF

RECORD

SHEET

OBSERVER

AGE

DATE

19

P-TEST

NUMBER

ND

FILTER

VIEWER

LIMITATIONS

AND

REMARKS

r

BROAD

BAND

N058-GREEN

No25-

RED

TRANS

vr

Bo

Bvl

i

MTF;

Bo

Bvl]

MTF

Bo

Bvl

MTF

Tmax

Tmin

3/8

. .. ...-__.

'

,

!

1

1

i

.... . ..

3A

L

i

11/2

... .. j

1

"

1

i

1

21/4

1

i

3

'

i

i 1

1

]

- -

--33A

!

i""

41/2

'

1 . .

i

6

!

\

i

i

.

J___L2J

.

. .. ..

*

OBSERVATION

CONDITIONS

FIGURE

IO

SAMPLE

CONFIDENCE

LIMITS

OBSEVER

FILTRATION

t

n

7.9

Xl

Xr>

(

X

2

!

O

j

<u

}

<

Ill

i

15.9

ill

!

24.8

1

j

j

1

1

!

1

!

!

31.8

1

1

!

1

47.7

63.6

;!

i

79.5

i

1

95.4

127.2

1

1

i

ADDITIONAL

COMMENTS

FIGURE II

George L.

Ayers

n

1

= = = =

^

=

-H

iii

I

':

m: - -- - -___ _:^ -|

: = r =[__- _n _j

j

j-5 -\-< _ ---r~.z mEE ~~

/VVl!R A G E MTF

CURVES

_

4

w

/

tufB

at

a a*

.

\

T

r1 -:=

=r

_=

P_____

^

\

*^

. V <

_t _ V

J_ ! A >

* * <

s t ft V 1 ^ m

\

2 ' *

\

i

x

I A,

:

1 >>-E|B

* 5 ~ ^r v

\

?

\

""*

' 1C

\

_fi%

\

>

"

V

k

>

l _L

E

liiiiillliil

E=7

=

II

=2

--i

|

s

;

|~\

'"i-iiiiiliiiii==l^==

1111111111!

gm

|_E

-~

-;

=E

E

1

=;

111111I1III1111I1IIII|II11II

4 :

3

>

JA 11/

ll r

kK

ui

N Z'l

'

1:

INA

2

J , . E .

inn 7CL_ __ _ =

a.

m _ ~ = E

=== ==_|^===j=^^|1=

=^=E=e|^^

/

40

/

-ii >r_i

0

-I

s

no ?o crs.

e

ss_______g__i-lo_

A

-RED

17!iRO'T

0 N Gl?F

FN

~

SI

IFtR

O

II

N

n

E

E I

R

-:

s

R

-= -= -= -= -= = = =:.

B

-GREEN

1

AFtG El C N R ElD

51IRROUND

-^

6_EEEEEEEE 5 . ========== E

E

4 3 _

11

22

2D

41.

J1

>J 6 J

/

.>

b

u

V

J

KHX

113

120

L-O

SIJ

U

_/i1

F k

:(Jl

II h K

i <

-i

M

/I

hi

^1

R

1;

r

1

1 1 II II

i

1

I

z

o

REFERENCES

1.

ALPERN,

M.

,

LOWENSTEIN,

0.,

and

LOWENFELD,

I.E.,

"Accomodation

the

Pupil"

in

Vol.

3

of

The

Eye

ed.

by

Hugh

Davson

(4

vols.,

London

Academic

Press.

1962)

pps.

191-229.

2.

ALTMAN,

J.H.

(1966)

Photography

of

fine

slits near

the

diffraction

limit.

J.

Photo

Sci.

&

Engrs.

(10)

No.

3

pps.

140-143.

3.

BARLOW,

H.B.,

(1965)

Visual

resolution and

the

diffraction

limit.

J.

Physiol

(181)

pps.

576-593.

4.

CAMPBELL,

F.W.

and

GREEN,

D.G.,

(1965)

Optical

and retinal

factors

affecting

visual resolution.

J.

Physiol

(181)

pps.

558-578.

5.

CAMPBELL,

F.W.,

and

GUBISCH,

R.W.

,

(1966)

Optical

quality

of

the

human

eye.

J.

Physiol

(186)

pps.

558-578.

6.

COBB.

P.W.,

(1915)

The

influence

of

pupillary

diameter

on

visual activity.

Am.

J.

Physiol

(36)

p.

335.

8.

DAVSON,

H.

,

Visual

Optics

and

the

Optical

Space

Sense

in

The

Eye

Vol.

4

(London,

Academic

Press,

1962).

7.

DAVSON,

H.

,

The

Visual

Process

in

The

Eye

Vol.

2

(London

Academic

Press,

1962)

9.

DE

GROTT,

S.G.,

and

GEBHARD

,

J.W.,

(1952)

Pupil

Size

as

determined

by Adapting

Luminances.

J.

Opt

Soc

Amer.

Vol.

42,

pps.

4 92-4 95.

10.

HARTLIDGE,

H.

,

(1918)

Chromatic

aberration and

the

resolving

power of

the

eye.

J.

Physiol

(52)

p.

175.

11.

HECT

and

MINTZ,

(1939)

Visibity

of single

lines

of various

luminances

and

the

retinal

basis

of visual resolution.

J.

Gen.

Physiol

(22)

pps.

593-612.

12.

GRAHAM,

G.H.

and

HARTLINE,

H.R.,

(1935)

The

response of single

visual sense cells

to

light

radiation of

different

wavelengths.

13.

IVANOFF,

A.,

(1947)

Les

abernations

de

chromatisme et

de

sphericite

de

l'oeil.

Revve

opt.

Theor

Instrum

(26)

pps.

889-919.

14.

KRAUSICOFF,

J.,

(1962)

Light

distribution

in

human

retinal

images

.

J.

Opt.

Soc.

Amer

(52)

pps.

1046-1050.

15.

LOWENSTEIN,

0.,

and

LOEWENFELD,

I.R.,

"The

Pupil"

in

Vol.

3

of

The

Eye

ed.

by

Hugh

Davson,

(4

vols.,

London

Academic

Press

1962)

pps.

231-267.

16.

LOWRY,

E.M.

and

DEPALMA,

J.J.,

(1961)

Sine-wave

response

of

the

visual system

I.

The

mach. phenomenon.

J.

Opt.

Soc.

Amer.

(51).

17.

LOWERY,

E.M.

, and

DEPALMA,

J.J.,

(1962)

Sine-wave

response

of

the

visual system

II

Sine-wave

and square-wave contrast

sensitivity.

J.

Opt.

Soc.

Amer.

(52)

pps.

328-335.

18.

OOUE,

S.,

(1959)

Response

Function

of

the

eye.

J.

Appl

.

Phys.,

Japan

28

(oyo

Buturi)

pps.

531-534.

19.

PIRENNE,

M.H.,

(1946)

On

the

variation of visual

acvity

with

light

intensity.

Proc

.

Camb.

Phil.

Soc.

(42)

pps.

7 8-82.

20.

PIRENNE,

M.H.

,

Vision

and

the

Eye

(London,

Chapman

and

Hall

LTd.

,

(1967)

.

21.

ROHLER,

R.

and

HILZ

,

R.

,

(1965)

Physical

and physiological

factors

in

the

visual modulation

transfer.

Proceedings

of

the

Colloquim

at

Delft

1965.

pps.

105-117.

22.

SCHLADE,

O.H.,

(1956)

Optical

and photoelectric

analog

of

the

eye.

J.

Opt.

Soc.

Amer

(46)

pps.

721-739.

23.

SHLAER,

S.,

(1937)

Relation

between

visual

acvity

and

illumination

.

J.

Gen

Physiol

(21)

pps.

165-188.

24.

SHLAER,

S.,

SMITH,

ELL.

and

CAASE,

A.M.

(1942)

Visual

Acvity

and

Illumination

in

different

spectral regions.

J.

of

Gen.

Physiol

(25)

pps.

553-567.

25.

VAN

NES,

F.L.

and

BOUMAN,

M.A.

(1965)

The

effects of wave

length

and

luminance

on

the

viual modulation

transfer.

Proceedings

of

the

Colloquim

at

Delft

1965

pps.

183-192.

26.

WESTHEIMER,

G.,

(1960)

Modulation

thresholds

for

sinusoidal

light

distributions

on

the

retina.

27.

WESTHEIMER,

G.,

and

CAMPBELL,

F.W.

(1962)

Light

distribution

in

the

image

formed

by

the

living

human

eye.

J.

Opt.

Soc.

Amer.

(52)

pps.

1040-1045.

ADDITIONAL

RELATED

LITERATURE

A.

DOUGHTY,

G.F.,

PORTEOUS

,

R.L.,

WALL,

F.J.B.,

Visual

factors

in

projected picture

quality,

J.

Phot

Sci.

Vol

12

1964

pps.

211-221.

B.

HUBBARD,

R;

KROPF

_,

A?

Molecular

isomers

in

vision

Sci.

Amer.

Vol

216,

No

6

June

1967

pps.

64-76.

C.

HUBEL,

D.H;

Visual

cortex of

the

brain

Sci.

Amer.

1963,

pps.

2-10.

D.

ITTELSON,

W.H;

KILPATRICK,

F.P.;

Experiments

in

perception.

Sci.

Amer.

Aug

1951,

pps.

2-7.

E.

LAND,

E.H.;

Experiments

in

color vision.

Sci.

Amer.

May

1959,

pps.

2-4.

F.

MacNICHOL,

E.F.;

Three

pigment color vision.

Sci.

Amer.

Dec

1964,

pps.

2-19.

G.

PRITCNARD,

R.M.;

Stabilized

images

on

the

retina.

SCI.

Amer.

June

1961.

H.

WALL,

F.J.B.;

STEEL,

B.G.;

Implication

of

the

methods chosen

for

the

measurement of

the

statistical properties of

photographic

images.