Psychophysical Stimuli Flashcards
What are sinewave gratings?
Basic stimulus
* Are one of most used stimulus
* Sinewave gratings are building blocks to construct more complex stimuli
* Lego blocks of making visual stimuli
Grating: repeating sequence of light & dark bars
Describe Square Wave Grating?
- These gratings have sharp edges
- Their luminance profile looks like a square
- When cut through square wave along x-axis and plot its luminance profile – will see abrupt change in luminance alternating between maximum and minimum – have same width
- Even though it looks simple – it is a special kind of grating made by combining infinite number of different sine wave gratings mathematically
- As luminance profile only changes in x-axis – it is a 1-dimensional stimulus
Describe Sine Wave Grating?
- Sine wave grating – black and white bars are smoothly undulating along the x-axis without sharp edges
- Can see undulation when cut through sine wave grating
- If plot luminance profile along x-axis, then will see sine wave characteristic
Sinewave Grating is mathematically v well-defined so v easy to manipulate & change the aspect of the grating that corresponds to a specific parameter of vision.
Describe Gabor Grating?
- Response profile of simple cell in V1 can be nicely approximated by this function
- Dark and light bar in centre, luminance intensity tapers off from centre to sides in every direction smoothly
- Gabor is a sine wave function modulated by 2D Gaussian (Bell-Curve) Function
- Well known application of Gabor function: model response profile of V1 neuron
o RF tuning characteristics of V1 neurons have been modelled using even- and odd- symmetric Gabor functions
What is the sinewave grating equation and describe the parameters? What is special about this equation?
L = Asin (fθ ± φ) + L (lower m)
L = luminance profile of sinewave grating
A = amplitude
Can replace ‘sin’ in equation for ‘cos’
f = spatial frequency
θ = angle (or orientation)
φ = phase
L (lower m) = average luminance
this is a 1 directional equation –> changing orientation here does not change actual orientation of sine wave grating
Describe the sinewave grating equation in more detail?
- L is a function of A
- Amplitude of sine wave function defines how high and low peaks and valleys are in undulating function
- Difference between peaks and valleys defines contrast of grating
- Sin function can be replaced by Cos function
- Appearance of sine wave function is a function of the spatial frequency (f)
o Spatial frequency defines how many cycles are contained within a defined space – e.g. a degree of visual angle
o High spatial frequency – width of each light and dark bar is v narrow
o Low spatial frequency – width of each light and dark bar is wide - Θ = angle or orientation of grating
o In practice need to have a 2D version of sine wave equation to change the orientation of the grating - ϕ (Phi) = phase of grating
o Phase determines relative location of peaks and valleys - Lm = average luminance
o Middle brightness of sine wave grating
o From average luminance, luminance profile of sine wave grating will fluctuate - If change any of the parameters, can change the appearance of the sine wave grating
What happens to the gratings when contrast increases? What happens when spatial frequency increases?
- Gratings become more visible as contrast increases
- Increasing spatial frequency can place more cycles in a unit visual angle
Describe the basic stimulus: Optotypes?
- Black letters on a white background -> basic letter stimuli for eye charts
- Snellen letters:
o Have serif – extra stroke - Sloan:
o Perfectly square
o 5x5 size – each side is 5 units with stroke width equal to 1/5 of a side
o Snellen E (tumbling E) or landolt C -> essentially sloan letters - British Standard:
o 5x4 size – 5 vertically and 4 across - Fonts w/o serif are called sans serif fonts
- Not all letters are used on vision charts due to legibility difference between letters
Describe stimulus parameters: contrast?
- Contrast parameter defined differently depending on type of stimulus used in experiment
- Relative difference in luminance between objects within a scene
- Important for psychophysics: human visual system is known to be more sensitive to contrast rather than absolute luminance of objects
Describe Weber Contrast (C lower W)?
- Used to define isolated features of image against large uniform background e.g. black optotype on white background
- Ratio between luminance difference between target and background divided by background luminance
- -ve Weber contrast occurs when black optotype on white background
- Can go to +ve infinity when letter is white on black background
What is the Weber Contrast equation and what is its units?
C lower W = ΔL / L lower b = (L lower t - L lower b) / (L lower b)
L lower t = luminance of target image
L lower b = background luminance
Ratio between luminance & difference between target & background over background luminance
cd/m2 = typical unit of measurement for luminance intensity using photometer
cd = candela = unit for luminance intensity
What is Michelson Contrast (C lower M) & what is the equation?
- Used for images/patterns where luminance profile is periodically fluctuating & repeating
- Can range from 0% to 100%
- Difference between peak & valley divided by average luminance of grating
What is Michelson Contrast (C subscript M) & what is the equation?
- Used for images/patterns where luminance profile is periodically fluctuating & repeating
- Can range from 0% to 100%
- Difference between peak & valley divided by average luminance of grating
C subscript M = (L subscript max - L subscript min) / (L subscript max + (L subscript min)
Difference between peak & valley divided by average luminance of grating
Where L subscript max = L subscript avg + ΔL
and where L subscript min = L subscript avg - ΔL
When plug these in, end up with Weber contrast equation
What is Root Mean Square Contrast (C lower RMS) & what is the equation?
- Used for complex images and patterns such as natural images or randomised stereograms
- Numerator is the standard deviation of pixel intensities or luminance in 2D images scales to mean luminance (which is the denominator)
- This contrast ranges from 0 to 1 assuming there is no outlier in the image
C lower RMS = square root of x / L bar
x = (1/M*N) * Sum upper N lower i=1 * Sum upper M lower j=1 * (L lower ij - L bar) ^2
M,N: rows * columns of pixels in a 2D image
L bar: average luminance of image
