I am interested in the topic of spatial frequency and its impact on face processing. I have been reading about previous studies that have used different spatial filtering thresholds for their stimuli. Some studies used frequencies below 8 cycles per image for Low Spatial Frequency (LSF) and above 32 cycles per image for High Spatial Frequency (HSF) stimuli.

However, I noticed that the size of the stimuli used in these studies varied, which could affect the degree of visual angle at which the stimuli were presented. For example, if one study used 250x250 pixel stimuli presented at 4 degrees of visual angle, and another study used 700x700 pixel stimuli presented at 12 degrees of visual angle (with the same 8 and 32 cycle per image values for LSF and HSF respectively), the cycle per degree values would be different in each study for both LSF and HSF conditions(8/4 = 2 vs 8/12 = 0,66 for LSF; 32/4 = 8 vs 32/12 = 2.66 for HSF). Does this mean that the participants in these studies perceived different spatial frequencies than those in other studies that used different stimuli sizes?

Should we scale the cycle per degree values to the total degrees of visual angle when we consider the perceived frequencies of participants and compare it with different studies? Or is there another approach that should be taken to address this issue?


1 Answer 1


It's common practice to express size and related matters, such as visual acuity measures, in terms of angles (minutes of arc) (Stronks & Dagnelie, 2014). E.g., grating visual acuity tests don't make sense when disregarding distance, because any size measure inherently changes when the distance changes to the eye, as you explain in your question. Using angular subtense instead of distance measures is therefore crucial (Fig. 1). Angular subtense can be expressed in minutes of arc.

The only way of directly comparing HSF and LSF stimuli with different contrasts, different sizes, and different distances (i.e., visual angles) is, hence, to look at their visual subtense, including images with spatial filters such as gratings.

Alternatively, you can project them onto a common reference, such as retinotopic coordinates and express the stimulus characteristics as distance measures occupied on the retinal surface (Stronks & Dagnelie, 2014).

- Paudel et al. IET Optoelectron (2013); 7(1): 1–8
- Stronks & Dagnelie, Expert Rev Med Devices (2014); 11(1): 23–30

angular subtense
source: Paudel et al. (2012)


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.