Internal noise and efficiency underlie performance asymmetries throughout the visual field

[Background] Visual performance peaks at the fovea and decreases with eccentricity and differs around polar angle: it is better along the horizontal meridian (HM) than the vertical meridian (VM) and at the lower than upper VM. Whether performance differences throughout the visual field are due to differential ability to extract task-relevant information in a noisy environment remains unknown. Here, we investigate how two factors limiting performance –internal noise (amount of internal variability in the system) and efficiency (ability to extract information from the target)– underlie performance differences throughout the visual field. [Method] We used an equivalent noise method, which maps contrast threshold as a function of noise contrast, to estimate and disentangle internal noise and efficiency at the fovea and around polar angle (left & right HM and upper & lower VM) in parafovea (4°) and perifovea (8°) . At each location, observers discriminated the orientation of a 5 cpd Gabor (±45º off the vertical axis) embedded in dynamic white noise. We obtained contrast thresholds at 7 noise contrast levels at each location. [Results] We found that (1) Internal noise increased with eccentricity and was lower at HM than VM in the parafovea, but did not differ around polar angles in perifovea; (2) Efficiency (a) was higher at the fovea and parafovea than perifovea; (b) did not differ around polar angle in the parafovea, and (c) was higher on the HM than VM, and at the lower- than the upper-VM in the perifovea. [Conclusion] Distinct computations limit performance throughout the visual field: Internal noise primarily underlies eccentricity differences, consistent with variation in the distribution of the cortical tissue. Efficiency primarily underlies polar angle differences and more so at perifovea than parafovea, consistent with variation in tuning properties.