Goldfish vision: the science and the numbers
Carassius auratus · order Cypriniformes · Fish: all the numbers
Goldfish have four colour channels, including ultraviolet (360, 455, 532 and 614 nm).[1][2][3] Their sharpest vision resolves 3 cycles per degree, against 63.75 for people in this dataset.[8] Goldfish stop seeing flicker at 53.6 Hz, against 60 Hz for people.[29][30]
- 4colour receptor classesMeasured*
- 3cycles per degree (sharpness)Measured
- 53.6hertz flicker fusion (motion)Measured
The goldfish (Carassius auratus) is a fish in the order Cypriniformes. Its eyes belong to the vision type Shallow-water fish tetrachromat: four cone types including ultraviolet and far red. Measured in this species: colour, sharpness and motion (flicker fusion). Measured core: measured values on at least 3 of the 6 dials.
This is a simulation built from published eye measurements, not what the animal experiences.
What goldfish see: colour receptors
What stands out
- It has 4 colour receptor classes, including ultraviolet; people have 3.
- Its sharpest vision resolves 3 cycles per degree: the finest stripe pattern it can tell apart from grey.
- It stops seeing flicker at 53.6 Hz, against 60 Hz for people in this dataset, so fast motion looks choppier to it.[29][31]
The six dials
Evidence levels: how the tiers work. "Measured" means a value measured in this species; "Estimated" values come from a close relative or an eye-size formula. The last column gives the values for people from the same catalogue.
| Dial | Value | Evidence | Sources | People |
|---|---|---|---|---|
| Colour | Colour receptors 4 receptor classes: 360 nm (UVS), 455 nm (SWS (blue)), 532 nm (MWS (green)), 614 nm (LWS (long)) measured in this species | Measured (not re-verified) | [1][2][3] | Colour receptors: 3 receptor classes: 421.5 nm (VS/SWS (violet)), 532 nm (MWS (green)), 558.4 nm (LWS (long)) Measured (not re-verified)[4][5][6][7] |
| Ultraviolet yes: at least one receptor peaks in the ultraviolet | Measured (not re-verified) | |||
| Sharpness | Acuity 3 cycles per degree median of 7 behavioural rows (method priority rule) | Measured | [8] | Acuity: 63.75 cycles per degree Measured[9][10] |
| Field of view | Binocular overlap 32.85° median of 2 relatives in family Cyprinidae: Danio rerio, Notemigonus crysoleucas | Estimated | [11] | Binocular overlap: 122.5° Measured[12][13] Total field of view: 200° Measured (not re-verified)[14] Blind area behind the head: 160° Derived[14] Eye placement: frontal Derived[12][13] |
| Eye placement lateral | Estimated | [11] | ||
| Sharp zones (foveas) | Number of foveas 0 group default: mode of tier-A values in vision type the "Shallow-water fish tetrachromat" type within phylum Chordata (2 species: Toxotes jaculatrix, Anableps anableps) | Group default | [15][16] | Number of foveas: 1 Measured[17] Fovea type: fovea Measured[17] |
| Fovea type ventrotemporal area (high rgc density) | Group default | [15][16] | ||
| Night vision | Activity pattern diurnal group default: mode of tier-A values in vision type the "Shallow-water fish tetrachromat" type within phylum Chordata (1 species: Amphiprion ocellaris) | Group default | [18] | Activity pattern: diurnal Measured (not re-verified)[19][20][21][4][22][23][24][25] Pupil shape: vertical Group default[26][27] Reflective layer (tapetum): no Measured[28] Rods vs cones: cone-dominated Derived[19][20][21][4][22][23][24][25] |
| Rods vs cones cone-dominated | Group default | [18] | ||
| Motion (flicker fusion) | Flicker fusion frequency 53.6 Hz median of 2 bright-light rows (behavioural/whole-eye ERG rows; all rows: [40.0, 67.2]) | Measured | [29][30] | Flicker fusion frequency: 60 Hz Measured[29][31] |
All fish side by side: Fish: every measurement. Method: how we know what animals see.
Sources
- Schweikert et al. 2018
- VPOD in-vivo (MSP / single-cell) lambda… 2025
- Hárosi et al. 1974
- Longcore 2023
- Kirwan
- Müller et al. 2009
- Thermal Activation and Photoactivation of Visual… 2004
- Caves et al. 2017
- Kirk et al. 2004
- Veilleux et al. 2014
- Pita et al. 2015
- Heesy 2004
- Heffner et al. 1992
- Campbell & Green 1965
- Schwab et al. 2001
- Temple et al. 2010
- Kopania et al. 2025
- Froese et al.
- Anderson et al. 2017
- Borges et al. 2018
- Wilman et al. 2014
- Maor et al. 2017
- Jones et al. 2009
- Schmitz et al. 2011
- Moura et al. 2024
- Banks et al. 2015
- Cervino et al. 2021
- Guareschi et al. 2025
- Healy et al. 2013
- Lafitte et al. 2022
- Inger et al. 2014
Every value cites its sources (all sources). Values were extracted from these works and converted (units, medians, derived values); changes are ours, and the listed sources do not endorse this site. Data: catalogue-v1, built 2026-09-29. Accuracy notes: how accurate is this? Method: how we know.