I am not asking the following question: Can humans ever see a photon in the same way we see a chair?

My question is: Can a human retina respond to a single photon? If so, how does this happen and why is the retina able to "sense" (detect) a single photon?


3 Answers 3


A single molecule of rhodopsin (actually the cis-retinal bound to it) can and actually does react to one photon (Purves et al. Chapter: Phototransduction in Neuroscience).

It has been estimated that a single light-activated rhodopsin molecule can activate 800 transducin molecules, roughly eight percent of the molecules on the disk surface. Although each transducin molecule activates only one phosphodiesterase molecule, each of these is in turn capable of catalyzing the breakdown of as many as six cGMP molecules. As a result, the absorption of a single photon by a rhodopsin molecule results in the closure of approximately 200 ion channels, or about 2% of the number of channels in each rod that are open in the dark. This number of channel closures causes a net change in the membrane potential of about 1 mV.

Photoreceptor cells continuously secrete the neurotransmitter glutamate in dark; with light exposure glutamate release would reduce. This leads to activation of the next line of cells (bipolar, horizontal cells and retinal ganglion cells).

Modelled relationship between glutamate release and rod membrane potential.
From Witkovsky et al. (1997).

From this model (Witkovsky et al. 1997) it appears that for a change in glutamate release rate, a larger change in membrane potential is needed. Moreover, the depolarization of the photoreceptors has to be sustained for a while so that downstream cells can react (light absorption is much much faster than biochemical processes). The lifetime of an activated rhodopsin molecule is ~40ms in mice (Gross and Burns, 2010) and ~400ms in salamanders (Lyubarsky et al., 1996). We can assume that human eye is more similar to that of mice. There are approximately 40,000,000 molecules of rhodopsin in a rod cell (BioNumbers). The most sensitive of retinal ganglion cells need ~0.04% of rhodopsin molecules per rod cell to be activated (Takeshita et al., 2017).

(b) Absolute threshold of On and Off retinal ganglion cells

Both On and Off RGCs carry rich information about the weakest light signals. Off cells are somewhat more sensitive than On cells but have a higher error rate in their gap-based coding. The absolute threshold for primate On and Off parasol cells in a two-alternative forced-choice task is extremely close to the limits posed by the quantal nature of light: On parasols reach 75% correct choices at a light level corresponding to approximately 0.0008 R* per rod per flash (mean, n = 6) and Off parasols at a light level corresponding to approximately 0.0004 R* per rod per flash (mean, n = 5).

A single activated rhodopsin molecule would be 0.0000025% of the total. This would be insufficient to activate an RGC.

So, I can make a reliable guess that you cannot actually "see" a photon although one of the retinal photoreceptor cells can sense it. However, I'm still looking for a solid reference that says how many photons are needed (and for how long) to produce a visual response.

  • 6
    $\begingroup$ A single photon is sufficient to elicit a somewhat reliable response (whether this counts as "seeing" is open to debate, but seems to fit the OP's criterion): nature.com/articles/ncomms12172#discussion $\endgroup$
    – llama
    Commented Jun 18, 2019 at 21:22
  • 5
    $\begingroup$ Another thing to add would be that if our eyes were good enough to reliably grab single photons, we'd have great night vision. Your cats might stand a better chance, though it's much harder to determine if they have actually "seen" anything or not. Moreover, if our eyes were good enough in that regard, we might even have discovered photons and quantum mechanics earlier than we did in our actual history, by trying to figure out why that very dim light gets "staticky". $\endgroup$ Commented Jun 19, 2019 at 19:47
  • 1
    $\begingroup$ I've been looking for a reference on how many photons are need for threshold perception, for a while. For some reason the number seven to ten photons knocks around physics teaching circles, but I have no idea where it comes from. $\endgroup$ Commented Jun 20, 2019 at 22:53
  • $\begingroup$ The numbers I remember having heard are 10 photons per second sustained exposure or 100 photons in a single burst $\endgroup$ Commented Jun 21, 2019 at 5:57

A recent study published in Nature by Tinsley et al. Direct detection of a single photon by humans found that it is possible for dark-adapted humans to respond to a single-photon stimulus, but only rarely. They used a source which created pairs of photons, and used one of the pair to determine whether the subject may have been exposed to a single photon. The subjects were asked to respond whether they had seen a single photon after an event, and to rate their confidence in their response on a scale of 1-3. Of the highest confidence trials, $60\% \pm 0.3 \% \space (P=0.001)$ of responses were correct. Not exactly reliable, but strong evidence that it is possible.

They also interestingly found that the detection efficiency was highly increased if another single photon was emitted in the previous ~3.5 seconds.

As far as mechanisms, I'm no biologist or biochemist so I'll suggest reading section III from Rieke & Baylor (1998) Single-photon detection by rod cells of the retina, which covers it in detail.

  • 1
    $\begingroup$ @NicHartley they state somewhere in there or the supplementals that they didn't have good enough individual-subject statistics to do that convincingly, but that people took a few sessions to get good at it $\endgroup$
    – llama
    Commented Jun 18, 2019 at 21:52
  • 2
    $\begingroup$ @DmitryGrigoryev a lot of biology uses p<0.05. At 0.001, its fairly strong evidence that people are better than a coin flip (but still pretty bad!) at detecting single photons. $\endgroup$
    – mbrig
    Commented Jun 18, 2019 at 22:10
  • 1
    $\begingroup$ @mbrig Or perhaps people are good at detecting something correlated with the photon emission (like noise), but not everyone figured that out. $\endgroup$ Commented Jun 18, 2019 at 22:15
  • 3
    $\begingroup$ @DmitryGrigoryev they both minimised (sound damping foam, fibre to couple light into room a distance from source) and tested for that. The efficiency of the single photon source is low, because if you crank it up you get significant numbers of multiphoton events. Most trials (92%) actually had no photons, although they would have had every other possible signal that would bias a subject. "in these trials the probability of correct response was not different from the baseline for both the combined and the high-confidence responses (0.505±0.003, P=0.08 and 0.507±0.01, P=0.3, respectively)" $\endgroup$
    – llama
    Commented Jun 19, 2019 at 0:07
  • 2
    $\begingroup$ I haven't read the paper but I wouldn't be surprised if the better than chance performance is an example of stochastic resonance. $\endgroup$
    – Bryan Krause
    Commented Jun 20, 2019 at 16:03

Technically, we can sense the individual photons. Here is an quote from a cellular biology textbook:

"Absorption of a single photon of light induces a conformational change in the rhodopsin molecule, which transmits a signal to a heterotrimeric G protein (called transducin ), which activates a coupled effector." (Karp's Cell and Molecular Biology 8e, 603).

However, we have neural filters (part of evolution/adaptation) that only "see" light if somewhere around 5-9 photons are detected within a 100 ms period. Here is a source for this.

According to that source, this neural filter is actually an evolutionary advantage, as it will prevent unnecessary visual data from being processed (which would add a lot of noise to the "images" we see).

However, something interesting is that frogs can actually detect and see single photons of light. I remember watching a video some time back where it was explained as follows: If a frog is launched to the outer edges of our solar system (where photon density from the sun is very little), time-to-time a single photon will enter the frog's eye, and the frog will see flashes of this light, while seeing darkness when photons aren't entering. You can Google more about these frogs, or read a summary of it on The seeing power of frogs: Frogs can detect single photons of light.

Edit: As WYSIWYG pointed out, frogs aren't so special after all! There are more than one organisms that can "see" (not just sense) these individual photons.

Hope this helps!

  • 2
    $\begingroup$ @WYSIWYG Hi, I read it from this source for my class last year: math.ucr.edu/home/baez/physics/Quantum/see_a_photon.html $\endgroup$
    – F16Falcon
    Commented Jun 18, 2019 at 22:53
  • 1
    $\begingroup$ @WYSIWYG I guess it's in one (or all?) of those: Julie Schnapf, "How Photoreceptors Respond to Light", Scientific American, April 1987 S. Hecht, S. Schlaer and M.H. Pirenne, "Energy, Quanta and vision." Journal of the Optical Society of America, 38, 196-208 (1942) D.A. Baylor, T.D. Lamb, K.W. Yau, "Response of retinal rods to single photons." Journal of Physiology, Lond. 288, 613-634 (1979) $\endgroup$
    – user38407
    Commented Jun 19, 2019 at 11:05
  • $\begingroup$ @F16Falcon I just found that the lifetime of activated rhodopsin is longer in salamander than in mouse. Not sure about frogs but photoreceptors of different organisms can have different sensitivities. $\endgroup$
    Commented Jun 19, 2019 at 17:10
  • $\begingroup$ @WYSIWYG Yes they can, and the frog was just an example that I wanted to give because it was fascinating to me. I didn't know about other organisms that did this, so thanks for sharing! Also, that is the original document I read (I didn't read the full research papers); Vaxquis is correct. $\endgroup$
    – F16Falcon
    Commented Jun 19, 2019 at 22:09

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .