Speed of eye response to red vs green LEDs

[Circlotron]

Just now I put together a little circuit that flashes a red and a green LED together at a 5Hz rate via two separate micro ports. The red LED leads the green LED by 2uS. The thing is, the red LED looks very much like it powers on maybe 20-30mS before the green. Is there some kind of difference in the response speed of the eye to these two different colours? Is persistence of vision related to colour?

[tooki]

Try swapping the LEDs and seeing if the delay moves with the LED or stays with the port. (LEDs have essentially instant response, easily capable of turning on and off at megahertz speeds. So it's not that.) I'd be willing to bet the delay moves with the port…

[hamster_nz]

Yes. It is one of the little projects I suggest for FPGA newbies.

Also sensitivity to flicker changes with colour too.

[Circlotron]

And further to this, if I look at both LEDs side by side at the same time using my peripheral vision, the red one flashes steadily but the green one appears slightly erratic.

[BrianHG]

Scope both LEDs.  1 led on channel 1 and the other on channel 2 and send us a scope shot.

[hamster_nz]

Also a dark room, if you move a red and green LED around, the red one tends to lag, more so.outside of center of vision.

[Circlotron]

Swapped LEDs to opposite ports and the red still appears to lead.

If they are separated by 50mm and viewed from 400mm with peripheral vision, if I concentrate on one it is difficult to perceive the other one flashing. Green is more difficult to see flashing under these conditions.

[tooki]

I wonder if brightness also matters. I expect it would, and that the two LEDs aren't perfectly matched. (Or indeed, they may not be even distantly matched, depending on what types they are!)

[ANTALIFE]

Use a scope to check voltage on each LED.
Maybe the voltage slowly ramps up and the red LED is on before the green one as it has a lower forward voltage; as in Vf-red = 1.7V & Vf-grn = 2.2V, meaning at 1.7V red led will definitely be on but green one will "not"

[james_s]

I had that thought for a moment but then realized if they're being driven by a square wave the rise time should be insignificant. I don't know for sure what's driving them but it says micro ports which implies a square wave from a microcontroller.

[viperidae]

You've discovered how amazing your brain is and how crap your eyes are.

Eyes are terrible. Gaping blind spots, slow to react, poor peripheral resolution, nonlinear, poor colour reproduction. Your brain fills in the blind spots by moving your eyes around automatically and using previously seem images and other replicating other shapes you can see.

[GigaJoe]

oh well,
relative amounts of red/green-sensitive pigments on the fovea produce perception .... it individual
some anomaly eye condition to reduce sensitivity to red green blue,  as an extreme condition ...
 

[Nusa]

oh well,
relative amounts of red/green-sensitive pigments on the fovea produce perception .... it individual
some anomaly eye condition to reduce sensitivity to red green blue,  as an extreme condition ...

It's not even that extreme a condition. About 8% of males and 0.5% of females (that's for European ancestry - stats may be higher or lower if you're something else) have some degree of red-green color deficiency. It's quite possible to have it and not even know it in the milder cases. If any maternal ancestors were color blind, your odds are much higher.

[David Hess]

And further to this, if I look at both LEDs side by side at the same time using my peripheral vision, the red one flashes steadily but the green one appears slightly erratic.

The rods which detect black and white instead of color are both more sensitive and have a faster response than the cones which detect color and are concentrated in the fovea (center of the field of vision).

So it is much easier to detect flickering using peripheral vision than by watching it directly.  I used to notice it all the time with computer CRTs operating at low refresh rates and sometimes with fluorescent lamps.  These days it shows up with LED brake lights that are deliberately strobed to attract attention.

[alexanderbrevig]

 Sounds like persistence of vision to me? Ever tried takin a few quick glances at a clock? Look away and randomly look at it for a few seconds. Half the time it will seem as if a second is longer than a second.   That explanation sucked!  Try it:)

[RoGeorge]

Nice finding about the red/green delays, will give it a try, thanks for sharing the results! 

I would have bet the green would be perceived first, because the eye sensitivity peak is at green AFAIK. Well, I would have lost. 

For whatever reason, my peripheral vision is very sensitive at blinking blue light. I noticed because they are some car alarms that have constantly blink a small blue LED. At night, I can spot from tens, maybe hundreds of meters the blinking blue light with the peripheral vision. The reflex is to instantly look at the blinking light, yet when I do that I see nothing, unless I came much closer to the blinking blue light (meters away instead of tens of meters).

There are lots and lots of oddities and tricks played by the brain in order to make the best future prediction. If it were to wait until all information is collected and processed, it would be too late. By then, we would have already been eaten by the tiger. 



 
Apart from being an attention test, that is also a good proof that we don't perceive what the reality is, but what the mind expect it to be. We are not aware about this, but what we perceive is a continuous fabricated daydream guided by reality, instead of perceiving the reality.

I wonder what will be the reason to perceive the red faster (or why the green can be left to lag behind a little), or is it just because some unwanted side effect of how the retina works.

[StillTrying]

It would be interesting to make a 'reaction tester' using 3 or 4 different coloured LEDs, and averaging the results sorted by colour.

Everybody is more sensitive to blue light in their peripheral vision. You can test it by looking at a very dim blue LED, when you look 15 to 20 degrees off, it looks 2 or 3 times brighter, you can also do it to see dim stars.

[David Hess]

Everybody is more sensitive to blue light in their peripheral vision. You can test it by looking at a very dim blue LED, when you look 15 to 20 degrees off, it looks 2 or 3 times brighter, you can also do it to see dim stars.

Everybody is more sensitive to light in general in their peripheral vision because the more sensitive rods are concentrated there instead of the less sensitive but color discriminating cones.

[RoGeorge]

Everybody is more sensitive to blue light in their peripheral vision. You can test it by looking at a very dim blue LED, when you look 15 to 20 degrees off, it looks 2 or 3 times brighter, you can also do it to see dim stars.

Everybody is more sensitive to light in general in their peripheral vision because the more sensitive rods are concentrated there instead of the less sensitive but color discriminating cones.

Combining the observations above, could the extra sensitivity to blue in the peripheral vision came just from the fact that the blue photons carry more energy than other visible colors?

Extra sensitivity to blue doesn't seem to me like an advantage in the process of natural selection, unless the subject is a fish living in the deep blue waters.


The reflex time measured when an RGB LED lights up at different colors and intensities will be a very interesting experiment.

Would be very interesting to see the results in a 3D chart, with the reaction time on the Z axis, and wavelength and intensity on XY.

[David Hess]

Combining the observations above, could the extra sensitivity to blue in the peripheral vision came just from the fact that the blue photons carry more energy than other visible colors?

Extra sensitivity to blue doesn't seem to me like an advantage in the process of natural selection, unless the subject is a fish living in the deep blue waters.

Rod cells have a peak sensitivity between blue and green but are 100 times more sensitive than cone cells.  The Wikipedia article says they are also slower but that sure is not my experience.

[bw2341]

I think we might be jumping to conclusions that visual perception is responsible.

As far as I know, there are three kinds of green LEDs.

The old-fashioned yellow-green
The modern pure green
The recently introduced phosphor green

The phosphor green LED will definitely lag a non-phosphor red LED. I don’t know if this would account for the perceived lag, but it should be ruled out. Maybe you could use a photodiode to measure the luminous response time.

With the blue LED chemistry so dominant in terms of market share, it’s probably cheaper to make all other colours with blue plus phosphor. For applications that don’t need pure colour, you probably have higher luminous efficacy as well.

[james_s]

I've never seen a phosphor green LED, what's the advantage of them? Wavelength? I'll have to look around and see if I can find a place to get a couple to play with.

The thing I don't like about phoshor LEDs is at least the small ones seem to wear out a lot faster than direct emission. The power emitters used in LED lighting hold up pretty well but I've had a lot of those 5mm white LEDs turn dim and gray after a year or less. A company I worked for made a product with a white 7-segment display and the segments most used by the clock would become noticeably dim and icky looking within less than a year.

[bw2341]

https://youtu.be/7povdS90Hyg?t=702

I heard about them in a bigclivedotcom video about Poundland Christmas lights. It looks like LED manufacturers are mixing and matching phosphors to expand the palette of colours for decorative applications. The wider spectrum of the phosphor-converted LEDs seems less harsh than the pure colours of direct emission LEDs.

https://www.ledsmagazine.com/articles/2014/11/osram-research-project-increases-efficiency-in-green-leds.html

I found this article on the topic. Basically, direct emission green is a laggard in luminous efficacy and phosphor-converted green is the solution.

[james_s]

Interesting, that's surprising. I would have thought direct emission would win hands down, at least for the high efficiency green color. They certainly appear very bright although that could be in part to the peak sensitivity of the human eye being in the green range.

[RoGeorge]

The phosphor green LED will definitely lag a non-phosphor red LED.

At turning on? Why? And by how much?