Testing IR Sensitivity: My Experiment

[Corkie2003]

Count me in.

I sell I.R. kit for outdoor sports, and the only way I know the illuminators are turned off, is by pushing the offer/onner so the little LED goes out and THEN checking at the business end to make sure the lights are indeed turned off.

This applies to I.R. lasers (which I don't point directly at my eyes), I.R. LEDs and lights with I.R. filters on.

. . . . and this whole thread has just started a big fight between me and my dad, who says that you just CAN'T see I.R.

 

[mgb_phys]

You can't see IR (pretty much by definition), what you can see is the 1 in a billion red photons that are 6 sigma away from the peak wavelength.

 

[Corkie2003]

Yeah, thanks, thanks a lot - sometimes it's more important to win an argument than tell the truth you know!

 

[unitone101]

Well, I tried this for a test. If I aim my remote towards my eye, and click the remote I can't see anything, and my pupils don't react. I tried it on a few of my family members and nothing happens. But if I aim it at my girlfriend, her pupils contract! So it's hard for her to lie or try watching my fingers or wathever. As far as I'm concerned, my girlfriend CAN see the infrared light on my remotes. Her pupils respond to them and I can't believe she can contract or dilate them at will. And anyways, she doesn't lie and doesn't care for seeing IR or not. She's actually just fed up of me pointing the remote at her eyes but I'm still fascinated by this.

God, I still remember seeing her using the remote as a flashlight.

 

[mgb_phys]

As far as I'm concerned, my girlfriend CAN see the infrared light on my remotes. Her pupils respond to them

If her pupils are V shaped and you seem to go through a lot of pet hamsters - you should be concerned!

 

[Redbelly98]

I used to work with lasers, and could see the IR from a 794 nm laser diode. 1/2 W, focused down to a 100 micron spot.

 

[DaveC426913]

You can't see IR (pretty much by definition)...

Why do you say this? The definition of IR isn't "below the visible light threshold of humans", it is simply "below red". And "red", as with "IR", is not well-defined in terms of frequencies.

 

[mgb_phys]

Why do you say this? The definition of IR isn't "below the visible light threshold of humans",

It's a pretty good definition!
CIE defines it as 700nm
Photographic astronomers define it as 730nm ( I band =806nm +- 75nm)
CCD astronomers define is as 1050nm cut-off of silicon
Modern astronomers define it as J band 1.2um same as comms fibre people - it's the first water band gap.

 

[DaveC426913]

It's a pretty good definition!
CIE defines it as 700nm
Photographic astronomers define it as 730nm ( I band =806nm +- 75nm)
CCD astronomers define is as 1050nm cut-off of silicon
Modern astronomers define it as J band 1.2um same as comms fibre people - it's the first water band gap.

That's fine. So then how can you state categorically that humans cannot see those wavelengths?

Or, at least, how can you correlate the threshold of IR with the limit of human colour perception?

 

[mgb_phys]

There is obviously some variation in the limit of human vision from person to person.
The limit is somewhere around 750nm it could easily vary by 10nm - but I don't think it extends to 880nm!
However cheap IR leds with a nominal peak at 880nm and a 30nm bandwidth can emit a small amount of light at several times this bandwidth. And remember that a dark adapted eye, especially peripheral vision, is amazingly sensitive - it can easily detect single photons.

 

[DaveC426913]

There is obviously some variation in the limit of human vision from person to person.
The limit is somewhere around 750nm it could easily vary by 10nm - but I don't think it extends to 880nm!
However cheap IR leds with a nominal peak at 880nm and a 30nm bandwidth can emit a small amount of light at several times this bandwidth. And remember that a dark adapted eye, especially peripheral vision, is amazingly sensitive - it can easily detect single photons.

But you're missing a critical factor: one person can see them, the rest cannot.

Assuming normal people can see "plain old red" it follows that the frequencies normal people cannot see must be infra-red. i.e. Whether or not the light from the LED is wide-spectrum, 99% of normal people cannot see it. That would make it IR.

Now we introduce a few rare people that still can see the light.


So, I guess I'm turning your definition on its head:

You said, "if people can see it, then, by definition it's not IR".
I'm saying, if (normal) people can't see it, then, by definition it is IR. This allows for the possibility that some can.

 

[mgb_phys]

But you're missing a critical factor: one person can see them, the rest cannot.

We don't exactly have a controlled experiment with the same LED being used by everyone. People's night vision definitely varies and it takes upto 20mins to become completely dark adapted.

I don't know what the ultimate cut off wavelength of rhodopsin is but it can't really vary from person to person - the molecule is the same. What varies is the number/density of rods, the ability of the brain to process low level signals, the amount of time they are prepared to stand around becoming dark adapted - and the stability of the IR led.

 

[DaveC426913]

We don't exactly have a controlled experiment with the same LED being used by everyone.

I did the experiment with a half dozen people and one remote. Only one person could see it. Not a big sample I'll grant but enough to demonstrate that it is not common and it definitely varies by the person, not by the electronics.

 

[mgb_phys]

The night vision acuity - their ability to see very faint red light - varies from person to person. None of them are able to see light much beyond 750nm.

 

[DaveC426913]

The night vision acuity - their ability to see very faint red light - varies from person to person. None of them are able to see light much beyond 750nm.

My experiment was in broad daylight with a thick blanket over their head. Zero time for acclimatization.

And as for "very faint", he was able to call it as fast as I could press the button. "OnOffOnOffOffOffOffOnOffOn..." and I couldn't fool him even by faking button-presses. I could press or unpress a zillion times and he'd get it right 100%.