Infra-red vision vs. dark vision...

[Johnny1A.2]

Assuming a fairly realistic human-like hominid eye, for a hominid adapted to spending large amounts of time (but not all the time) underground, which adaptation likely would be more viable, and more sense biologically/physically?

(I'm visualizing something closely evolutionarily akin to Homo sapiens, analogous to a Dwarf from fantasy fiction, but not magical, rather evolved by natural selection for semi-underground life.)

Dark vision works by hypersensitivity to what light there is. Now in a closed underground environment, dark vision still isn't very useful, there's darkness and then there's the darkness in an unlit cave, which is a whole new level of dark. But dark vision could make small, weak light sources viable in a way they wouldn't be to humans. A 'dwarf' with dark vision might see as well by the light of a single candle as we would from a bright electric light.

But the downside might well include tremendous sensitivity to bright lights and daylight, unless there are additional adaptations to cope. I'm also not sure just how realistically 'dark-sensitive' a near-human eye could plausibly be.

Infra-red vision might work in an apparently totally unlit cave or mine, from the natural heat radiation, at least to a degree. But again, there is the issue of realistic infra-red sensitivity, as well as the size of the eyeball to cope with the longer wavelengths.

Infra-red vision might be easier to pair with normal optical vision in daylight, though.

(For context, these creatures are near-humans average 4 feet tall or so, heavily built like the traditional fantasy dwarf, but while they might be happier in the night they aren't supposed to be nocturnal as such.)

[Fred Brackin]

Originally Posted by Johnny1A.2

Dark vision works by hypersensitivity to what light there is. .)

I think you mean _Night vision_ which works as you describe. Dark Vision works without any light at all..

[Flyndaran]

Infravision, as in seeing long wavelength infrared, requires a sensor colder than what you're detecting. So not plausible for a endotherm at all, unless you need to see really hot objects for some reason.
That's also assuming you conceive of a biological infrared focusing lens which hasn't evolved on earth. Pit vipers' heat vision only barely works to within 6 feet.
As for extreme night vision, lack of color perception is not a given. Some lizards lack rods all together, but are still adapted to the dark using sensitized cones.

[Johnny1A.2]

Originally Posted by Fred Brackin

I think you mean _Night vision_ which works as you describe. Dark Vision works without any light at all..

True enough, I should have said night vision.

[Anthony]

Originally Posted by Johnny1A.2

Assuming a fairly realistic human-like hominid eye, for a hominid adapted to spending large amounts of time (but not all the time) underground, which adaptation likely would be more viable, and more sense biologically/physically?

Infravision sufficient to detect a person is probably not possible for an exotherm, though there are endotherms (pit vipers) with a primitive version, and there's certainly no easy evolutionary path that leads to it. Night vision adaptations are more possible but won't do a lot of good underground, as they only work in conditions of low light, not no light. Echolocation (in GURPS, probably a low res air sonar) is an option but again there's no easy evolutionary path.

[Dustin]

I've toyed with the idea that dwarven beards are highly sensitive to air currents, and that they can use them to get a sense of underground areas, especially where the air is entering and exiting as the cave "breathes".

As for echolocation, it can apparently be done with human senses well enough for a blind man to ride a bicycle and describe trees. So, I suspect a hominid trained from youth, and perhaps with additional hearing advantages, could do pretty well in that regard.

[whswhs]

Originally Posted by Anthony

Infravision sufficient to detect a person is probably not possible for an exotherm, though there are endotherms (pit vipers) with a primitive version, and there's certainly no easy evolutionary path that leads to it. Night vision adaptations are more possible but won't do a lot of good underground, as they only work in conditions of low light, not no light. Echolocation (in GURPS, probably a low res air sonar) is an option but again there's no easy evolutionary path.

You've got those terms reversed. An endotherm is an organism that generates heat internally ("endo"); an ectotherm is one that takes in heat from its external environment ("ecto"). Pit vipers are ectotherms.

Realistically, infrared vision that responds to human body temperature has poor resolution; the wavelength of peak emission at human body temperature is about 16x that of monochromatic green light at peak human visual sensitivity, and for an eye of a given size, resolution is limited by wavelength. So you'd get around 20:320 visual acuity at best. You can do a lot better with near infrared, which could get you to around 20:40, but NIR emission from a human body is nearly zero; you'd need a temperature of around 4000°F to get peak emission in the NIR, if I'm figuring Wien's law right.

[Anthony]

Originally Posted by whswhs

You've got those terms reversed.

Whoops, so I do.

Originally Posted by whswhs

Realistically, infrared vision that responds to human body temperature has poor resolution; the wavelength of peak emission at human body temperature is about 16x that of monochromatic green light at peak human visual sensitivity, and for an eye of a given size, resolution is limited by wavelength. So you'd get around 20:320 visual acuity at best.

Well, a bit better -- 20:20 vision requires a pupil size of around 4,000 wavelengths, which is 1.6-2.8mm for the visual spectrum, and probably around 35mm for a reasonable thermograph, so a pupil dilated to 7mm might approach 20:100.

Originally Posted by whswhs

You can do a lot better with near infrared, which could get you to around 20:40, but NIR emission from a human body is nearly zero; you'd need a temperature of around 4000°F to get peak emission in the NIR, if I'm figuring Wien's law right.

True, though you don't need peak emissions to detect something (last I looked at this, I figured NIR might reduce darkness penalties for total from -10 to -9). Mostly what NIR will give you is a modest bonus to night vision when the primary light source is incandescent, and a larger bonus when it's something like a candle or torch.

[whswhs]

Originally Posted by Anthony

Well, a bit better -- 20:20 vision requires a pupil size of around 4,000 wavelengths, which is 1.6-2.8mm for the visual spectrum, and probably around 35mm for a reasonable thermograph, so a pupil dilated to 7mm might approach 20:100.

But it's not only pupil size. It's also how closely spaced the receptors can be. Human retinal neurons are as close together as is permitted by the wavelength of visible light. If you go to thermal infrared wavelengths you need the neurons to be more widely spaced.

[Ji ji]

You can develop huge eyes then. They could have a different shape like an elongated sphere, hosted in an elongated skull as well. This would imply less eye mobility yet, and a narrower field of vision.

Another way to make up for the need of retinal surface is a spherical eye with multiple pupils and retinas. Three pupils closedly spaced, each with its own elongated camera, hosted in a single spherical eye, could double retinal surface within the same space.

Another solution is 3+ eyes.

Or maybe the vitreous humor could be substituted by biological optical fibers bringing the light to a multilayer retina, multiplying the retinal surface by the number of layers.