Even better: the Thermal Flashlight. Makes a cruder picture, but way cheaper:

http://publiclaboratory.org/tool/thermal-photography

PLOTS (Public Laboratory for Open Technology and Science) is worth checking out. They're a group that's trying to put technology into the hands of citizens to make a difference.

I was under the impression alot if not all digital camera's had a filter for IR that on some is easier to remove than others.

Seen many a article on how to convert a common digital camera though have not tried any myself so for example this one http://www.petapixel.com/2010/10/20/how-to-convert-a-cheap-d...

Note that removing the IR filter the camera will still pick up visable light so you could add a filter to remove that so you just get IR or if the firmware is hackable then you can could see what opensource flavours are out. Though failing that most photo edit software will cater for that with after filtering. But you will be doing most IR in the dark to eliminate the sun factor, unless you want to see how much the sun warms area's.

I'm sure somebody has done this type of hack and can comment better with converting digital camera's to IR and what models are best for the price, possibly ebay 2nd hand DLSR's but who knows?

Yeah, I modified an old CVS disposable digital camera to do this once. Those filters only block near IR (as mentioned in the article), basically light just outside our vision spectrum. You can use infrared LED's and a camera thus modified to make a cheap night vision camera (This is, I believe, how the night mode on old sony cameras would work).

Measuring heat, however, requires far IR, which won't be picked up by removing the filter.

Exactly. Near and far IR are different beasts. Not only are the sensors different, but you need to make the lenses out of different material (or use mirrors).

Interesting I didn't think it could be done.

From that I understand normal thermal cameras can't have a silicon CCD and need to be actively cooled.

Actively cooled IR sensors can achieve higher resolution and accuracy, but ambient temperature IR sensors have been around for a while. As long as the incident IR can drive the sensor element away from ambient temperature, then you can have sensing.

IR sensors don't work by detecting heat changes. They convert infrared photons directly to electrical current.

Now I don't actually know jack about IR sensors, but I would assume that increased temperature means increased noise, as it does with visible-light CCDs. People refrigerate visible-light CCDs too, for taking low-noise astronomical photos.

Posting this because some people seem confused about IR.

I'd just like to point out here that infrared is a really wide chunk of spectrum. It goes from 0.74µm to 300 µm (according to Wikipedia), which is a factor of about 400. By comparison, the visual spectrum has a width of about a factor of 2, and yet our biology still gives us four different sensors crammed within that little sliver. Imagine if you had a rainbow with red, orange, yellow, green, blue, indigo, violet, and 50 more colors which were just as distinct from each other, and you get an appreciation of how big IR is.

Near infrared is the stuff used by artistic IR photographers (film and digital), remote controls, and night vision cameras. Regular CCDs designed for visible light also pick up near IR, so most cameras have a built-in filter which blocks it out. This is what you remove to turn your camera into an IR camera, and it also accounts for why some things just show up wrong in some digital pictures. Maybe you've seen a fireplace with purple flames -- that's because the IR filter isn't getting rid of all the near IR.

Near infrared is really just light that you can't see. Most things don't emit near infrared unless they're very hot (like fire) or designed to do so (like LEDs).

Far infrared is used in thermal imaging. It is a completely different beast, with much longer wavelengths.

1) Thermal infrared is low-resolution. The resolution of an imaging system is limited by the aperture of the lens relative to the wavelength of light that you are using. A proficient photographer with a nice prime lens and a tripod can come close to the resolution limit of visible light with no trouble, but the (much) longer wavelength of far infrared means that the relative size of a lens is much smaller. You simply cannot get a sharp picture of thermal infrared due to the increased diffraction.

2) Thermal infrared requires different sensors. Your standard CCD simply does not pick it up, you need something like HeCdTe.

3) Thermal infrared requires different lenses. Far infrared does not pass through glass like it does through air, you need to use a different material to make the lens, or just use a mirror.

4) Thermal imaging is more sensitive to thermal noise. That's why you cool the sensors. All sensors have thermal noise, even visible-light sensors. You will get cleaner pictures by refrigerating your visible-light CCD too, but it's a much bigger deal in far IR.

5) Atmosphere absorbs IR. Well, not all of IR. But the atmosphere is fairly opaque in some IR bands. If you build a sensor at 6-7µm you might as well submerge the thing in pea soup, the atmosphere is that opaque.

So [near] IR photography and [far IR] thermal photography are completely different beasts.

So they built basically an IR tracker. Talk about dual use.

Nice, happy ghost hunting everyone!

AFAIK, most houses (not apartments) in the US have forced air circulation systems along with heating. Basically this means a fan will suck air out of the house and push air from outside, even with all windows closed.

So your advice is likely wrong for a lot of people. YMMV.

Forced hot air systems get their air from inside the house via the "cold air return", then they heat it or air condition it then pump it out the vents. Nothing comes from outside.

IMO A few air leaks is not a bad thing as they provide some nice fresh air.

Or use a heat exchanger.

Classic Engg.