4
\$\begingroup\$

So, I recently saw this question and thought it looked simple enough. Instead of a webcam though i used an actual (but old, 3MP) digital camera. The camera was a bit broken anyway(always exposed with weird colors and too dark usually). Well, I've confirmed the IR works now by holding my TV remote up to it. I don't actually have any IR LEDs on hand though.

The two prevalent kinds I see seem to be 950nm and 850nm. For the purposes of making a night vision camera, which one would be more bright? I would think 950nm because it's closer to the visible spectrum, but everyone else seems to say 850nm is brighter. So which one works better? Or should I just buy one of each and see for myself which one is brighter?(as in, it depends on the camera)

\$\endgroup\$
  • \$\begingroup\$ Note that 850nm is closer to visible spectrum. Also, that won't matter as the CCD/CMOS pixels simply map intensity, regardless of wavelength. Color is determined by filtering. \$\endgroup\$ – tyblu Dec 12 '10 at 11:20
  • 4
    \$\begingroup\$ Not true - all light sensors have some difference in sensitivity at different wavelengths. You'd normally expect wavelengths closer to visible to be better, however other factors like different LED efficiency and possible slight eye visibility may mean that 950nm is a better choice. LEDs are cheap - just buy both types and try it. \$\endgroup\$ – mikeselectricstuff Dec 12 '10 at 11:51
  • \$\begingroup\$ Actually, yeah, buy both. \$\endgroup\$ – tyblu Dec 12 '10 at 11:54
5
\$\begingroup\$

EDIT:After seeing mikeselectricstuff's comment, I agree: buy both. Then upvote his comment.

Pick the 850nm one. As shown in the following image (from here), it is attenuated less than 950nm.

http://www.microscopyu.com/articles/digitalimaging/ccdintro.html

The green and blue portion of your camera's Bayer filter will attenuate the IR light considerably, making their channels noisy, so you may want to look into getting raw image data output such that it can be processed on your PC, rejecting those channels. The filter cannot be removed.

Be careful with your IR "flash bulb", or whatever it ends up being. Although we can't see it, it still hurts and damages eyes. Many animals/insects are more sensitive to and can even see IR.

\$\endgroup\$
  • \$\begingroup\$ Sadly this is an el cheapo camera without a raw output function. But I will get one or two of both wavelengths and see which one works best. \$\endgroup\$ – Earlz Dec 12 '10 at 17:46
  • 3
    \$\begingroup\$ This is a very old answer, but I have to say: QE curves (counts/photon) are not the same as responsivity curves (A/W). Since for every watt of optical power there are more photons for a longer wavelength than a shorter one, and typical detectors detect photon flux rather than incident energy, the QE curve needs to be weighted by wavelength. The responsivity maximum for an unfiltered CCD will normally be around 900nm. \$\endgroup\$ – Oleksandr R. Apr 7 '15 at 22:12

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.