I've done a lot of research regarding filters as I'm about to do a modification on my DSLR for astrophotography. I've settled on removing the existing IR filter and replacing it with one the will still allow H-alpha (656 nm) and S-II (672 nm) plus the visible spectrum below that. My question is..if the eye can see red to approximately 700nm (before it reaches near-IR), why do most stock IR filters cut off below 650 nm ?

With few exceptions, data for infrared filters are not published for the region from 700 to 900nm. This is because most infrared filters are offshoots of the Wratten gelatin filters recipe. These were made by dissolving dye in gelatin and then floating the liquid gelatin on the surface of water. The gelatin was then allowed to gel and then lifted from underneath via a wire frame. The dried gelatin filter was then over-coated with lacquer. Often these gelatin filters were sandwiched between two sheets of optically flat glass.

Anyway, infrared radiation is strongly absorbed by matter. Most gelatin filters begin to absorb infrared beginning at about 2000nm and fully absorb at about 3000nm. Most glass infrared filters fare worse. The typical absorption curve is generally published for the region 750nm to 3000nm

To pass the frequencies you need, the filter would likely need to be made of quartz. I think none are available at affordable prices. Perhaps you can find such a filter under “specialized filters for science”. I don’t think you will find them in the general photographic community.

  • Why are you mentioning 2000 and 3000nm? The question was asking up to about 700nm which is fine for glass. Silicon detectors for the most part don’t go to 3000nm anyway. – Eric Shain Nov 7 at 0:20
  • @ Eric Shain -- Photographers generally think about colors (frequencies) a filter blocks, and rarely think in terms of frequencies transmitted. UV filters are used to block ultraviolet but an IR is generally mounted to transmit IR. Most photographers never venture into scientific imaging (medical, astronomical, etc.) these often image in ranges far from the visual range. For the most part: Infrared spans 700 900 nm. Since infrared radiation is emitted by warm bodies, by tungsten lamps, by electronic flash, and LED, etc. there are many potential situations to image in this region. – Alan Marcus Nov 7 at 17:20
  • I completely understand and agree with what you just related, its just that it is not relevant to the question which was about wavelengths within human perception (<700nm). – Eric Shain Nov 7 at 18:05

Camera imaging sensors are much more sensitive to longer wavelengths than human vision. The normalized response of the "red" sensitive cone cells drops off rapidly above 640nm. A camera without ir filters will be much more sensitive to red light that contains material percentages of photons with longer wavelengths. Digital color cameras have such filters built-in just to prevent overly strong response to images containing high amounts of longer wavelength light. These filters attenuate some long wavelength light that is visible to more accurately render color perception. Such items include flames and red hot embers, but also everyday things such as vegetation. Green leaves typically have high reflectance levels at long wavelenths. Without an IR filter a color camera will produce a yellowish cast to green leaves in daylight.

However, these IR filters have too sharp a cutoff when trying to capture longer wavelengths in astro-photography. Replacing these with filters that pass specifically desired, longer wavelengths allows better capture of this light. It comes at the cost of poorer color rendition for general camera use.

Eric, in the comments, posted a link that discusses a wide range of wavelength issues in color rendition. A good overview which includes graphs of various sensor and filter responses.

  • Thanks for the insight guys...makes sense. Spencer photography and LifePixel both do an Astro conversion with a replacement filter that passes visible up to 700nm. That gives me the option to filter for H-beta, OIII, H-alpha , S II and RGB as needed. – 77pro Nov 7 at 3:57
  • You link doesn't really compare human perception to silicon detectors very clearly. I think fen-net.de/walter.preiss/e/slomoinf.html has a pretty good plot. – Eric Shain Nov 7 at 18:14
  • @EricShain. Camera sensors, sans filters, have responses that don't vary much between 600nm and 800nm. Some technology becomes more sensitive while others the reverse. But they all are quite responsive in comparison to human vision with it's strong drop above 640nm. Your link contains good info. Here's another than is specific to why color cameras need IR filters. adimec.com/… – doug Nov 7 at 20:48

Your Answer


By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

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