Nick’s infrared & ultraviolet photography page

This website is compatible with Lynx, Netscape, or any other browser you may choose to use. For more information, see Best Viewed With Any Browser.

Non conçu pour un fureteur particulier. Utilisez ce que vous voulez.

Spectral response of Olympus C-750. Wavelengths are approximate. Sunlight shining through freshwater fish tank, which probably accounts for the dark (absorption) band around 1100 nm. Infrared image uses filter with a cutoff of 850 nm (sharpened to improve contrast; not so grainy in original), superimposed on visible-light image (no filter). Wikipedia values (from article "color") are supplied for reference but are not aligned with the photo’s colors.

Diagram from 1929 Encyclopaedia Brittanica (vol. 18, p. 877), showing the entire known elecromagnetic spectrum at that time. Usually nowadays, 700 nm (7000 angstroms), or the low 700's are taken to be the visible-invisible IR boundary; has human vision gotten worse since the 1920's? Actually, 800 nm light is said to be barely visible if very bright (much like barely visible UVA light), and LED's emitting in the 800's are usually seen as dim red, while LED's emitting in the 900's are totally invisible to the human eye, even when very bright at close range, without a trace of light seen (e.g. those 920 nm LED's tested below). Whereas a 375 nm UVA LED is visible at close range, but it is otherwise invisible; the fluorescence it causes suggests that it is much brighter than it seems.

Ultraviolet seems to be just as invisible to the C-750 as to the human eye. The lens glass and cement both are said to strongly attenuate UV. A 375 nm UVA LED is only visible at close range.
 
The UV blocking is performed mostly by the eye's lens, which absorbs the UV and converts it to heat and/or light of a higher wavelength (the lens fluoresces very noticeably under a blacklight). Some people who have had cataract surgery, such as a professor at St. Louis University, are able to see UVA and even a trace of UVB down to about 300 nm, if their lens is not replaced, or is replaced with a UV-transmitting lens, as was common in the 1980's and before. As the wavelength of light decreases in the visible range, we see blue then violet; barely-visible UV-A may appear lilac, but a "blacklight blue" lamp will appear dim violet from its 404 nm mercury line, with the lower UV wavelengths invisible. However, a person without a UV-absorbing lens will see that "invisible" UV as light blue since the retina responds to wavelengths down to about 300 nm. But unlike other wavelengths of monochromatic light, visible UV "doubles back" in appearance so that, for example, 360 nm and 450 nm look similar, whereas "visible" monochromatic light's appearance has a unique correspondence to its wavelength (see chromaticity diagram below).
 
The painter Monet documented the bluish appearance of UV by painting one of his favorite scenes, water lilies, before and after his surgery:
 

Monet's water lilies with and without UV vision

On December 10, 2007 I received a 375 nm UV LED which I had ordered. Its wavelength is "just below the visible range". Its light is easily visible if looking directly at the LED, but shining the light on something other than a mirror or an object that is right next to the LED shows its invisibility. Is the light purple appearance of the lit LED due to a trace of higher wavelengths being emitted which are visible, or to 375 nm being visible above a certain intensity? The visible-invisible boundary appears similar to that of the eye on the C-750, and very long exposures at maximum ISO do not pick up anything other than what the eye sees. Sometimes white-balance adjustments may be needed to make the photo appear as it does to my eye; this is even necessary with some dark violet flowers (whose visible wavelength is close to the UV, and which also likely has a UV pattern visible to birds and insects).

First 375 nm UV photo (in the dark) Envelopes and textbook

Envelopes and textbook with room light on

With the (60 W overhead incandescent) room light on, the UV is not visibly reflected from the textbook page (which does not fluoresce noticeably, while envelopes and postal markings fluoresce brightly from the incident UV. The purple seen on the envelope, except perhaps for the very center of the beam, is visible light from fluorescence, not UV from the LED.

With the room light off, one can see a circular beam, including over the non-fluorescing textbook which reflects that pale lilac that seems to be the visible color of 375 nm light. The center of the beam is more washed out, whitish-blue, and the parts shining on the envelope would be showing a mixture of reflected UV and visible-light fluorescence.

(I'm using the term "visible light" loosely here to refer to the wavelengths normally considered visible, over about 400 nm; it does not include UV light visible to the human eye and camera.)

Head-on view of LED light appears purple; washed-out white from intensity at center

Long exposure (F2.8 16 seconds, ISO 400) shining LED from foreground in dark room towards DVD player at wall at night. Sodium-vapor streetlight across street visible from window. DVD player's LCD visible, and fluorescence of paper in foreground visible, but no UV visible.

The intent was to see if a long exposure would reveal light that was not visible to my eye.
 
Compare an invisible infrared LED (see 920 nm tests below) which are easily visible to the camera. A film camera without zoom would likely be very sensitive, as ordinary film responds well to UV (but not to IR).

Looking in the same direction with room light on; the paper which fluoresced in the previous photo is visible in the foreground, and the paper below it is weakly fluorescing from the 375 nm LED which is being shined through my camera's "UV filter" which does not appear to noticeably reduce longwave UV transmission (it's likely just plain glass, useful to protect the lens only, though it may be useful in reducing haze at high altitude with a large amount of solar UVA/UVB). The unlit LED's around the rim of the filter are the 920 nm infrared LED's tested below.

Viewing the security strip on a $5 bill on a sheet (which also fluoresces from UV).

Closeup of lit 375 nm LED showing its internals, lit only by its own light. Reddish area at left may be chromatic aberration.	Another closeup of 375 nm LED using its own light. Bluish hue due to camera white balance; appears pale violet to eye, as shown in the previous photos.
Underexposure results in truer "color". Side view of lit 375 nm LED using its own light.	Appearance of 375 nm LED in room light, with LED off. Same $5 bill in background.

Violet: DEC VT 220 (yes, it still works!)	Light green: DEC Professional 350 monitor (it still works too!) Note reflection of LED's light, to lower left of fluorescence, is the color of the LED, while the fluorescence is obviously a higher wavelength.
Light blue: Toshiba TV/VCR	Light blue: NEC Multisync 50 (while typing in the results of these tests using EDT)

Shining LED on hallway carpet with lights on: nothing visible.
Unretouched photo of UV-illuminated carpet with lights off. Underexposed on camera but easily visible to the eye.	Contrast, brightness and gamma adjusted to see image clearly, as with the eye; there's a stain on the carpet that would otherwise be invisible but for the fluorescence and/or UV reflectivity of the stain.

Fluorescence of orange plastic elephant (from a margarita at Maracas). The envelope at lower right, and scrap of paper at upper left, also fluoresce, but the wooden desk does not; it only reflects the orange light.	Control picture of same objects illuminated only by tungsten room light (i.e. 60 W overhead incandescent). Both pictures using F8 1/5 sec, ISO 400, tungsten white balance, no flash, so this one is intentionally underexposed.
Weak fluorescence of Lysol solution in plastic container when UV is shone through the plastic.	Fluorescence of Lysol much stronger when no plastic to block the UV. Width of fluorescent area suggests the incident UV is much brighter than it appears. The Lysol is normally yellow.
Shining 375 nm UV through two layers of plastic, the lid and the bottom, with the Lysol solution in between. Container is polypropylene (recycling code "5"), brand name C.P. Auto white balance with ISO 400, exposure compensation -2.0, results in reddish cast of white ceilings and walls under incandescent room lights, and underexposed graininess, respectively, so container and LED are not overexposed.

The two lamps in the kitchen, viewed with Auto white balance to match their appearance to the eye. The faint purple reflecting off the walls makes me wonder if there's some UV getting through as well, but could not find any data on UV emission by these bulbs. Lamp to the left is 60 W incandescent.	Same two lamps on sheet (that was shown fluorescing, above). Control picture with same white balance as following infrared photos. Taken with room lights on, which is mostly removed by the infrared filters. The left lamp has its plastic cover on, and the right one is bare.
Same, with 720 nm infrared filter (Hoya R72), and 16 second exposure.	Same exposure level, with 850 nm infrared filter, showing greatly reduced intensity above this wavelength. The bulb cover no longer diffuses light at this wavelength, and the IR is primarily emitted from the ends of the bulbs where the filament may be incandescing. There is also a small amount of increased intensity in the lower bulb at center right, due apparently to reflection.
Same exposure level, with 950 nm infrared filter, showing almost no IR emission above this wavelength, except for some emission from the ends and that center-right bright spot. This totally invisible (to the eye) light appears violet to the camera; one could say "extremes meet" with regard to the two ends of the visible spectrum as seen by a silicon CCD in a camera.	Same exposure level, with 950 nm infrared filter, and room light turned off (the 60 W incandescent). This shows that the center-right spot was a reflection from the incandescent, as was much of the light which appeared to be from the fluorescents, which indeed emit almost no light above 950 nm except for a very weak emission at the ends.

Sparta, NJ July 2005 (720 nm)	Waymart, PA July 2005 (720 nm)	Delhi, NY July 2005 (720 nm)	Hamden, NY July 2005 (720 nm, "snow in midsummer" effect)
Bear Mountain, NY July 2005 (visible light, note hazy sky)	Bear Mountain, NY July 2005 (720 nm, cuts through haze)

Simulated visible-infrared image (all other images are as seen by camera, not the unaided retina

Approximation of appearance to me of naked-eye-visible IR + far red (only R72 passes any visible light; the visible disk of the sun* is probably from an ordinarily infinitesimal passage of visible frequencies in the 700’s, as it’s a desaturated red vs. the whitish foliage seen otherwise. *Warning: looking at bright IR sources like the sun is still hazardous even if most of the IR is not visible to the eye! (it can still burn the retina after maybe 1/4 second, and IR penetrates closed eyelids easily.)  February 2007

Local & Test Photos
Mineola Long Island July 2005 (visible & 720 nm, "snow effect")		Mineola Long Island July 2005 (visible & 720 nm)	Comparing IR LED on top (from TV remote) to red LED's on bottom (from taillight) (no filters used; some IR filtered by camera)
Sun & tree Mineola, L.I. July 2005 (720 nm)	Camera vs. webcam (single 920 nm LED)	Relative brightness of 920 nm test LED & TV remote LED's
Building IR illuminator for camera (920 nm LEDs seen in visible light)	Darkroom photo of hand using IR illuminator (920 nm LEDs)	RGB+ test pattern (visible light & 920 nm IR illuminator)	Animated GIF of 32-LED finished IR illuminator at various voltages (webcam)
920 nm IR illuminator (using 12 of 32 LEDs) August 2006	same, in dark (using 12 of 32 LEDs) August 2006
South Beach Staten Island May 2007 (950 nm)	South Beach Staten Island May 2007 (950 nm)	In my garden Staten Island May 2007 (950 nm)

Your comments are welcome: Click here to send me mail!

This site has been viewed times since May 3, 2001.

There have been visits to this home page since Thursday, July 12, 2007.

Page created:   July 12, 2007
Last modified:  Mar. 10, 2008