The glory days of infra-red (IR) are past. Most IR photography was military, forensic or otherwise technical in nature, and 'pictorial' IR photographers rode on the coat-tails of heavy users such as the armed services (IR is excellent for detecting camouflage) and the police. Today these heavy users have almost all converted to digital imaging or to other imaging techniques, inclusing thermal imaging, so IR films are ever rarer.

Kodak's High-Speed Infrared (HIE) has been threatened with extinction for many years, and by the time you read this, it will probably have disappeared altogether. Likewise, barring (extremely unlikely) reintroductions, false-colour IR film is a thing of the past, though we have included a couple of images shot on Kodak's last Infra-Red Ektachrome, mainly for nostalgia's sake and completeness.

Even so, monochrome IR films reappear periodically, and there is always the digital option. Most digital sensors are inherently highly sensitive to IR, and there are wide variations in the efficacy of the IR filters built into cameras to counter this. Some are so effective as to leave little or no IR sensitivity, while others are removable, non-existent (though this is rare) or simply not very effective.


Reservoir, Malta


This is possibly Roger's favourite among all his IR pictures, not least because it is an absolutely standard landscape with very conventional composition, lent drama by the IR rendition. Strong and intricate contrasts of light and dark often make for attractive pictures, as much with IR as with any other monochrome film.

This was shot on Maco-branded IR, which is no longer available but is almost certainly the same film as Efke IR820c. Roger used an M-series Leica and 35/1.4 Summilux (pre-aspheric) with a B+W 092 filter (T50 695nm -- the meaning of T50 is explained rather later in the module). Generous exposure -- EI 3 on a hand-held meter -- gave the best exposure in a series of brackets at 3, 6 and 12.

reservoir near mdina
st. jouin

St Jouin de Marnes

The Leica M8 is notorious for its high IR sensitivity, a result of the necessarily thin IR filter which is in turn a consequence of the small flange/film distance of the M-series Leica. It can therefore be used as an IR camera simply by fitting a filter with a high T50 such as the B+W 092 used here. The results are not outstandingly sharp or tonally attractive, but they are certainly easy: you can even leave the camera on 'auto', though you may need to put in a small correction factor if you don't like the exposure.

What is IR?

'Infra-Red' literally means 'below red', i.e. of longer wavelengths (lower frequencies) than the human eye can see. The human eye does not however have a sharp cut-off between deep red and IR. In common photographic parlance, therefore, 'IR' includes light beyond about 700nm which is quite readily visible through a 'black' or 'opaque' IR filter after a few seconds' accommodation. The practical limit of human vision, except for very bright flashes of light, is probably beyond 750nm for the vast majority of people. The distinction between 'near IR' and 'true IR' is flexible but 750nm is one possible cut-off point. We prefer 800nm; this more accurately reflects the range of sensitizing dyes available.

Pots, Knossos

Frances shot this with Konica 750 (no longer available) and, as far as she recalls, a Nikkormat and 35/2.8 PC-Nikkor fitted with B+W 092 filter (T50 695nm). The camera was tripod mounted; she focused, then put the filter on and made the necessary focus correction.

Often, with IR of any kind, more is down to the quality of the light than anything else. Here, a deep blue Cretan sky has registered as a solid black, and fairly generous development of the film has given high contrast. Similar results could have been obtained in this light with almost any IR film, 'near' or 'true' -- but every film has its own signature, and 'similar' is not the same as 'the same'.

Anything that reflects a lot of IR will appear very light. This includes foliage, which explains the classic 'white foliage' look. Subjects with no IR will look very dark, or even black. There is next to no IR in a blue sky, so it will appear black with appropriate filtration.

IR is not, however, the same as heat, at least in photographic terms. Yes, there are thermal-imaging cameras, but they respond to much longer wavelengths than photographic IR and have correspondingly lower resolution. The idea that the pallor of IR portraits is a result of the film recording the heat of the human body is complete codswallop. At anything below about 400 degrees C, the heat is unlikely to record; a severe fever by anyone's standards.

American Gothic

Tonality of IR films varies widely, according to exposure and development, and you don't always get the corpse-like pallor that many associate with IR portraits. What you normally do get, though, is very dark, inky shadows, though these can be ameliorated with extremely generous exposure. The flower is of course a sunflower.

Roger shot this on Efke IR 820c with a B+W 092 filter having a T50 of 695nm. We have forgotten what the lens was, but it may have been the (new) 50/1.5 Voigtländer Nokton as this is one of the few rangefinder lenses we have that accepts our 52mm 092.

film choice ir
False-colour IR films

These worked by using a conventional multi-layer colour film, but sensitizing one of the layers to IR. This was normally recorded as bright red, though a moment's thought will reveal that this was pure convention. Once you represent something you can't see, as something you can see, it can hardly be 'realistic'.

Place des Vosges, Paris

The sky looks comparatively natural but the foliage is red, flesh tones are a sickly yellow, and blue jeans (the figure on the right in the group) register as red or orange. The yellow path is rather lighter than in reality and -- as already noted, a typical phenomenon -- the shadows are very dense indeed. Technical details as below.

notre dame

Notre Dame de Paris

Like the picture above, this was taken (by Roger) on Kodak's last E6-compatible Ektachrome Infra-Red (EIR) using a fixed-lens Konica rangefinder camera with a medium yellow filter; it was easier to use a separate camera, pre-loaded, than to risk reloading.

The stonework is curiously desaturated against the deep blue of the sky and the red of the foliage, which points up the very low latitude of the film: +/- half a stop produced very marked differences. Here, half a stop more would have 'blown' the stonework while half a stop less would have meant blocked-up foliage and sky. It was an interesting film, but we never warmed to it.

The 'Green Gap'

Silver halides are inherently sensitive only to blue, violet and ultra-violet light. In order to sensitize them to other colours, dyes have to be added to the emulsions. First came green sensitization (orthochromatic or 'right-coloured'), then orange and red sensitization (panchromatic or 'all-coloured'), then deeper red sensitization (hyperpanchromatic -- almost certainly the origin of the name of Ilford's 'HP' series of films) and IR.

All dye sensitization gives extra speed, at the cost (with hyperpan films) of excessive red sensitivity, which can make reds and even skin tones look light: this is why some people like very weak blue filters with hyperpan films for portraiture.

Because a single dye will not give reasonably even sensitization to all colours, it is normal to use two or three or even four dyes in a hyperpan or near IR film. If you look at the sensitization curves you can see the peak for each dye. If you use only an IR dye -- the traditional approach with 'true IR' films (see below) -- then the sensitivity of the film to green will be negligible: this is the 'green gap' that you sometimes hear about.

A green gap means that you get strong IR effects with quite modest red, orange or even yellow filters. These are sufficient to block the inherent sensitivity of the film to blue/violet/UV light, so that the film records only red and IR. If however the film is sensitized to green as well, you need a much deeper red filter or the effects of visible light will swamp the effects of IR. There is more about filters below.

clouds, mono lake

Mono Lake, California

There is no white foliage here, because Frances used only a light red filter, with a T50 of about 500nm, with Ilford SFX. She could probably have got a very similar image using Ilford HP5 Plus, but with a red filter, she prefers the tonality of SFX. The clouds were not noticeable to the naked eye; she was surprised when she examined the negative.

Hyperpan, near IR and 'True IR'

The distinction between 'hyperpanchromatic' and 'near IR' is obviously blurred and indeed many recent 'near IR' films originated as hyperpanchomatic traffic control films sensitized only into the near IR, about 740-750nm or so, and with no green gap. This is the origin of both Ilford SFX (formerly SP815/816T) and Maco Cube, and probably of Konica 750 as well.

The numbers in IR film names are typically well rounded, and refer to the upper limits of sensitivity. Thus Konica 750 has its last dye peak rather before 750, and Efke's 820c, a 'true IR' film, is sensitized to longer wavelengths than the Konica but does not have a dye peak at 820nm; rather, this is the limit of its sensitivity, tapering off from a dye peak at an earlier wavelength.

ir sunflowers

Apocalyptic sunflowers

Never neglect the possibility of abusing materials, if it gives you the effect you want. This was the least exposed image of three bracketed shots on Efke 820c, and in the conventional sense it is underexposed and excessively contrasty. On the other hand, the sunflowers seem almost to be attacking the photographer, as if they were Triffids, and the sky is certainly dramatic enough. Roger finds this picture (taken by Frances) to be rather disturbing. She used a 50/2 dual-range Summicron and a B+W 092 filter (T50 695 nm) on a tripod-mounted Voigtlander R2.

Kodak's great HIE was sensitized almost to 900nm and there were special-application (usually military) films sensitized to 1100nm and more. Sensitization beyond about 1300-1350nm is pointless for most applications as the IR is absorbed by water vapour in the atmosphere.

The longer the wavelength to which the film is sensitized, the greater the technical problems. The dyes are not very stable, and decay; the film becomes more and more sensitive to poor storage (warmth); and there is ever greater difficulty in bringing both visible light and IR to a common focus. Some lenses are much better than others (see below) but all benefit from stopping down to f/8 or less, at which point depth of focus takes ever more care of the discrepancies between visible focus and IR focus, and depth of field takes ever more care of absolute focusing errors. Even so, IR pictures from 35mm are seldom super-sharp: bigger formats, enlarged less, are usually quite a lot sharper. Many of the really sharp IR shots from the 1930s were shot on 4x5 inch film.

pylons and buildings, malta

Pylon and buildings

Frances (who shot this) says that she has 'a love-hate relationship with it'. To Roger, it's a classic indictment of insensitive holiday-island development, a parody of the neatly terraced fields that characterized Malta when he lived there as a boy in the 1950s. It's a typical example of using Ilford SFX for its grittiness rather than as a 'true IR' film. Nikkormat FT3, 35/2.8 PC-Nikkor, deep red tri-cut filter (T50 c. 645 nm).

IR film choice

As noted at the beginning of the module, IR films are increasingly a 'boutique' product, and choosing among the limited range available may depend as much upon what you can get as upon preference. Even so, two points are worth making. One is that the 'dripping with light' effect that so many associate with IR is to a large extent a matter of halation and irradiation: light bouncing around inside the emulsion, as well as being reflected off the back of the film. At least one IR film was offered without an anti-halation layer to help create this effect (Maco Aura) and some others have clear film bases to the same end (most 35mm films have grey bases to reduce 'light piping' and to assist in suppressing halation).

The other is that most hyperpan/near IR films (such as Ilford SFX) can be loaded conventionally under subdued light, but 'true IR' films (sensitized beyond about 800 nm) need to be loaded and unloaded in pitch darkness or there is a virtual certainty of some degree of fogging. At worst, most of the film may be unusable; this is by no means uncommon. If the instructions advise loading in complete darkness, heed them, and remember this includes unloading too. Most (but apparently not all) changing bags are IR-tight.

Note also that not all film canisters are totally opaque to IR, so if you are carrying true IR films in anything other than subdued domestic lighting, it may be a good idea to wrap the canister (or the whole box) in aluminium foil.

Fog, bridge

Efke IR820c is a 'true IR' film that must be loaded and unloaded in pitch darkness, and as you can see from the tonality of the unfogged bit, the results are not necessarily all that different from 'near IR' films; but you can also see the fogging that resulted from not loading in pitch darkness.The clear film base (the 'c' in '820c') does not help. Leica M-series; 50/2 DR Summicron, B+W 092 filter.

ir fog
Choosing cameras and lenses for IR

Concerns about loading bring us to another potential problem. Some modern 35mm SLRs do not count frames mechanically, but instead use an infra-red sensor to count the sprocket holes. We have never used such a camera for IR and can give no advice, but only a warning. Apparently some cameras fog a lot worse than others. Older, mechanical cameras without motorized advance should always be safe.

A problem that is not obvious until you think about it is focusing through a near-opaque filter. There are three main options here.

The first is what we normally do: use a rangefinder or other direct-vision camera so that you do not have to focus through the filter.

The second is our preferred choice with cameras where you have to focus through the lens: quickly-detachable (QD) IR filters. In this case you need to use the camera on a tripod or to use a separate direct-vision viewfinder on the camera. You can equally well use a normal screw-in filter if you are willing to accept the extra delay or to scale-focus and use the ancillary finder.

ir camera options

The third approach is one we have not tried: cut a sheet of IR gel filter (Ilford's is fine) and fix it just in front of the shutter on an SLR camera. Of course you can do the same in many leaf-shutter cameras too, if you merely want to protect the filter.

The drawback to the last approach is of course that any dust or dirt on the filter will record as a shadow on the film. You also have the difficulties of fixing the filter in place so that it is not dislodged by the normal operation of the camera, and of being sure it is still there.

Filter options

The M2 Leica (lower right) is fitted with a B+W 092 filter. If you don't want to spring for a Leica, consider a low-cost fixed-lens rangefinder such as the Konica, lower right. The old Yashica SLR can be used in one of two ways with the QD spring-clip gel filter holder, either focused without the filter, or scale-focused. Framing for hand-held shots is via the Tewe viewfinder on top.

A very few cameras are reputedly less than opaque to IR and will fog the film. This was apparently more of a problem in the days of bellows cameras than today, and even with bellows cameras, we have never had a problem; but if you get otherwise inexplicable fogging, this is another possibility, as are film sheaths in darkslides that are not IR-opaque (we have yet to encounter any that are not). You might also think twice about using plastic-bodied cameras with true IR films.


Traditional 'achromatic' lenses were designed to bring light of two colours to a common focus (red and blue). 'Apochromatic' or 'Apo' lenses may be cubic (three wavelengths, red, green and blue) or more rarely quartic (four wavelengths), but unless you compute a lens specifically for IR it will not necessarily bring IR to a common focus.

In order to compensate for this, you normally need to focus the lens slightly closer than you would for visible light. The classical advice was to rack it out by about 1 to 1.5 per cent of the focal length, e.g. 1.5 to 2mm for a 150mm lens, but with a helical focusing mount this is rarely an easy option, so some lenses are furnished with an IR focusing mark. Once you have focused visually or by rangefinder, you transfer the reading from the main index to the IR index. After a while, you can learn to make the tiny correction that is required by touch alone.

st. martin

Church of St. Martin, Noizé

Even on an overcast day, a strong IR filter such as the B+W 092 used here (T50 695nm) can deliver dramatic white foliage effects, but a cloudy sky will not go black in the same way as a clear blue sky -- compare this shot with the Cretan picture (the pots as Knossos) above.

Even quite small amounts of haze can lighten the sky considerably: here, the sky has been burned in somewhat (in Adobe Photoshop) to make it darker, though the road is quite an impressive river of black without any manipulation whatsoever.

Roger shot this using his M8 and (as far as he recalls) the 50/1.5 Voigtländer Nokton. The odd thing is that he had used the 50/2 Dual-Range Summicron several times for close-ups before realizing that it will not focus further than a few metres on the M8, because of the cam design.

If the camera does not have an IR focusing mark, try the depth-of-field mark for f/4, f/5.6 or f/8 (focusing the lens closer), but be aware that there may be a significant focus shift on stopping down, so that different degrees of correction may be required for different distances. This is equally true if there is an IR focusing mark, and is far from invariably the case.

Leica DR Summicron

The IR focusing mark on this lens is the little dot beside the red R, between the f/4 and the (unmarked) f/5.6; here, the infinity mark has been set against it. In practice, with Efke 820c, the f/4 depth of field mark seemed to work better. Such tests have to be performed on a lens-by-lens and film-by-film basis.

ir focusing mark
Developing tanks

Yet another problem that we have read about, but never encountered, is plastic developing tanks that are not fully opaque to IR. Once again, we mention it in the interests of completeness.

palms, goa

If the film has a green gap, as described above, then even a dark yellow filter will give clear IR effects (white foliage, black skies). Most current films don't have a green gap, though, and digital sensors in consumer cameras never do, so if you want these effects you will have to use a deep red or very deep red (visually all but opaque) filter.

IR filters are normally described in terms of their 'T50' (the number is often written as a subscript) which is the wavelength at which they transmit 50 per cent of the light. An ordinary deep red tri-cut filter typically has a T50 on just under 650nm but the weakest red that will normally give significant IR effects has a T50 of around 695nm; you may do even better with a T50 of 715-720nm. Such filters are visually opaque at first, though you can see through them after a while if you hold themn close to your eye and exclude all other light. You can buy filters with a T50 of 780nm or more but for obvious reasons these cannot be used with those films whose sensitivity has either disappeared altogether by this wavelength or has at least declined sharply, i.e. many or possibly all current IR films.

Palms, Goa

This is one of Frances's favourite Ilford SFX pictures, taken with a Nikkormat and deep red (but still visually transparent) tri-cut red filter (T50 c. 645nm) on her 35/2.8 PC-Nikkor. It is rather unusual in that the tonality of the print is 'upside down', with the lightest areas at the bottom and the darkest at the top.

Film speed and metering

The inevitable problem of filtering out more and more light is that film speed falls, so that by the time you are using filters with a T50 of 695 to 720nm, most IR films except HIE are best rated at something between EI 3 and 12 on a hand-held meter. The exact speed will depend on your meter, the subject, the time of day (there's more IR around noon) and personal preference: more exposure often gives the 'dripping with light' effect that most people associate with infra-red. We've never used enough HIE to make recommendations here but it is faster. Through-lens metering may or may not work: it's surprisingly good with the Leica M8, for example.

ir farmland

Pastures, Pyrenees

Later in the day -- this is an hour or so before sunset -- you often need to prolong your exposures by a stop or two, even though the light appears to be quite red. After sunset, exposures become very long indeed; too long to be practical, in many cases. Leica M-series, Efke IR 820c, B+W 092 filter (T50 695nm)

The very low shutter speeds that are necessitated by these film speeds limit hand-holding, the more so as it is generally a good idea to shoot at f/8 or below in order to cover up errors in focusing. Using the 'sunny 11' rule (more realistic than 'sunny 16' with most negative films) you'd need 1/3 to 1/12 second (call it 1/2 to 1/15 second on most shutter speed dials) at f/11, or 1/4 to 1/30 at f/8. Shooting at apertures wider than f/5.6, unless you have a reliable IR focusing mark or have extensively tested the film/lens combination, is pushing your luck; that's only 1/8 to 1/60 second, with 1/30 a safer bet than 1/60.

ir church portugal

Chapel, Portugal

This shot clearly illustrates that you can get 'true IR' effects with Ilford SFX, using an IR filter; we forget whether this was the B+W 092, T50 695nm, or the Ilford gel with its T50 of 715nm, but both would give much the same effect. It also shows clearly that exposure can be tricky: much more IR was falling on the sunlit side of the building than on the shaded side. Frances shot this from the platform on top of our Land Rover, using a Nikkormat and her 35/2.8 PC-Nikkor.

What else do you need to know?
ir castle

Not very much, really. Shooting IR is one of those things where, once you have absorbed the minimum amount of theory, the only thing is to go out and try it. Our own view is that for certain subjects, in particular ruins, graveyards, old churches and similar 'spooky' subjects, IR has become so over-used as to run a great risk of slipping into cliché. On the other hand, the black skies and haze-cutting ability can be dramatic; the latter can even be achieved with quite modest tri-cut filters with a T50 of 645nm.

ir portugal

Ruined castle, Pyrenees

A classic cliché shot, taking with a Leica on Efke 820 IR with a B+W 092 filter (T50 895nm); the lens was probably either a 35/1.7 Voigtländer or a 50/2 Summicron, as we have screw-in UV filters only in 39mm and 52mm, and both lenses accept the former. The ruined castle glowers against a black sky, with eerie white foliage tumbling down in front of it. So?

Southern Portugal

Frances loaded her Nikkormat with Ilford SFX for this shot, and put a tri-cut red (T50 about 645nm) on the lens; as far as we recall, her 35/2.8 PC-Nikkor. The camera was tripod mounted and the picture was taken at about f/11 in the interests of maximum sharpness. The unsharpness of many IR shots is the result of camera shake: low effective film speeds beg for tripods.

The bottom line

As we were working on this module we realized that many of our favourite IR shots were actually made with deep red filters (T50 of around 650nm) rather than with filters having a T50 of 695nm or more. Dramatic skies, and haze-cutting ability seem to us to produce better pictures, more often, than classic IR 'white foliage' shots.

Of course, this may be sheer personal prejudice, but one of the things we try to do in the Photo School is to present contrarian views: all too many people just don't think to question standard Received Wisdom, but just blindly attempt to reproduce results they have seen elsewhere without being entirely sure how to do this (which we hope this module has addressed) and, more importantly still, without thinking about whether the technique they are trying to emulate is suitable either for the sort of subjects they shoot or for their personal vision.

ir mani

Mani landscape

The Mani in the Pelopponese is an unsettling area, dotted with fortified houses like minature castles: you can see one in the distance. These are relics of centuries of blood feuds. Somehow, this photograph seems to capture the savagery of the place. Frances used a Nikkormat with a 35/2.8 PC-Nikkor and deep red tri-cut filter, loading her camera with Ilford SFX, her favourite among all the IR films we have tried. Yes, it's grainy and moody. And grain and mood are what you want (or need) sometimes...


Go to the list of modules

or go to the home page

or support the site with a small donation.


© 2006 Roger W. Hicks