Beyond Visible Light - Shooting in Infrared

A personal journey into the world of digital infrared photography.

First published in Nikon Owner Magazine in 2012, updated May 2014 with information about the Fuji X-E1"IR"

Chianti Vineyards, Italy  - August 2005

Modified Nikon D100 IR  - Nikkor 17-55 at 19mm 1/180 sec, f/8, ISO200
From my very first experiments in infrared I knew I had found my photographic medium - a place where texture and drama collide beyond reach of the visible.

There is always something unexpected in infrared photography - infrared light ‘sees’ much further than visible light, cutting through haze and pollution and revealing distant detail in the landscape invisible to the naked eye. Clouds take on three-dimensional shapes hanging in a deep black sky and water is ominously dark and cold. Trees and foliage become luminous white demonstrating the “Wood Effect” first discovered in the early 20th century.

Now in the 21st century digital technology and modified infrared cameras allow hand held shooting at a resolution which was unthinkable in the days of infrared film and we are entering a new era of detail and clarity - beyond visible light.
Avenue of Palm Trees, Valencia, Spain - September 2004

Unmodified Nikon D100
Cokin Infrared filter (740nm) - Nikkor 18-35 at 18mm, 10 sec, f/22, ISO200
My own journey in infrared photography started almost ten years ago in the Yorkshire Dales where I was taking part in a workshop with the celebrated landscape photographer Charlie Waite. There had been much discussion about the use of neutral density filters and what would happen if you stacked several filters together. As I was one of the only photographers in the group shooting digitally, with my brand new Nikon D100, it fell to me to take the “experimental” shot through four stacked NDs. The resulting image took on an unexpectedly reddish glow and we realised that something more was going on than just the attenuation of visible light. We were seeing the invisible effects of infrared light.

To understand this phenomenon we need to travel back not ten years but one hundred years. Although photographs made by infrared radiation had been made in the late 19th century, the first published infrared photograph was taken in 1910 by Robert Williams Wood (1868-1955), an American physicist and professor of optical physics at John Hopkins University in Baltimore. Wood had been fascinated for many years by what he called “invisible rays” and was credited with the discovery of electromagnetic radiation beyond the visible spectrum. He went on to develop photographic emulsions that could capture these rays and produced the first photographs of both infrared and ultraviolet radiation. Wood’s image of his house in East Hampton, reproduced to illustrate his scientific papers in the February 1910 edition of Century Magazine and the October 1910 edition of the Royal Photographic Society Journal, is thought to be the first published infrared photograph. The following year another image by Wood of a quarry in Syracuse, Sicily was exhibited in the Royal Photographic Society exhibition and published in the Illustrated London News. Both images show the characteristic whitening of healthy foliage that eventually became known as the “Wood Effect” and is a distinctive hallmark of many infrared photographs.

San Quirico d'Orcia, Tuscany, Italy  - August 2005

Modified Nikon D100 IR
 - Nikkor 17-55 at 55mm
1/100 sec, f/8, ISO200
The next landmark in infrared imaging came from the inventor of television, the Scotsman John Logie Baird (1888-1946). In the 1920s and 30s Baird developed a number of advanced television systems including large screen, colour and stereoscopic television and in January 1926 he demonstrated the first viable television system which could be sent by radio or over ordinary telephone lines. In that December he also showed the “Noctovisor” where visible images were sent from a room that appeared to be in complete darkness. Baird had originally used an ultraviolet light to illuminate his subjects but on discovering that this could damage their eyes he tried an ordinary light bulb coated with ebonite to block the visible radiation and allow the infrared light to pass. The following year the eminent scientist Sir Oliver Lodge was successfully “Noctovised” at a meeting of the British Association held in Leeds and in 1929 Baird held a public demonstration of the system at Box Hill in Surrey which drew the attention of huge crowds as well as the military, who thought this device could be used to see at night or in fog, a reality only fully achieved by RADAR in World War II.

During the 1930s the development of photographic emulsions was largely driven by the movie industry. Early colour film or orthochromatic film was sensitive to wavelengths in the 400-600nm band, principally blues and greens. This proved troublesome for motion pictures, rendering blue skies as perpetually overcast, blond hair as washed-out, blue eyes nearly white, and red lips nearly black. To some degree this could be corrected by makeup, lens filters, and lighting, but never completely satisfactorily. Kodak then developed a black and white panchromatic film, extending the sensitivity up to 700nm into the red band, producing a more realistic reproduction of a scene as it appears to the human eye. They also developed extended-red panchromatic emulsions that were sensitive well beyond the visible spectrum up to about 900nm, what we now call near-infrared. This early infrared cine film was often used to capture supposed night scenes under full daylight, a technique known as “Day-for-Night” or “nuit américaine” (“American night”).

Straw Bales, Normandy, France - August 2007

Modified Nikon D100 IR - Nikkor 17-55 at 55mm
1/500 sec, f/8, ISO200
By 1937 some 33 different types of black and white infrared films were available from manufacturers such as Agfa, Kodak and Ilford and The Times regularly published landscape and aerial infrared photographs. During the second World War a team at Kodak, working on camouflage detection, developed the first false colour infrared film, Kodachrome Aero-Reversal-Film, which became more widely available to the public in 35mm form during the 1960s and was popularised by the iconic images of Bob Dylan and The Band taken by Elliott Landy (b. 1942) at the Woodstock festival in 1969 where he was official photographer. The surprising colours and effects that Kodak’s colour infrared film produced fitted well with the psychedelic aesthetic emerging in the late 1960s and was used to great effect on album covers by Jimi Hendrix, Donovan, Frank Zappa and the Grateful Dead.

By this time the difficulties of shooting with infrared film were well understood. The film had to be kept refrigerated, loaded in complete darkness and used with a camera that didn’t leak infrared light. The slight front focussing of infrared wavelengths could be adjusted by using the red infrared index engraved on many lenses and special film pressure plates were available that didn’t reflect infrared light back into the film emulsion. Generally prime lenses were preferred as they produced more contrast in infrared images than the equivalent zooms and a wide range of visible light blocking filters were available ranging from the very dark red (29) to the opaque (79, 89b, 87c and 72), each producing unique results.

Yellowstone National Park, USA - January 2014
Modified Fuji X-E1 - Fuji XF14mm - 1/60th, f/8, ISO200
Then, in the early 1970’s, the digital era began. Following on from research done by Philips in the late 60’s, Steven Sasson, an engineer at Eastman Kodak developed the first digital camera using a charge-coupled device (CCD) image sensor in 1975. The camera weighed 3.6 kg, recorded black and white images to a cassette tape, had 10,000 pixels and took 23 seconds to capture an image. It wasn’t until the 1990s that the first commercially available digital cameras were available including Kodak’s DCS-100, based on a Nikon film body, and it was in 1999 that Nikon released the D1, a 2.7 megapixel camera costing some £5,000. Because both CCD and CMOS sensors are just as sensitive to near-infrared light as they are to visible light, the D1, like all subsequent Nikon D-SLRs and digital compacts, incorporated an infrared blocking filter just in front of the imaging sensor.

Although this filter is intended to block all of the infrared light from reaching the sensor, in practice a small amount does get through, especially in long exposures and this is exactly what we saw in Yorkshire on my D100 with the four ND filters in front of the lens: an image on the screen that looked like it had been shot through a deep red filter but with a strange glow to it. The filters were attenuating the visible light as expected but by chance were allowing far more infrared light to pass through and I had captured my first digital infrared image. A few weeks later I bought a Cokin 007 (89b) filter that appears almost totally opaque to the eye (it filters out all light below about 650nm) and began my infrared experiments in earnest!

When shooting infrared images with an unmodified digital camera you cant see anything through the viewfinder with the filter in place as it is there to block most or all of the visible light and only allow the near infrared wavelengths to pass. As the infrared light is then being severely attenuated by the camera’s blocking filter, exposures can be quite long even under bright sunlight, often several seconds at a moderate aperture which means that you are restricted to shooting with a tripod. As with infrared film, the focusing also has to be adjusted slightly and I would often find that to get a single useable image I would have to bracket both exposure and focus – the failure rate was high but over time I came up with a workable procedure which would usually yield a useable RAW file within a couple of frames. On the back of the camera all you see is a low contrast reddish image with a rather narrow histogram but once processed in Photoshop this system can produce beautiful black and white images with a slight softness and graininess that is not unlike infrared film.

Cloud Mountains, Islington, London - May 2011
Modified Fuji S5Pro IR - Nikkor 18-200 at 44mm, 1/30 sec, f/8, ISO125
By 2005 the first few images from modified DSLRs were beginning to appear on the Internet. The infrared blocking filter had been removed from these cameras and replaced with an infrared pass filter similar to the 89b I was already using. By this time my main camera was a D2X and my D100 was sitting in the back of a cupboard so I took the plunge and set about modifying it to work as an infrared only camera. Looking back now, I was extremely lucky that this modification worked so well!

As the only supplier of pre-made infrared filters for the D100 at that time was in the USA and they were mainly of the resin type, so easily scratched, I decided to have a bespoke Schott Glass RG780 filter made by an specialist optical firm in Sussex. This meant dismantling my D100 to reach the sensor, removing the blocking filter, measuring its size, shape and most importantly thickness and ordering an infrared replacement to be hand crafted. As its name suggests this filter transmits all light beyond 780nm only allowing the very deepest visible red light to pass. Once the filter arrived the process of reassembling the camera began, trying all the time to keep it free of dust, which proved to be rather harder than I thought. In the end I dismantled and reassembled the camera three times before I was satisfied that there was no dust trapped between the filter and the sensor. Dust is the single reason why I would advise anyone thinking of doing this today to have their camera professionally converted: you may think you have a dust free environment at home but you will never come close to the levels of cleanliness achieved in a professional camera-servicing workshop.

So more by luck than judgement I had a fully working infrared only D100 and what a joy it was to use. Now exposures were much as they would be for daylight, focussing could be checked on the LCD display and even the autofocus worked accurately after a little tweaking. As it turned out the Schott Glass filter was a perfect match for this camera and the results still please me today. This is in part because of the removal of the anti-aliasing filter, part of the same glass as the infrared blocking filter, which showed just what the D100’s 6 megapixel sensor was really able to produce. As the RG780 filter allows a small amount of visible light to pass this modified camera can produce false colour infrared images as well as black and white, much like the Kodachrome Aero film of old, though my preference has always been for the purely tonal world of black and white.

Cottage, Normandy, France - August 2007

Modified Nikon D100 IR - Nikkor 17-55 at 38mm, 1/160 sec, f/8, ISO200
The choice of lens for infrared photography depends on a number of factors. There are some lenses that never seem to achieve good focus with infrared light, others that exhibit terrible flare under practically all lighting conditions and some that are prone to central areas of over exposure or hotspots. All camera lenses have some degree of multi-coating on both external and internal optics but these are only designed to work with light at visible wavelengths. Establishing how they behave under near infrared light is a matter of trial and error and seems to vary from camera to camera but the one lens I have found to work consistently well under all lighting conditions is the AFS-Nikkor 18-200mm f/3.5-5.6 VR G ED DX. This zoom lens is unusual in that it retains its infrared focus point across all focal lengths allowing a modified camera’s autofocus to be calibrated consistently across the full range and it has become my firm favourite for infrared shooting.

Around this time a handful of other manufacturers produced D-SLRs without an infrared blocking filter, most notably Fujifilm who had a long-standing arrangement with Nikon, rebuilding Nikon film camera bodies around their own proprietary sensors to produce the FinePix S series cameras. In 2006 they released the FinePix S3 Pro UVIR, based on a Nikon F80 film body, aimed principally at scientific, medical and forensic markets, and this was succeeded the following year by the FinePix IS Pro, a special version of the S5 Pro based on a Nikon D200 body, without any filter blocking ultraviolet or infrared light. The unique Fuji designed SR sensor in this camera is highly sensitive to infrared light and its two photodiodes per photosite increase the captured dynamic range. Although these specialist cameras were never made available to the general public I bought a second hand S5 Pro in 2010 and had it professionally converted to a black and white only infrared camera with a 830nm filter cutting out all of the visible light and producing wonderfully smooth and contrasty images straight out of the camera.

Blenheim Palace, Woodstock, England - May 2011

Modified Fuji S5Pro IR
 - Nikkor 18-200 at 28mm, 1/50 sec, f/8, ISO125
Like its D200 cousin, the S5 Pro had the first version of LiveView, a revelation for infrared photographers. For the first time you were able to preview infrared images before shooting and in later cameras like the D300 use the LiveView focusing system to get accurate focus with out having to mechanically recalibrate the camera. However as cameras have become more complex they also become harder to convert successfully. Some full frame Nikons like the D700 use infrared sensors within the mirror box to monitor its travel and these can lead to fogging in longer infrared exposures unless disabled, and most current Nikons use small piezoelectric devices to shake dust loose from the blocking filter which sometimes have to be removed during conversion. Gone also are the days of manually tweaking the autofocus with an allen key, now it requires complex equipment and software.  Fortunately though there are a number of specialist camera workshops offering conversion services so there has never been a better time to reach into the back of that cupboard and give your old D-SLR an new lease of life as a dedicated infrared camera.

In 2011 Fujifilm launched their X system cameras starting with the X100 which I can honestly say I fell in love with at first sight.  I soon tried it out as an unmodified infrared camera with some success - the Fuji chip proved to have good sensitivity to IR wavelengths and at high ISOs it was possible to shoot in IR handheld, previewing the "live" infrared through the electronic viewfinder.  This live preview is nothing short of game changing for infrared photographers - now we can see the world in infrared before we shoot...

Hyde Park, London - April 2012
Unmodified Fuji X100
 - 1/6 sec, f/2, ISO400
Shortly after Fuji launched their X-E1 interchangeable lens camera I bought a body to be modified into a B&W infrared only camera with an 830nm filter.  It was a bit of a gamble as it was the first X-E1 in the world to be modified as far as I am aware but as soon as I received it I knew I had done the right thing.  Combined with Fuji's fabulous XF14mm prime lens this instantly became my favourite infrared camera with a dynamic range, sensitivity and resolution far in excess of anything the modified Fuji S5-Pro or Nikon D100 could achieve.
Yellowstone National Park, USA - January 2014
Modified Fuji X-E1 - Fuji XF14mm - 1/60th, f/8, ISO200
For me the joy of photographing in infrared comes from that constant element of surprise. However much I pre-visualise or pre-view the image I am never quite able to predict the way infrared light will interact with the landscape and the transformation from the out-of-camera RAW to the finished black and white image is as much part of the creative process as pressing the shutter. For the new infrared photographer that creative process will often be challenging but ultimately rewarding. After all, photographing the invisible was never meant to be easy!

Kanha National Park, India - March 2014
Modified Fuji X-E1 - Fuji XF14mm - 1/80th, f/8, ISO200
A fine-art book entitled “Beyond Visible Light – Infrared Photography” containing over 40 infrared photographs is available to purchase in both hardback and softback – go to www.simonweir.com and click on “Books”.

Yellowstone National Park, USA - January 2014
Modified Fuji X-E1 - Fuji XF14mm - 1/60th, f/8, ISO200 - Stitched from 3 exposures


Sources and links for further information:

Camera Conversion:
· Advanced Camera Services (Norwich) - www.advancedcameraservices.co.uk
· Life Pixel (USA) - www.lifepixel.com

Infrared filters:
· Hoya R72 (720nm) - www.hoyafilter.com
· Cokin A007 Infrared 89B - www.cokin.co.uk
· Lee 87 IR Poly Filter - www.leefilters.com
· B+W Dark Red (092) Infrared filter - www.bpluswfilters.co.uk/

Historical Figures in Infrared:
· Robert W Wood - http://en.wikipedia.org/wiki/Robert_W._Wood
· John Logie Baird - the inventor of Television - www.bairdtelevision.com/

Other sources:
· RPS / Infrared 100 - Andy Finney’s blog celebration 100 years since the 1st published IR photograph - www.infrared100.org
· Bjørn Rørslett - IR/UV photographer and Nikon Lens reviewer - www.naturfotograf.com
· Front and back focus adjustment on Nikon cameras - www.howtoshot.com/82/how- to-fix-front-focus-nikon-d90-d80-back-focus-problem
· Scientific information on sensor coatings and hotspots - www.sensor-film.com
· Scientific information on diffraction and IQ in infrared - www.infraredphoto.eu/Site/Diffraction_and_Image_Quality.html
· Clive Haynes’ very detailed and clear explanations on IR postproduction workflows - 
www.crhfoto.co.uk/crh/digital infra-red/digital-ir.htm
· Even more links can be found at Andy Finney’s excellent website - www.atsf.co.uk/ilight/ilightlinks.html

10 comments:

  1. Really cool pics Simon, but why so oversharpen?

    ReplyDelete
  2. These are low resolution images optimised for screen viewing so, depending on the device, they can appear over sharp. Rest assured they print beautifully !!! Cheers - Simon

    ReplyDelete
  3. Hello Simon
    Had already found your site and then today you were name dropped by ACS. I am going to convert my X100T (the X Pro 2 has rendered it redundant). I'm only interested in black and white (have had some good experiences with a screw in Hoya filter). I am agonising over 720nm vs 830nm. I can see you went with the latter. Any regrets? Are the exposure times still quick enough to shoot handheld? Is there any reason I should still be considering the 720nm?
    Sorry for all the questions!
    Cheers
    Ben

    ReplyDelete
  4. Hi Ben

    Thanks for your message - this is a question I get asked quite often!

    The choice of what filter to go for depends on your aesthetic more than anything:

    An 830nm filter only allows infrared light through - no visible light whatsoever. The RGB photo sites on the Fuji sensor are all equally transmissive to infrared so the converted camera will produce a monochrome image of the infrared light with a dynamic range comparable to an unconverted camera but with a slightly reduced sensitivity. This can then be processed just like any monochrome image in Lightroom / Photoshop or in Silver Effex Pro (now free thanks to Google so do try it if you haven’t already)

    A 720nm filter will see all of the infrared spectrum and some of the visible red spectrum as well. The visible red will be recorded at the red photo sites and the infrared will be recorded equally across all the photo sites - the resultant image will therefore have some limited “colour” information that can be enhanced in post-production to produce what is termed “false-colour infrared”. The channel swapping is simply to produce images with bluish skies rather than purple skies - more what the human brain expects to see.

    A 720 conversion will be slightly more sensitive than an 830 conversion by about 1.5 stops but as the Fuji sensor is already very IR sensitive this is not a deciding factor - in both cases the camera is plenty sensitive enough to be used handheld.

    If you intend to produce black and white infrared images then, although both conversions can be used, the 830 converted camera will produce images that are sharper and have a wider dynamic range as well as a fuller infrared effect. Of course an 830 converted camera cannot produce false colour images…

    Starting with a false colour image from a 720nm camera and “folding it down” to black and white means two compromises: a softer image because the visible and infrared light will have focussed in slightly different planes, and a reduced IR effect because your exposure will have to allow for the visible red light as well as the IR light in the red channel, reducing the amount of IR hitting the green and blue photo sites and therefore reducing the recorded dynamic range of the IR light.

    If all this sound mystifyingly technical than it boils down to:

    720nm - colour IR with a soft B&W option in post
    830nm - B&W IR only

    Hope this helps more than it confuses !! - Cheers - Simon

    ReplyDelete
  5. Thanks Simon

    Incredibly helpful. Thanks for taking the time for such a complete reply.

    I am a lazy, lazy JPEG shooter most of the time these days (spoilt by Fuji's JPEGs) so I can't imagine I am going to have the patience to much about with faux colour, channel mixing etc. (especially since they look awful if you don't get them spot-on). As long as the 830 is hand-holdable it sounds like it makes way more sense.

    I've also now realised that I can sell the X100T and have enough money for a body + conversion.... (way more sensible!). Any thoughts on XE1 vs XE2 vs XT10 (much of a muchness on the pricing).


    Thanks again!

    Ben

    ReplyDelete
  6. Hi Ben

    Sensor-wise there isn't much in it but of the three I would go for the X-E2. It has significant improvements over the X-E1 and will sit well with your X-Pro2 in terms of layout.

    A good place to look for one is on the Fujifilm website under "Refurbished" - essentially brand new cameras returned as "open box" by dealers and sold at a good discount.

    Check my blog as to which lenses work well in IR - some do and some don't...!

    Cheers - Simon

    ReplyDelete
  7. Thanks again Simon

    That does make sense. I did have an XT-1 for a while and never loved it but am used to that type of layout. Interestingly there appear to be more secondhand XT-10s around than XE-1s (including at Fuji). Might come down to availability.

    I did see your lens list - very useful (and similar to the Kolari results). Seems most of my primes would be pretty useless but I do have (but rarely use) the 18-135 which seems a good one by all accounts. At some point I will fold and buy the 14mm - has been on my mind for a very long time. I don't suppose you have tried the new 35 F2?

    Cheers

    Ben

    ReplyDelete
  8. I'd hold out for an X-E2 to come up either reconditioned or 2nd hand - remember ACS will give the camera a full service as part of the conversion - another reason why I like their work so much.

    Yes the 18-135 zoom is excellent in IR and I am sure you will weaken and buy the 14mm soon - it is my favourite! The old 35mm f/1.4 is superb in IR but I haven't yet tried the new F/2 version...

    Cheers - Simon

    ReplyDelete
  9. Hi Simon,
    Great images! Just getting started. Saw you at the sailing club recently when my wife and I visited to see your presentation.
    I would like to combine infrared with the 'big stopper' effect but it doesn't seem to work. Is there a way to lengthen the exposure time beyond stopping down and altering ISO?

    ReplyDelete
    Replies
    1. Hi Chris

      You will need a special infrared big stopper as normal visible light ones don't work at IR wavelengths.

      The one I have is a "Formatt Hitech 4x4 inch 16 Stops Firecrest IRND Neutral Density 4.8 Filter" which will set you back about £100 from Amazon.

      There is also a new company NiSi who I met with at the photography show who claim their BigStoppers work in IR but I haven't tested them yet... Search NiSi IR ND

      When using a filter avoid stoping down beyond f/8 as it will soften the image considerably - diffraction kicks in early in IR !

      Cheers - Simon

      Delete