Nigel Richards Photography

Experiments in Infrared

About     Exhibitions     Contact     Links     Home   Buy 

 

 

Experiments in Infrared by Nigel Richards

I was always interested in the results from infrared photography but with film cameras never really got around to actually trying it for myself. 

Growing plants and vegetation reflect infrared light and look lighter than normal when taken with an IR sensitive camera, blue sky looks very dark and will contrast greatly with clouds. It's an interesting, dramatic effect.

Digital cameras normally have an infrared cut filter built into the filter that is mounted internally in front of the sensor, this is to stop the infrared light from reaching the sensor and causing unwanted effects however all filters are not the same strength and some let more infrared light through than others. Experimentation is needed.   

The purchase of a cheap infrared pass filter gave me the opportunity to experiment with various digital compact cameras I have access to. 

The Infrared pass filter is mounted in front of the lens and tries to stop any visible light reaching the sensor, only infrared light is allowed to pass. The internal infrared cut filter immediately in front of the sensor tries to block  infrared light but invariably lets some through depending on the strength of the filter. The consequence of this is that some infrared light gets to the sensor and very little visible light. 

Initially I bought a gelatine Ilford SFX filter in Cokin P mount size, not a pure IR filter but close, to the eye it is opaque. I held this in front of a number of digital compact cameras and Digital SLRs to see the effect, most of them gave a reasonable infrared effect but the exposure needed varied enormously from “instantaneous” to long exposures of 30 seconds or more.

Canon Ixus 950 IS

The most sensitive camera to IR light was a Canon Ixus 950 IS camera, exposure were as short as 1/50 of a second in sunlight allowing the camera to be hand held, the results were very interesting and showed a dramatic infrared effect, however some images did tend to show a hot spot, a circular light area in the middle of the photo and also exhibited a lot of digital noise making good quality large prints very difficult.

To mount the gel filter on the camera for serious testing I improvised with  a holder made from a discarded plastic 35mm film case with a hole cut in the end to mount a cut circle of the gel filter, this  coincidentally just pushed over the front of the Ixus lens when extended.

Photograph taken with Ixus 950 IS and SFX filter. 1/20 second exposure.

Canon 5D

The next experiment was with a 77mm Wratten 72 equivalent infrared filter bought very cheaply from China via e-bay for my Canon 5d full frame sensor camera and the Canon 17-40mm lens.  Exposures were very long, with the lens stopped down sometimes in the region of 30 seconds so a tripod was essential. The 5D mk1 doesn't offer live view so composing the shot was difficult because the SLR viewfinder was black (the filter is opaque),  a clip on finder in the hot shoe would make life easier. The results were excellent, sharp and with little noise even when iso 800 was used. The slow shutter speed added an almost ethereal quality to the images. No hot spot was evident. 

Biddulph Grange, Canon 5d, 10 second exposure, Wratten 72 equivalent

Middlewich, Canon 5d, 1 second exposure, Wratten 72 equivalent

Olympus D380.

The long exposures and the black viewfinder made me wish for a better solution so having bought an Olympus D-380, 2 megapixel camera from a car boot sale for £1, I decided to strip it and see if it was possible to replace the in built infrared cut filter that is in front of the sensor with plain glass to allow infrared light to reach the sensor. From research I believed that this would give me short exposure “hand held” photography, with the option to compose the image on the screen or in the viewfinder. I do repair the cameras in my collection so this wasn't as daunting a task as it sounds.

First make sure the camera is actually working………

The following is not a precise description of the process but will give a good indication of the major steps in the conversion.

Searching on the internet can give useful information on disassembly of many cameras.  The modification is totally at your risk, I hold no responsibility if you damage the camera or give yourself a nasty shock from the flash capacitor.

 

The converted Olympus D-380 2.0 megapixel camera is shown above with a small piece of Ilford SFX Infrared filter held under the plastic lens surround.

Stripping the camera to get to the sensor was easier than I had thought, in this particular model there is basically one printed circuit board, a metal shield, the screen with a ribbon connector and various screws that hold the board and the lens assembly in position, once removed the sensor is accessible.  

It is a good idea to take digital photographs of every stage of the process in case there are problems when it comes to re-assembly. I keep a digital camera permanently on my desk for that process. 

The lens assembly is held in place by a couple of gold coloured screws, no unsoldering was needed.

  

This shows the D-380 with the back cover removed showing the screen, the shield and the main circuit board. 

The camera is shown with the lens assembly separated from the main board revealing the sensor.  The flash capacitor is the black tube on the right of the picture.

Care should be taken with the flash capacitor because even with the batteries removed it can still hold a lethal charge. It may be safer to learn how to discharge the capacitor before proceeding, search on the web, it is easy to touch the contacts with your fingers, I speak from experience.

Once the lens assembly is separated from the board the sensor can be seen, don't touch the surface of the sensor because it will be almost impossible to clean.

The IR cut filter is held loosely in the lens assembly by a rubber plate, it can all fall out when you separate the lens assembly from the sensor. The wires for the auto iris mechanism can be unplugged from the board to make things easier. 

The back of the lens assembly showing the IR cut filter behind a rubber plate.

The rubber plate is easily removed and then the filter will fall out. 

I carefully measured the filter size and thickness so I could make the clear glass replacement.

 

The replacement clear glass filter shown mounted on the rubber plate. The rough edges can be seen.

The replacement clear glass filter is made of several layers of clean microscope slide and cover glasses bonded together with a UV curing glass glue. It must the same thickness as the original filter to make the camera optics focus correctly. I rough cut the glass before bonding and then trimmed it to the correct size with a diamond tipped cutting disk. The actual sensor sensitive area is much smaller than the new glass so you can get away with slightly rough edges.  I cleaned and re-cleaned the glass at every step to avoid bonding dust into the assembly and covered the outer surfaces with tape when I trimmed it to size. 

Re-assembly is simply the reverse of the above process.

As well as replacing the IR cut filter with clear glass you also need to introduce the IR pass filter in the light path to stop the visible light from entering.

 With the D-380 there is a thin plastic surround to the lens glued on to the front of the camera. It has the openings for the lens and viewfinder, if this is lifted up it is possible to put a small piece of  Ilford SFX gel filter under the plastic and in front of the lens. (see the first photo of the camera above)  It may also be possible to adjust the critical focus of the lens whilst this cover is removed as there is a hole giving access to the lens mechanism. This hole may need to be enlarged.

Once the camera is re-assembled batteries can be put in and the camera tested.  If it doesn’t power up, dismantle it again and check everything is ok, refer to the photographs you took during the disassembly process.  First tests were from the bedroom window.  

First test of the converted D-380, hand held with an automatic 1/125sec exposure, no post processing in Photoshop, the colours are typical.

With a fresh set of batteries, sun in the sky and an empty Smart Media card the first real photographs were taken on a trip to the Wirral.  The light house at New Brighton with a blue sky, white clouds and low tide proved an ideal subject.  Again the camera was hand held and the resulting image was converted to monochrome and the levels adjusted in Photoshop.

The first competition I entered with it was as a digitally projected image which from a perspective of ultimate resolution is less critical  and won – not bad for a two megapixel camera.

 

New Brighton lighthouse taken hand held 1/400 second with the converted D-380

Olympus D-550

The next camera to be converted was an Olympus D-550 3.2 megapixel, auto focus, zoom camera, another car boot sale find at £2.  The conversion was remarkably similar to the D-380 but I was very careful to get the replacement clear glass filter exactly the correct thickness so the autofocus would focus correctly, this is very important, I was even able to mount the gel infrared filter internally in front of the new filter.  Results were sharper and far less noisy than the D-380 but recorded images with a slightly different colour, presumably because it is because of a different design of sensor.  The zoom lens and auto focus helped give results that could be printed to an acceptable size for print competitions.

 

Hand held 1/320 second photograph taken with the IR modified Olympus D-550

 What's next?

The results and the short exposures of the IR converted camera stand for themselves, even when the camera is a simple low resolution compact, I think this is the way to go for the future. 

I did tried converting an old 2 megapixel Canon Powershot, it was very complicated compared to the Olympus cameras with several circuit boards and many ribbon connectors, unfortunately the focus drifted badly when it was zoomed, sharp at the tele end but miles out at the wide, so for the time being it has been discarded, next will probably be a digital SLR. With this intent a £5, car boot, non working Canon 350D is sitting on my desk waiting to be repaired and then internally converted.

The new filter for a DSLR will be much more of a problem as it is far bigger and from experience will show dust etc. much easier. Commercially made replacements can be obtained but they are very expensive.  As is self evident my price limit for a camera to convert is about £5.  I am considering buying a small sheet of optical quality IR Pass plastic and installing this instead of the existing anti aliasing/IR cut filter that is immediately in front of the sensor. It would have to be the correct thickness and compatible with the different lenses that may be mounted on the camera.  Again searching on the internet will indicate which companies can supply suitable filters. I am expecting accurate auto focus will be a problem, complicated because I lack the equipment to re calibrate the focus of the camera body electronically. 

The cameras are one thing and we all love cameras (don’t we?) but photographically you need to understand what the infrared sensitivity does to the image, I suspect that you can have too many photographs that look like they were taken in mid winter after a snow storm.  Once you have an IR capable camera get out and take photographs in lots of different conditions and learn from the results obtained.

The gel Ilford SFX filter now seems almost impossible to obtain and alternatives need to be found, not tested yet but a possible alternative is a Lee 87 polyester filter. Some suggest that unexposed but developed E6 slide film also works!

 

Nigel Richards MA ARPS - Aug 2010

 

Contact:  nigel.richards01@virgin.net

From Above - Transition no.1 From Above - Transition no.2
Fruit & Veg Trafford Park Village 1984
EDF Building - La Defence Experiments in Infrared
 

All images & text on this page © 2010 Nigel Richards