Can I use my old Canon Xti camera as an IR camera without converting it?
In answer to a fellow photographer's question about IR filters and digital cmaeras vs. converting a camera to IR.
The XTi does not do native IR. Almost all digital SLR's have the capability to do IR and into several other spectrums too. However 99% of the cameras have an IR filter in front of the CCD or CMOS chip with various frequency cut-offs. When you have a camera converted basically they remove the IR glass filter and replace it with a clear (or cutoffs that allow IR while still blocking UV for example) piece of glass (that is over simplifying it).
The XTi has a somewhat narrower range on the IR filter in what light it blocks out which is how some people use it to shoot a quasi IR image. You may read about a lot of hobbyists doing this because it’s just a hobby and not worth the investment for them. Basically like using a butter knife for a screwdriver, it sort of works in a pinch and if you don’t need a screwdriver very often why not. Anyway it’s the attempt to add a Hoya R72 or similar filter (B+W 89B and 87C) to a digital camera that was made to be used with a film camera and IR film. It does not really work but it can be interesting, fun and inexpensive but IR photography it’s not. Oddly though some of these filters are very expensive ($70-$300.00) and you can get a camera converted for about $300.00 or do it yourself for much less. See http://www.lifepixel.com/ for conversion costs and image examples.
Here is what you have to consider:
With an IR converted digital camera you can shoot as normal and hand hold the camera shooting at normal speeds. You get to see the results on your screen just like any other shot you’d take on a normal camera. When shooting with a Canon XTi or other camera using an IR filter like a Hoya R72 (which is nearly a black filter to the eye) you will almost never be able to hand hold the camera. Typical bright daylight shots will require 10 to 20 seconds to make the exposure.
With a converted camera you do not need a R72 filter and you can shoot at regular speeds.
After you've made the exposure with an XTi camera and a R72 filter you have to decide which channel to use to make the quasi IR image. The RGB image will just look wrong with funky colors and halos, not very useful though. Most people will choose the red or green channel and then do a lot of tweaking in Photoshop. As a side note, the green channel captures about twice the image data on most chips. Again over simplifying but this is why green is the most often used channel to take data form when making a B&W image (also why the green channel is sometimes used to sharpen in normal shots while leaving the other channels alone to avoid noise issues, try this sometime). Channel sharpening is a whole other conversation though.
With the R72 filtered image you’ll spend a lot of time cranking out the contrast and doing channel mixes in an attempt to get close to IR. After all the effort you still do not get an actual IR image with the trademark black skies, milky smooth skin tones, and white flora.
With a converted camera you get a real RGB IR image and have all channels at your disposal. Converted IR cameras don't have to mess with the contrast and channel mixes or other Photoshop filters to get useful results. And when converting your camera you can even choose which kind of IR you want to do (color IR, B&W, and various levels in the frequency cut off). I'll stop there as to not get into the science too much here.
Next problem, the built in IR filter creates hot spots in the center of the image when using an IR cutoff filter. The tell-tail sign of a R72 filter on a digital camera is a hot spot in the center of the image that looks like a sunspot or flair. Many forums and blogs where amateurs discuss IR think this is just part of IR photography and come to expect it. Surprisingly when it’s not there some people think the shot is not IR because they’ve educated themselves into expecting the flaw. More experienced photographers will hide this as best they can in the composition. Since you're shooting so slowly with the filter you tend to slow down and consider composition a little more, which is a good thing. No firing off tons of shots and then going back to the studio to pick the one from 50 or 100 that you liked. Anyway with an IR converted camera you don’t have the hot spots, halos, and fewer lens flairs.
When comparing results of an XTi image with an R72 filter the results are much like many of the free filters and actions that do faux IR. They can be weird and artsy but not really a true IR result. For example the foliage although lightened will still have a green or blue tint where the real thing will be pure white flora while retaining detail and will not be blown out or haloing. Another mark of the faux IR regardless of if you do it with software, actions, or a R72 filter is blurry images. In the software it’s done on purpose as it is expected with the hobbyist crowd, again they don’t know better. The blurring is treated as if it were part of the art of IR where in fact it is not and was not with IR film and IR filters either.
Now since I've gone on too long I'll keep going for those insomniacs who I have not bored to death yet. You used to be able to buy a Canon Xt (350 I think) in Japan that has an IR code on it that Canon did not sell in the US. It had a limited run in Japan and is hard to find used today. Basically it was a narrower IR (red) filter on the chip that allowed IR photography while still blocking flair and haze. However, you had to use another filter on every lens to take normal shots. The thinking was that you would be allowed to do both IR and normal photography but you lost 1-2 stops on every lens you owned.
Leica also built a digital camera (the M8) recently that required another filter on all the lenses to take normal images. This however was because the IR was too sensitive and allowed more IR light in unintentionally. They tried to solve it with software and had recalls and other technical issues with the camera yet it’s still considered a very fine camera once the issues were resolved.
http://www.digitalcamerainfo.com/content/Leica-Upgrades-M8.htm
Really geeky IR filter info from B+W:
The nearly opaque Infrared Filter 092, which looks dark purplish red when held in front of a light source, blocks visible light up to 650 nm, and passes only 50% of the radiation just below 700 nm (thus the dark red color). From 730 nm to 2000 nm, transmission is greater than 90%. This makes photographs of pure red and infrared images possible making the best use of the relatively low sensitivity of infrared films. As the sensitization of infrared black-and-white films barely extends beyond 1000 nm, the red portion that is transmitted still makes a relevant contribution to the exposure. This makes the 092 the preferred filter for pictorial photography on IR black-and-white film. Its filter factor is 20 to 40.
This Infrared Filter blocks the entire visible spectrum, so to our eyes it looks completely opaque. Unlike the 092 infrared filters it makes pure infrared photographs possible without any residual visible red component. Its transmission only begins to exceed 1% at 800 nm, rising to 88% at 900 nm, and remains that high far beyond the upper limit of sensitization covered by infrared films. This filter is used less frequently in pictorial photography because of the dramatic loss of effective ISO. But in the scientific field, materials research and forensics, transmission strictly limited to the infrared range is often important. The filter factor is somewhat dependent on the illumination and on the characteristics of the film.
In addition, the 091 Dark Red may be used for more subtle Infrared Black & White photography; it transmits more visible light from the red end of the spectrum than the 092.
The XTi does not do native IR. Almost all digital SLR's have the capability to do IR and into several other spectrums too. However 99% of the cameras have an IR filter in front of the CCD or CMOS chip with various frequency cut-offs. When you have a camera converted basically they remove the IR glass filter and replace it with a clear (or cutoffs that allow IR while still blocking UV for example) piece of glass (that is over simplifying it).
The XTi has a somewhat narrower range on the IR filter in what light it blocks out which is how some people use it to shoot a quasi IR image. You may read about a lot of hobbyists doing this because it’s just a hobby and not worth the investment for them. Basically like using a butter knife for a screwdriver, it sort of works in a pinch and if you don’t need a screwdriver very often why not. Anyway it’s the attempt to add a Hoya R72 or similar filter (B+W 89B and 87C) to a digital camera that was made to be used with a film camera and IR film. It does not really work but it can be interesting, fun and inexpensive but IR photography it’s not. Oddly though some of these filters are very expensive ($70-$300.00) and you can get a camera converted for about $300.00 or do it yourself for much less. See http://www.lifepixel.com/ for conversion costs and image examples.
Here is what you have to consider:
With an IR converted digital camera you can shoot as normal and hand hold the camera shooting at normal speeds. You get to see the results on your screen just like any other shot you’d take on a normal camera. When shooting with a Canon XTi or other camera using an IR filter like a Hoya R72 (which is nearly a black filter to the eye) you will almost never be able to hand hold the camera. Typical bright daylight shots will require 10 to 20 seconds to make the exposure.
With a converted camera you do not need a R72 filter and you can shoot at regular speeds.
After you've made the exposure with an XTi camera and a R72 filter you have to decide which channel to use to make the quasi IR image. The RGB image will just look wrong with funky colors and halos, not very useful though. Most people will choose the red or green channel and then do a lot of tweaking in Photoshop. As a side note, the green channel captures about twice the image data on most chips. Again over simplifying but this is why green is the most often used channel to take data form when making a B&W image (also why the green channel is sometimes used to sharpen in normal shots while leaving the other channels alone to avoid noise issues, try this sometime). Channel sharpening is a whole other conversation though.
With the R72 filtered image you’ll spend a lot of time cranking out the contrast and doing channel mixes in an attempt to get close to IR. After all the effort you still do not get an actual IR image with the trademark black skies, milky smooth skin tones, and white flora.
With a converted camera you get a real RGB IR image and have all channels at your disposal. Converted IR cameras don't have to mess with the contrast and channel mixes or other Photoshop filters to get useful results. And when converting your camera you can even choose which kind of IR you want to do (color IR, B&W, and various levels in the frequency cut off). I'll stop there as to not get into the science too much here.
Next problem, the built in IR filter creates hot spots in the center of the image when using an IR cutoff filter. The tell-tail sign of a R72 filter on a digital camera is a hot spot in the center of the image that looks like a sunspot or flair. Many forums and blogs where amateurs discuss IR think this is just part of IR photography and come to expect it. Surprisingly when it’s not there some people think the shot is not IR because they’ve educated themselves into expecting the flaw. More experienced photographers will hide this as best they can in the composition. Since you're shooting so slowly with the filter you tend to slow down and consider composition a little more, which is a good thing. No firing off tons of shots and then going back to the studio to pick the one from 50 or 100 that you liked. Anyway with an IR converted camera you don’t have the hot spots, halos, and fewer lens flairs.
When comparing results of an XTi image with an R72 filter the results are much like many of the free filters and actions that do faux IR. They can be weird and artsy but not really a true IR result. For example the foliage although lightened will still have a green or blue tint where the real thing will be pure white flora while retaining detail and will not be blown out or haloing. Another mark of the faux IR regardless of if you do it with software, actions, or a R72 filter is blurry images. In the software it’s done on purpose as it is expected with the hobbyist crowd, again they don’t know better. The blurring is treated as if it were part of the art of IR where in fact it is not and was not with IR film and IR filters either.
Now since I've gone on too long I'll keep going for those insomniacs who I have not bored to death yet. You used to be able to buy a Canon Xt (350 I think) in Japan that has an IR code on it that Canon did not sell in the US. It had a limited run in Japan and is hard to find used today. Basically it was a narrower IR (red) filter on the chip that allowed IR photography while still blocking flair and haze. However, you had to use another filter on every lens to take normal shots. The thinking was that you would be allowed to do both IR and normal photography but you lost 1-2 stops on every lens you owned.
Leica also built a digital camera (the M8) recently that required another filter on all the lenses to take normal images. This however was because the IR was too sensitive and allowed more IR light in unintentionally. They tried to solve it with software and had recalls and other technical issues with the camera yet it’s still considered a very fine camera once the issues were resolved.
http://www.digitalcamerainfo.com/content/Leica-Upgrades-M8.htm
Really geeky IR filter info from B+W:
The nearly opaque Infrared Filter 092, which looks dark purplish red when held in front of a light source, blocks visible light up to 650 nm, and passes only 50% of the radiation just below 700 nm (thus the dark red color). From 730 nm to 2000 nm, transmission is greater than 90%. This makes photographs of pure red and infrared images possible making the best use of the relatively low sensitivity of infrared films. As the sensitization of infrared black-and-white films barely extends beyond 1000 nm, the red portion that is transmitted still makes a relevant contribution to the exposure. This makes the 092 the preferred filter for pictorial photography on IR black-and-white film. Its filter factor is 20 to 40.
This Infrared Filter blocks the entire visible spectrum, so to our eyes it looks completely opaque. Unlike the 092 infrared filters it makes pure infrared photographs possible without any residual visible red component. Its transmission only begins to exceed 1% at 800 nm, rising to 88% at 900 nm, and remains that high far beyond the upper limit of sensitization covered by infrared films. This filter is used less frequently in pictorial photography because of the dramatic loss of effective ISO. But in the scientific field, materials research and forensics, transmission strictly limited to the infrared range is often important. The filter factor is somewhat dependent on the illumination and on the characteristics of the film.
In addition, the 091 Dark Red may be used for more subtle Infrared Black & White photography; it transmits more visible light from the red end of the spectrum than the 092.
Comments