HOME | DD

Description
A hot spot is a noticeable color aberration, usually circular, that appears in varying amounts in digital infrared photographs. Some lens and camera combinations seem to be hot spot free, but most lenses show some degree of this problem.

Photographic lenses are not designed with IR in mind. IR focuses differently than visible light. It reflects and diffracts differently than visible light. The lens elements and the inside of the barrel of any given lens may be prone to uncontrolled reflection and diffusion of IR. There may even be reflection from surfaces inside the camera body. Also, different cut-off frequencies of IR filters can reveal differing degrees of hot spotting.

In its most obvious form, the hot spot is an area in the center of the photo that is a contrasting color to the rest of the image, fairly saturated and rendering the elements in that area flat and softer focus. In many cases the hot spot is less obvious and can be managed during post processing.

The most subtle hot spot effect is when there is a very slight soft-focus/glare effect across most of the image, with some vignetting in the corners. This is easily edited in post processing. It may be cropped, or corrected without too much effort. When the initial contrast adjustment is made, focus appears to sharpen up and the haze will often disappear. The impression of a color cast will be reduced, and further color editing will eliminate it completely. (When rendering an image black and white the hot spot often disappears.)

The obvious hot spot is the ruin of many images. It may be difficult to perceive on the camera display as you are producing the picture. So there are several strategies to reduce the presence of hot spots in digital IR images. They may differ by equipment, as using a filter over the lens is vastly different from using a converted camera with a filter in front of the sensor.

What both methods can benefit from is avoiding shooting angles that would cause flare and glare in normal photography. In fact, a little bit MORE care  is necessary because IR is a bit more diffuse than visible light, and a shaded lens for visual light may not block every stray ray of IR light. Much of the charm of IR is the halo created by the light on each piece of vegetation; but that shouldn't render as general glare and lack of sharpness.

The amount of hot spot can vary unpredictably on a given lens, whether zoom or prime, relative to the aperture. For example, it may be present wide open, then disappear one f-stop down, BUT then become progressively more evident at each smaller aperture. With a zoom, that progression can vary with the focal length used. There are many lenses which show little hot spot until they reach a middle aperture, but quickly degrade when stopped down. If a lens is going to produce a hot spot, the smaller apertures will almost always make it more distinct.

If you are lucky, you may find a truly IR compatible lens . . . when used with a filter on the lens. That same lens used with a converted camera may yield really obvious aberration. Since the 'hot mirror' is meant to exclude IR in an unconverted camera, when combined with an IR passing filter on the lens, weaker and indirect rays of IR may not reach the sensor. Still, as you stop down, the dreaded aberration may appear due to IR bouncing around among the glass elements in the lens.

I encountered that problem with my oldest favorite work-horse lens, a Tamron 28-105m f2.8 that was fine with filters on the lens. But when I got my converted Canon Xs/1000D - OH! what a disappointment - at any aperture smaller than f4 the hot spot is beyond editing. There is distinct vignetting wide open. I won't be using this lens for IR.

I tested each of my lenses. None of them are entirely free of the problem. The 50mm f1.8 prime comes closest. Even that has a slight degree of vignetting at all settings. All that being said, I have produced editable images from each lens! Not that I want to spend the effort editing if it is unnecessary, but knowing the IR characteristics of each lens means I have options.

Most lenses show the most aberration at smaller apertures. So if you have a lens that is acceptable at 5.6, then you simply use it to that limit. If you pay attention to the elements that cause stronger hot spotting - shooting angles near the sun, specular highlights or high contrast in the image, and so on, then you can get remarkable results from even the most limited lens. Use a lens shade whenever possible, and/or shoot from a shaded spot- (Maybe have your assistant carry an umbrella? Maybe not...)

I will mention that my converted point and shoot shows no hot spotting AT ALL. NONE!  (A Canon Powershot A630 converted with a 665nm filter. ) My DSLR, with its larger files, RAW capabilities, and generally sharper captures is also hot spot prone. So far. I have not had the opportunity to try any Canon L series lenses; I suspect there may be considerable differences due to superior optics and coatings and build quality.

I have found each of my Canon lenses on lists of lenses supposedly 'good' for IR and 'not suited' for IR. This suggests that the performance of an average lens for IR photography is somewhat subjective and variable. Test your lens and you will have the knowledge to make the best use of it.
-----------------------------------------------------
               My Tests and Some Results

Technical disclaimer: I am not very scientific, or proficient at engineering, so you may very well want to research and develop your own testing protocol. . .

The tests I performed were somewhat casual but sufficient for my purposes. All I did was set the camera on a tripod on a sunny day, and make a reasonable exposure of an average scene at each aperture. For the zoom lenses I did that at three focal lengths: the widest, the longest and at somewhere near the middle point.

I set up automated processing with a curves adjustment for contrast and a channel swap. No other enhancement. I compared the images in aperture sequence for each lens, and focal length, and rated the degree of hot spot on a 5 point scale, with 1 being no visible aberration and 5 being very defined and saturated. Using this approach I now have a good idea of what the performance of each lens will be.

One thing I'd never realized before was that the hot spot can be wider diameter than the frame, thus creating a subtle vignetting effect. That kind of effect is editable, as I said earlier. Even the best lens showed some of that. Of my six lenses I will be able to get the most use from my EF-S 18-55, which is pretty good from f11 to wide open. I have an old EF 28-80 which, strangely, needs a custom white balance versus my other Canon lenses. It is very sharp and shows no hot spot down to between f11 and 16. Since I will rarely need  telephoto when I shoot IR, these lenses will be fine for walk-around shooting.

I hope this article is useful for anyone new to IR or who has been frustrated by the hot spot problem.

Questions, comments, corrections always welcome.

Description
A hot spot is a noticeable color aberration, usually circular, that appears in varying amounts in digital infrared photographs. Some lens and camera combinations seem to be hot spot free, but most lenses show some degree of this problem.

Photographic lenses are not designed with IR in mind. IR focuses differently than visible light. It reflects and diffracts differently than visible light. The lens elements and the inside of the barrel of any given lens may be prone to uncontrolled reflection and diffusion of IR. There may even be reflection from surfaces inside the camera body. Also, different cut-off frequencies of IR filters can reveal differing degrees of hot spotting.

In its most obvious form, the hot spot is an area in the center of the photo that is a contrasting color to the rest of the image, fairly saturated and rendering the elements in that area flat and softer focus. In many cases the hot spot is less obvious and can be managed during post processing.

The most subtle hot spot effect is when there is a very slight soft-focus/glare effect across most of the image, with some vignetting in the corners. This is easily edited in post processing. It may be cropped, or corrected without too much effort. When the initial contrast adjustment is made, focus appears to sharpen up and the haze will often disappear. The impression of a color cast will be reduced, and further color editing will eliminate it completely. (When rendering an image black and white the hot spot often disappears.)

The obvious hot spot is the ruin of many images. It may be difficult to perceive on the camera display as you are producing the picture. So there are several strategies to reduce the presence of hot spots in digital IR images. They may differ by equipment, as using a filter over the lens is vastly different from using a converted camera with a filter in front of the sensor.

What both methods can benefit from is avoiding shooting angles that would cause flare and glare in normal photography. In fact, a little bit MORE care  is necessary because IR is a bit more diffuse than visible light, and a shaded lens for visual light may not block every stray ray of IR light. Much of the charm of IR is the halo created by the light on each piece of vegetation; but that shouldn't render as general glare and lack of sharpness.

The amount of hot spot can vary unpredictably on a given lens, whether zoom or prime, relative to the aperture. For example, it may be present wide open, then disappear one f-stop down, BUT then become progressively more evident at each smaller aperture. With a zoom, that progression can vary with the focal length used. There are many lenses which show little hot spot until they reach a middle aperture, but quickly degrade when stopped down. If a lens is going to produce a hot spot, the smaller apertures will almost always make it more distinct.

If you are lucky, you may find a truly IR compatible lens . . . when used with a filter on the lens. That same lens used with a converted camera may yield really obvious aberration. Since the 'hot mirror' is meant to exclude IR in an unconverted camera, when combined with an IR passing filter on the lens, weaker and indirect rays of IR may not reach the sensor. Still, as you stop down, the dreaded aberration may appear due to IR bouncing around among the glass elements in the lens.

I encountered that problem with my oldest favorite work-horse lens, a Tamron 28-105m f2.8 that was fine with filters on the lens. But when I got my converted Canon Xs/1000D - OH! what a disappointment - at any aperture smaller than f4 the hot spot is beyond editing. There is distinct vignetting wide open. I won't be using this lens for IR.

I tested each of my lenses. None of them are entirely free of the problem. The 50mm f1.8 prime comes closest. Even that has a slight degree of vignetting at all settings. All that being said, I have produced editable images from each lens! Not that I want to spend the effort editing if it is unnecessary, but knowing the IR characteristics of each lens means I have options.

Most lenses show the most aberration at smaller apertures. So if you have a lens that is acceptable at 5.6, then you simply use it to that limit. If you pay attention to the elements that cause stronger hot spotting - shooting angles near the sun, specular highlights or high contrast in the image, and so on, then you can get remarkable results from even the most limited lens. Use a lens shade whenever possible, and/or shoot from a shaded spot- (Maybe have your assistant carry an umbrella? Maybe not...)

I will mention that my converted point and shoot shows no hot spotting AT ALL. NONE!  (A Canon Powershot A630 converted with a 665nm filter. ) My DSLR, with its larger files, RAW capabilities, and generally sharper captures is also hot spot prone. So far. I have not had the opportunity to try any Canon L series lenses; I suspect there may be considerable differences due to superior optics and coatings and build quality.

I have found each of my Canon lenses on lists of lenses supposedly 'good' for IR and 'not suited' for IR. This suggests that the performance of an average lens for IR photography is somewhat subjective and variable. Test your lens and you will have the knowledge to make the best use of it.
-----------------------------------------------------
               My Tests and Some Results

Technical disclaimer: I am not very scientific, or proficient at engineering, so you may very well want to research and develop your own testing protocol. . .

The tests I performed were somewhat casual but sufficient for my purposes. All I did was set the camera on a tripod on a sunny day, and make a reasonable exposure of an average scene at each aperture. For the zoom lenses I did that at three focal lengths: the widest, the longest and at somewhere near the middle point.

I set up automated processing with a curves adjustment for contrast and a channel swap. No other enhancement. I compared the images in aperture sequence for each lens, and focal length, and rated the degree of hot spot on a 5 point scale, with 1 being no visible aberration and 5 being very defined and saturated. Using this approach I now have a good idea of what the performance of each lens will be.

One thing I'd never realized before was that the hot spot can be wider diameter than the frame, thus creating a subtle vignetting effect. That kind of effect is editable, as I said earlier. Even the best lens showed some of that. Of my six lenses I will be able to get the most use from my EF-S 18-55, which is pretty good from f11 to wide open. I have an old EF 28-80 which, strangely, needs a custom white balance versus my other Canon lenses. It is very sharp and shows no hot spot down to between f11 and 16. Since I will rarely need  telephoto when I shoot IR, these lenses will be fine for walk-around shooting.

I hope this article is useful for anyone new to IR or who has been frustrated by the hot spot problem.

Questions, comments, corrections always welcome.