Nick

A couple weeks ago I was doing some testing to verify the infrared focus point offset on some of my lenses. For the unfamiliar, infrared light generally focuses to a different point than visible light. For this reason, it was actually common practice for a long time to include an IR index on lenses, usually a red mark of some kind on the focusing scale. After focusing the camera, you would then adjust the focus based on that mark. (The underlying phenomenon of spectra having different focal points is also related to the phenomenon of chromatic aberration.)

Lamentably, current camera systems not only lack IR indices, they mostly lack focusing scales entirely, because everyone is autofocus-pilled. Cameras do generally have some kind of electronic focus scale display option, but these are rarely designed in such a way as to be actually informative, even for visible light shooting.

Anywho, one thing I noticed in testing is that the Canon RF28mm pancake lens, while tragically unsuitable to shooting true infrared due to a hotspot that kicks in around f/4, has barely any gap between its IR focus point and its visible focus point.

This made me wonder if it might be a particularly good choice for use with the IR Chrome filter, since that filter is combining infrared and visible light in the same exposure. I've noticed in the past that IR chrome images could sometimes have a kind of glitchy effect that was somewhat similar to red/blue 3D viewing or trichrome photos with imperfect registration, where it seems like there is some offset between the channels. Using a lens that focuses the IR to almost the same point as the visible might help alleviate that.

I did some very basic testing with some of the lenses I have available that have filter sizes that I can adapt to my IR Chrome filter: the 28mm pancake, the classic 2.8cm f/3.5 Nikkor, one of the great manual focus lenses for black and white infrared, and the adorable RF16mm. I should have brought the RF 50mm, but I forgot, and I didn't use a tripod because I couldn't find any of my QR plates for some reason. So, I don't think this necessarily does much to test my actual hypothesis, but goddamn, the corner sharpness on the 28mm pancake is wild relative to what I"m used to. It truly is a shame about the lens's hotspots in black and white IR.

For what it's worth, these are 100% crops with a basic preset applied for color and contrast, but no adjustments to sharpness.

For those who are curious about those IR offsets, see below. Unfortunately the specified focal points aren't the same across the lenses because I was working with the cameras' displayed distance scale values.

RF28mm f/2.8:

Setting	Focused Distance
1.5'	19"
3'	37"

RF16mm f/2.8:

Setting	Focused Distance
1'	14"
2'	2.67'
2' plus 1-2 ticks	4'
2' plus 4-5 ticks	6'

Nikkor 2.8cm f/3.5:

Setting	Focused Distance
3'	3'4"
4'	5'10"

RF 15-30:

Length	Setting	Focused Distance
30mm	2'	26"
30mm	3'	about 3.5'
24mm	2'	28"
24mm	3'	Not quite 4'
20mm	2'	29"
20mm	3'	About 4', little more?
15mm	2'	Maybe 3.5'
15mm	3'	??

IR-converted GRIII:

Setting	Focused Distance (metric)	Focused Distance (imperial)
0.3m	0.46	18"
Halfway btw 0.3 and 1	0.99	3.24'
1m	1.83m	6'

it does sort of demonstrate what I'm talking about with IR Chrome sometimes giving an appearance that is vaguely reminiscent of a trichrome with slightly off registration. Top is with the 2.8cm f/3.5 Nikkor-H, bottom is with the Canon 28mm pancake. The Canon one isn't necessarily a better photo, mind you, but certainly much cleaner. (Which is only to be expected, there's something like 60 years between the two lenses.)

It's unfortunate that I can't do a direct comparison with another modern 28 without buying something. I have a 15-30 zoom but it takes too large a filter size for me to use a step down.

A couple weeks ago I was doing some testing to verify the infrared focus point offset on some of my lenses. For the unfamiliar, infrared light generally focuses to a different point than visible light. For this reason, it was actually common practice for a long time to include an IR index on lenses, usually a red mark of some kind on the focusing scale. After focusing the camera, you would then adjust the focus based on that mark. (The underlying phenomenon of spectra having different focal points is also related to the phenomenon of chromatic aberration.)

Lamentably, current camera systems not only lack IR indices, they mostly lack focusing scales entirely, because everyone is autofocus-pilled. Cameras do generally have some kind of electronic focus scale display option, but these are rarely designed in such a way as to be actually informative, even for visible light shooting.

Anywho, one thing I noticed in testing is that the Canon RF28mm pancake lens, while tragically unsuitable to shooting true infrared due to a hotspot that kicks in around f/4, has barely any gap between its IR focus point and its visible focus point.

This made me wonder if it might be a particularly good choice for use with the IR Chrome filter, since that filter is combining infrared and visible light in the same exposure. I've noticed in the past that IR chrome images could sometimes have a kind of glitchy effect that was somewhat similar to red/blue 3D viewing or trichrome photos with imperfect registration, where it seems like there is some offset between the channels. Using a lens that focuses the IR to almost the same point as the visible might help alleviate that.

I did some very basic testing with some of the lenses I have available that have filter sizes that I can adapt to my IR Chrome filter: the 28mm pancake, the classic 2.8cm f/3.5 Nikkor, one of the great manual focus lenses for black and white infrared, and the adorable RF16mm. I should have brought the RF 50mm, but I forgot, and I didn't use a tripod because I couldn't find any of my QR plates for some reason. So, I don't think this necessarily does much to test my actual hypothesis, but goddamn, the corner sharpness on the 28mm pancake is wild relative to what I"m used to. It truly is a shame about the lens's hotspots in black and white IR.

For what it's worth, these are 100% crops with a basic preset applied for color and contrast, but no adjustments to sharpness.

For those who are curious about those IR offsets, see below. Unfortunately the specified focal points aren't the same across the lenses because I was working with the cameras' displayed distance scale values.

RF28mm f/2.8:

Setting	Focused Distance
1.5'	19"
3'	37"

RF16mm f/2.8:

Setting	Focused Distance
1'	14"
2'	2.67'
2' plus 1-2 ticks	4'
2' plus 4-5 ticks	6'

Nikkor 2.8cm f/3.5:

Setting	Focused Distance
3'	3'4"
4'	5'10"

RF 15-30:

Length	Setting	Focused Distance
30mm	2'	26"
30mm	3'	about 3.5'
24mm	2'	28"
24mm	3'	Not quite 4'
20mm	2'	29"
20mm	3'	About 4', little more?
15mm	2'	Maybe 3.5'
15mm	3'	??

IR-converted GRIII:

Setting	Focused Distance (metric)	Focused Distance (imperial)
0.3m	0.46	18"
Halfway btw 0.3 and 1	0.99	3.24'
1m	1.83m	6'