IR distance adjustment formula

[xkaes]

I've SEARCHED all over but all I can see are the typical "You can't there from here" responses. I'm trying to find a calculator / formula for computing the adjusted distance for IR B&W film from the visible light distance.

I have software that can compute the formula based on the focal lengths of the lenses I have, and spend a lot of time doing it, but I thought someone would know.

Here is a sample of lenses (on the left in millimeters) and the IR focus distance (on the right in feet) when the subject is at infinity. For example, a 80mm lens needs to be focused at about 50 feet, while a 17mm lens needs to be focused around 5 feet.

These are just rough estimates based on the scales on my lenses.

Like I said, there must be a SIMPLE formula -- or I can create one.

I hope not to open up a can of worms, but we'll see.

 

[ntenny]

What wavelength are you assuming for “IR”? It covers a lot of ground.

I did some experiments with Efke IR (nominally sensitive out to 820 nm), photographing a measuring tape, and didn’t find a big focus shift in practice. It was a sloppy experiment and others may have tested more precisely.

-NT

 

[Sirius Glass]

Fortunately for me, my lenses have the IR focus mark engraved on the lenses and they are near the infinity but offset a bit closer.

 

[MattKing]

Any lens with an IR focus mark will likely be set up with Kodak HIE in mind. That means film with with its peak sensitivity between 750nm and 840nm, and some sensitivity extending to approximately 900nm.
If the film is not sensitive as far into the IR as HIE was - and no current films are - then the IR mark will over-compensate for the difference.

 

[retina_restoration]

Any lens with an IR focus mark will likely be set up with Kodak HIE in mind. That means film with with its peak sensitivity between 750nm and 840nm, and some sensitivity extending to approximately 900nm.
If the film is not sensitive as far into the IR as HIE was - and no current films are - then the IR mark will over-compensate for the difference.

This is true. The so-called IR films we have access to now do not need that focus adjustment, and in fact - it will potentially lead to focus errors if applied to these new films.

 

[xkaes]

Not every lens in the world has an IR mark, and many of them don't even have a focusing ring.

And I'm referring to B&W IR films that are currently available. That's a pretty narrow spectrum.

And while longer lenses don't need much adjustment at all, as my table shows, shorter lenses (which are NOT wide-angle on smaller formats) require significant adjustment. A 17mm lens needs to be focused at around five feet when you want the horizon in focus.

 

[reddesert]

There is no simple general formula. It depends on the lens design, the types of glass used, and so on. Different lenses will have somewhat different effective focal length variation as a function of wavelength.

There is a general rule of thumb, which you can find in older books about IR photography. That rule is: add an extension of 1/400 of the focal length of the lens. So for a 50mm lens, you would turn the focus ring so as to extend the lens 1/8 mm (0.125 mm) further from the film. For a typical 50mm lens on 35mm film, that corresponds to about the 20 meter focus mark or the f/4 DOF mark. I've compared this formula to the manufacturer's IR marks on various lenses and it's generally within a factor of 2 or so.

This formula dates from the Kodak HIE days, and the correction would be smaller for currently available films.

The reason your table of focal length versus focus mark shows longer focus-adjusted distances (less offset from infinity) for longer focal length lenses, is that the longer focal length lenses need more extension to focus at a given distance.

 

[xkaes]

Thanks, but as I mentioned, I've seen all that, and it's of little help since most of us don't/can't measure how far out the lens moves forward from infinity -- especially in fractions of millimeters. Turning a focusing ring to a closer distance would work, however. That's what my quick & dirty table does.

And as I said, I can always compute the formula for curves for a bunch of focal length lenses at various distances with SPSS. I just didn't want to re-invent the wheel, as I too often have to do.

But I'm not going to be shooting my new 16mm IR film with my 20mm & 22mm lenses anytime soon. I have to run some exposure and development tests first, and distance is not a factor.

 

[reddesert]

It is easy to compute the conversion from the rule of thumb
extension = (focal length)/400
to where the IR mark should be on the focus scale of the lens, using the simple lens equation:
1/focal_length = 1/d_image + 1/d_subject

d_image = focal_length + extension
So you just need to combine these equations and compute d_subject for a given focal length. That will tell you the rule-of-thumb location of the IR mark for a given focal length. That's the formula you asked for in the original post.

For example, for an fl = 35mm lens, working through the equations gives extension 0.0875mm, and that suggests that the IR infinity focus is at the location of the optical focus at d_subject ~= 14000mm, 14 meters, 46 feet. This is somewhat different than the number in your table, because every lens design is different; if you compare several different 35mm lenses, you'll likely find several slightly different locations of the IR mark.

 

[Klaus_H]

In 2008 I did ask Zeiss about the IR marks on Contax and Hasselblad lenses.
Their answer:

The amount of focus correction to compensate for the longitudinal color error as a function of the wavelength on a lens depends on various factors, e.g. the spectral sensitivity of the sensor or film, the spectral distribution of the light source, the reflectivity of the subject, the transmission of a filter used and the focal length. An infrared index on a lens can therefore only ever provide an approximate recommendation as to the direction in which focus correction should take place when taking pictures with an increased IR component.
The IR index can only be used as a focus index when working exclusively with monochromatic light with a wavelength of approx. 820 nm (i.e. all visible light is blocked out, e.g. by using an interference filter).
If the visible wavelength range is to be included in the image, the extent to which a focus correction in the direction of the index marking on the lens must take place depends heavily on the intensity distribution of the individual wavelengths.
Unless all of the above factors can be precisely determined during an exposure, it is not possible to specify a specific correction factor for focusing. The infrared index on lenses that are primarily intended for general photography with visible light should therefore always be seen as a rough guide for a recommended correction direction.
We therefore recommend stopping down by several stops if possible and, if necessary, creating focus brackets in order to achieve optimum results.

 

[ic-racer]

Plot 'indicated distance' vs 'actual distance' you measure from each lens with your intended film. From there you can get the equation for each of your lenses.

 

[Petrochemist]

Lenses differ in how they behave to IR.
My lenses that have an IR focusing mark, generally shift the lens to focus closer, but not all do, and the distance the mark is from the visible mark is not simply a matter of focal length - it will depend on the types of glass used in the lens...

If you find a graph of wavelength to focus distance achromatic lenses have a quadratic (U shaped) line while apochromatic lenses will have a cubic (vaguely S shaped). Lens designers generally try to keep the visual wavelengths as near to flat on this graph as possible, but most of the time they don't really care about the portions outside the visible - where the slope could rapidly become very steep.

I think there is a simpler solution than a formula, if your lens doesn't have an IR focusing mark you can simply add one.

This is much easier to do with an IR sensitive digital camera than an analogue one, but it is still possible - use a set of markers at known distances, focus visually at a known distance & take a shot in IR. Then process the photo & find out the distance the was actually focused. Mark the degree of shift as a new focus mark on your lens.

If this procedure is too much of a pain simply guess
Three of the four lenses I have to hand that have IR marks (all legacy achromatic primes) have these marks close to where the infinity mark is when the lens is focused to 30m. This might prove close enough for moderate focal lengths & relatively simple designs. The exception is a 400mm lens where the mark corresponds to infinity when focused at 100m.

 

[xkaes]

I'm not looking to compute the additional amount of extension -- although that might be useful for some people, such as those with flexible bellows cameras. I'm looking to convert subject physical distance in focusing distance. For example, in my table, with a 17mm lens, and a subject at infinity, the lens must be focused at 5 feet. I was hoping that someone had computed a formula to convert subject distance into IR focus distance by focal length of the lens.

Looks like I'll just have to do it myself!!!

K.I.S.S.

 

[Sirius Glass]

Any lens with an IR focus mark will likely be set up with Kodak HIE in mind. That means film with with its peak sensitivity between 750nm and 840nm, and some sensitivity extending to approximately 900nm.
If the film is not sensitive as far into the IR as HIE was - and no current films are - then the IR mark will over-compensate for the difference.

This is true. The so-called IR films we have access to now do not need that focus adjustment, and in fact - it will potentially lead to focus errors if applied to these new films.

I have one roll of HIE 35mm film left. When I use Rollei IR 400 I use small lens openings and do not adjust the focus for IR film.

 

[reddesert]

I'm not looking to compute the additional amount of extension -- although that might be useful for some people, such as those with flexible bellows cameras. I'm looking to convert subject physical distance in focusing distance. For example, in my table, with a 17mm lens, and a subject at infinity, the lens must be focused at 5 feet. I was hoping that someone had computed a formula to convert subject distance into IR focus distance by focal length of the lens.

Looks like I'll just have to do it myself!!!

K.I.S.S.

I explained how to do this in post #9. There is no general formula that applies to all lenses (which is one reason that Zeiss was refusing to give a useful answer). So you have to take the rule of thumb for extension, and convert it into a rule of thumb for focusing distance. You can convert this with some algebra, or just use a computer to do the math and make a table of equivalent-subject-distance for your desired focal lengths.

 

[Andrew O'Neill]

The only IR film I ever had to focus compensate for was HIE, with the 87C filter). I didn't bother to do so with the #25. HIE was a real IR film that went out beyond 900nm. Today's IR films go no where near there, and do not need any focussing compensation.

 

[xkaes]

The only IR film I ever had to focus compensate for was HIE, with the 87C filter). I didn't bother to do so with the #25. HIE was a real IR film that went out beyond 900nm. Today's IR films go no where near there, and do not need any focussing compensation.

As the focal length gets shorter, the need to compensate increases -- see my table. Since I'm going to be using very short lenses -- some as fast as 2.0 -- compensation is likely to be needed.

 

[Andrew O'Neill]

I often use short lenses with Rollei IR. I've never focus corrected. f/11-f/22 was usually where I exposed at.
You could do a side by side comparison. One with compensation, one without. Which I have done. Saw no difference. But... maybe there is with lens wide open. If there is, I'll bet it's so small, it goes undetected.
Please report back. I'd be curious with your results.

 

[xkaes]

Since I need to "open-up" about six f-stops with an IF filter, using a fast lens is a big benefit, but results in a big drop in DOF -- not all short lenses are wide-angle.

 

[MattKing]

It wouldn't surprise me to learn that the distance markings on lenses are calculated based on visible light that leans already toward the near IR end of the spectrum. So with modern near IR films ....

 

[xkaes]

The formulas are luckily working out in my favor. I am interested in using IR with two 16mm cameras -- a Feinwerk MEC 16 SB with a 22mm f2.0 lens and a Minolta 16 MG with a 20mm f2.8 lens. The MEC lens focuses and I can make a small IR mark on the camera body for the noticeable adjustment. The MG has a fixed lens at 10 feet, but the camera has a built-in, sliding close-up lens for focusing at 4 feet -- which is almost perfect for IR adjustment.

For focusing lenses that lack an IR mark, it's easy to determine where the mark ought to be.

 

[xkaes]

Just for the heck of it, I've decided to throw some Kodak 4143 High Speed IR film into my Minox IIIs. Using the formula, to focus at infinity, I'll have to set the 15mm f3.5 lens at 2.5 feet. Even though it's a very short focal length lens, it is not a wide-angle lens. And since it is fixed at f3.5, the DOF is relatively narrow, so focusing is critical. Fortunately, focusing very close is no problem with the IIIs.

 

[reddesert]

So to make the rule-of-thumb formula a little more digestible from what I said in post 9 and 15, I realized there is a way to write it using the magnification. The rule of thumb is that the IR extension past infinity e is:
e = f / 400.
Let's write this as e = f * x, with x = 1/400, just in case somebody decides that 200 is better or whatever.
And x is small, x << 1.

The image distance d_i = f + e. And the magnification m is: m = d_i / d_o, ratio of image to object distances. It is possible to then do some algebra to rewrite the simple thin lens formula 1/f = 1/d_i + 1/d_o as:

f = d_i / (1 + m),

so d_i = f * (1 + m)
[this is a pretty common formula in macrophotography]

and so e = m * f. So the factor x that we defined above is actually the same as the magnification m. And d_o = f * (1 + 1/m).

Now, the total distance as marked on the lens scale is basically film to object distance, which is
total dist = d_i + d_o = f * (1/m + 1 + 1)

Remember that m is small, 1/m is large (400 or so), so that extra +2 is insignificant, this is a rule of thumb not a law - the difference between 400 and 402 is immaterial. So the rule of thumb just boils down to the idea that the IR focusing mark should be at a distance of 400 times the focal length.

When I looked at actual lenses, this 400*f is usually within a factor of ~2 of the original IR mark. Again, this was probably for Kodak HIE, and you can expect less of an offset for current extended-red films. For wide angle lenses, DOF will often cover the entire possible range of IR marks if one stops down a little.

 

[xkaes]

For wide angle lenses, DOF will often cover the entire possible range of IR marks if one stops down a little.

That's the kicker. With smaller formats, short focal length lenses are not wide-angle lenses. Take the 15mm f3.5 on a Minox or a 22mm f2.0 on a MEC 16. Plus, to expose IR film it's necessary to open up the lens tremendously if you want to avoid using a tripod -- so the DOF can be very shallow.

I hope to expose my first roll of IR today in my MEC 16. This will be basically to start to determine the correct ISO for the film & processing time, but it will give me an idea of where the infinity focusing point is at f2.0.