I have to test a camera that is focused to infinity, but I cannot take an image of an object that is more then 1.5 meters away from the camera because I can't get it out of the room itself. (The room is a clean room, used to store the hardware in order for it to still be clean when we have to use it eventually.)

Is there a way to create some sort of a distance illusion? Or a any way to test the focus?

  • \$\begingroup\$ What kind of camera and lens? For a very wide angle lens, or for a small enough system, 1.5m could effectively be infinity (e.g. my 16-35mm lens focus distance markings only go up to 1m before infinity, so that's close-ish). What "effective infinity" means is highly lens-dependent, and somewhat related to hyperfocal distance... \$\endgroup\$
    – twalberg
    Jul 27, 2018 at 13:11
  • \$\begingroup\$ I haven't gotten the camera yet so the hyper focal distance is still unknown to me, but if I get measurments of the lens I will be able to know the approximate length in which I can test it ? \$\endgroup\$
    – Xalien
    Jul 27, 2018 at 13:16
  • \$\begingroup\$ Yes - there are numerous web-based calculators and Android/iPhone apps out there to assist with this - plug in your camera's sensor size, and the focal length of the lens, and any other required info (distance to subject, f-number, etc.) and it'll calculate things for you. \$\endgroup\$
    – twalberg
    Jul 27, 2018 at 13:21
  • \$\begingroup\$ thank you ! can you please write it as an answer so I can accept it and the question can be solved ? \$\endgroup\$
    – Xalien
    Jul 27, 2018 at 13:22
  • 3
    \$\begingroup\$ I'm voting to close this question as off-topic because it indicates no photographic purpose. \$\endgroup\$
    – Michael C
    Jul 27, 2018 at 23:10

2 Answers 2


This sort of testing is normally performed with an infinite conjugate optical system. One example of such is a/an (auto)collimator.

Collimators and AutoCollimators can be used interchangably for the task of testing a lens at infinity. The testing procedure is simple.

  1. Obtain an (auto)collimator which is focused to infinity and contains a resolving power target
  2. Focus the camera on the "infinite" distance image from the collimator
  3. Read the resolving power indicated by the target

This system has the advantage of not only measuring the resolving power at infinity but also allowing you to focus the camera with the aid of the collimator. To use the collimator to assist focus, simply perform a binary or progressive search while focusing. The peak resolution represents infinite focus.

Autocollimators may also offer the ability to focus at distant, far conjugate distances. For example, with a 300mm lens the hyperfocal distance could be thousands of feet. A properly configured autocollimator could be focused to, a specific distance to allow you to focus.

Collimators can be found on ebay for a couple hundred and, if not damaged, tend to last forever. Edmund Optics and Thor Labs carry collimators in a $1k range. Newport is a mid-range supplier. The Cadillac of production collimators is the Moller Wedel but expect to pay for that quality.

A point source system projects a collimated beam of light into your camera. By measuring the diffraction via a PSF measurement of the image produced by the camera, an MTF value can be generated. Systems which use point sources generally include a goniometer for off-axis testing but that is not strictly necessary.

It is possible to construct your own Point Source using a normal bulb (LED preferred) and the collimating eyepiece of a telescope. The precision of your device will be a product of your facilities for alignment and calibration of the focus rig. PSF/MTF measurement is fairly easy to implement if you are computer savvy and have access to the raw images. OpenCV and various Python distros include libraries to perform such a task.

Cost for such a system could be as little as $50 (the cost of the eyepiece) assuming you've already got a suitable hardware rig and computer. On the other hand, point source focus testing is generally considered the most precise form of optical system testing (most metrology utilizes Point sources) so there are a variety of commercial systems. The cheapest is probably about $5k and prices range up to $1million including a truly impressive systems from Optikos and Trioptics.

Optical Mirrors are the simplest solution. If the size of your cleanroom represents a large fraction (say, 1/2 or 1/4) of your hyperfocal distance, consider using mirrors. 1 mirror will allow you to double the distance to your target and you can keep adding mirrors. Use high quality mirrors such as first surface optical mirrors. I recommend enhanced aluminum for visible applications due to its superior performance in the blue region. For many folds (greater than 3) consider upping the quality of the mirrors to 7/lambda or 10/lambda.

The longer the focal length of the lens, the smaller the mirror needed but the higher the quality needs to be in order to match the increased angular resolving power. To determine the size of optical mirror needed, image your target through a handheld or wall mirror. Mark the edges of your resolving power target with tape using a live feed from the camera to judge the limits of cropping. Then you can measure the taped off area.

Optical mirrors can be had for a couple of dollars but I suggest going for a reasonable level of quality in anything more than 1 fold. As such, expect to pay $30-$90 per mirror. Edmund Optics or Thor labs are good sources for new mirrors. Avoid purchasing new mirrors from ebay or amazon where counterfeits are common. Used scientific mirrors, however, tend to be of high quality and it's easy to see if something is wrong with the mirror in an image of it.


Infinity (Latin as far as the eye can see) optically is when light arrives from an object as parallel rays. As a rule of thumb, this distance will be about 3000 times the diameter of the entrance pupil. For all intent and purposes, the entrance pupil is the diameter of the iris diaphragm (aperture). You can measure the diameter of the entrance pupil by shining a flashlight with a spotlight beam, into the lens from the rear. Place a piece of white paper before the lens. Measure the diameter of the projected circle that plays on the paper. Also you can calculate the diameter of the entrance pupil. Divide the focal length by the f-number setting. Thus a 20mm lens operating at f/8 has a 2.5mm diameter entrance pupil. Infinity for this lash-up is 2.5 X 3000 = 7,500mm = 7.5 meters = 295 inches = 25 feet.

If none of this works for you, procure a small telescope or binocular. Focus them so that an object is sharp to your eye. The nature of the optics causes the light to exit as parallel rays. Allows your camera to peer through the telescope. Place the camera so that the camera lens is almost touching the eyepiece of the scope. The camera lens set to infinity; this lash-up simulates infinity.

  • \$\begingroup\$ Interesting rule of thumb I may use that in the future.. Is that assuming full frame / film-sized CoC's? Using your approximation I get CoC's of 25, 18, and 11 micron for 80, 50, and 35mm lenses, respectively. Seems OK for a full-frame but not quite enough for many of the smaller sensors out there. \$\endgroup\$ Jul 27, 2018 at 20:12
  • \$\begingroup\$ @ PhotoScientist -- 3000 diameters distance from a disk results in that object under the most favorable conditions -- usually to stringent for photo applications. This is equal to a 3 foot wagon viewed from 1.3 miles. Depth of field is generally based on 1/1000 of the viewing distance = 3.4 minutes of arc equivalent to 1/100 of and inch viewed from 20 inches. \$\endgroup\$ Jul 27, 2018 at 20:25

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.