Category Archives: Lens Types

What happens when you focus lenses from infinity to shorter distances?

To focus on shorter distances, s-mount lenses have to be “unscrewed” . On first glance this seems to be very different from larger Lenses like C- or CS-mount Lenses, that provide a focus ring. If the focusing ring is turned however also in larger lenses the position of a lens package from the sensor is increased. (There are a few exception of lenses that offer more than one lens package that are moved synchrously, one of this maybe towards the sensor) .
Well over 95% of all lenses work as described here:

May the lens be focussed on infinity. What happens when you increase the distance of the lens package from the sensor, for example by unscrewing or by turning the focus ring?

  • The _maximum_ object side viewing angle of the lens stays the same… because the optics does not change
  • The _maximum_ image side viewing angle of the lens stays the same… because the optics does not change
  • The _maximum_ amount of pixels the lens can handle is still the same … because the optics does not change
  • The F# changes to the “Working F#” (also called effective F#) :
    The F# of a lens is only defined for infinite distance. When focused to infinity the focal point of the lens is right on the sensor.The F# is then defined as : Focal length, divided by the Entry-Pupil-Diameter (= the appearant diameter of the “hole”  in the lens when you look from object side = “EPD”)The working F# (‘wF#) is defined a little different : wF# = (focal length + amount unscrewed) /EPD = F# + (amount unscrewed / EPD)
In the general case, that the Magnification is M ( where M = sensor size / object size) we get a formula :

    \[wF\# = (1 + M) * F\#\]

Example 1:
For a 1:1 magnification (object size = sensor size), you have to unscrew by the focal length. 12mm for a 12mm lens : 4mm for a f=4mm lens , 50mm for a 50mm lens)
For the 1:1 case we get

    \[wF\# = 2* F\#\]

(because the amount unscrewed = f , so we get

    \[wF\# = 2 * (focal length / EPD ) = 2*F\# \]

We get the same result from the above formula with M=1 :

    \[wF\# = 2 F\#\]

Example 2:
For the object at infinity, no unscrewing is needed, so we have wF# = F#.
The magnification for objects at infinity is zero, because the sensor is small and at infinite distances the lens sees infinite much. (Hundreds of galaxies at the night sky, for example). So we get from the formula above :

    \[wF\# = (1 + M) F\# = (1 + 0)F\# = F\#\]

  • The brightness of the image changes:
    The amount of light that reaches the sensor is determined by wF#, the working F#. wF# depends on the diameter of the Entry pupil, but the brightness depends on the area of the entry pupil.
When the wF# is increased by factor x, the image brightnness is decreased by factor x^2
Brightness of a standard 1:1 lens:
wF#, the working F# of a 1:1 lens is
wF# = 2 F# = twice the wF# at infinity \]
So the brightness decreases by factor 2^2 = 4 compared to the brightness at infinity
  • In general the resolution decreases: The smallest possible point diameter that a diffraction limited (read”perfect”) lens can generate is given by the Rayleigh diameter :

        \[ D = 2 * 1.22 * wF\# * Wavelength \]

    The resolution is half that diameter R = D/2.

The resolution of a (non telecentric) 1:1 lens is about half the resolution of the lens in infinity position , both in x and y direction. If the lens could resolve 5 Megapixel at infinity before, the resolution drops to 1.3 Megapixel if it is used in an 1:1 setup.
    • The Field of view gets smaller: Because the lens is not telecentric (but “entocentric”), the light arrives at some Angle > 0 in the corners of the sensor. We can imagine this as an image side (half) viewing angle. That angle is called (max) Chief ray angle) When the lens is in infinity position. When we increase the distance to the sensor, tha maximum angle stays the same, but some of the light will no longer reach the sensor. This means, only a smaller fan angle on image side can be used. This implicated that also only a smaller angle on object side can be used!
    • The magnification changes. This is because the sensor keeps its size and the visible Object size gets smaller, see above
    • The distortion in general gets better : Distortion of each lens is larger in the corners of the field of view than in the center. Because we use a smaller object and image side angle now, we also don’t use the old colrners of the image any more. Therefore we don’t use the rim of the lens elements
    • There working distance changes, because the ratio of object distance and image distance is the Magnification, which changed.
    • The Chief Ray Angle CRA changes. This it the off axix angle at which the light arrives in the sensor corners.
Sensors with shifted microlenses can have troubles with short workingdistances if they need a minimum CRA. Microlens-Vignetting would be the result
Lenses that were designed for infinity, assume that the light arrives about parallel at the lens. This clearly defines angles at with the light arrives at the sensor surface.
By changing the working distance from infinity to a shorter distance, maybe even below the MOD, these angles change. But for these new angles the lens was never designed, so the performance MUST suffer.
Whether the performance is still good enough depend on you application and especially on the sensor pixel size
Telecentric lenses behave different: The image side F# is Magnification times the object side F#.
For an 1:1 lens for example is the imageF# = object side F#, the resolution on image and object side is the same (and not factor 2 lower than with entocentric lenses. Also the image brightness does not decrease by factor 4.

Little siblings: s-Mount lenses as replacement for c-mount lenses

If we want to replace c-mount lenses by s-mount lenses it begs the question …
Can s-mount Lenses replace c-mount lenses in general?

Of course lenses should be of similar quality …
Can s-mount Lenses be as good as c-mount Lenses?

To answer this question, we should define what’s “good”. At least we should get an idea about …

How to improve Lenses …

In general we can assume that objectives are optimized in a way. Why is that so? Let’s assume an objective has, say, 6 lens elements. We can assume that all of these lens elements are needed, to achieve the quality of the lens. Not needed, the manufacturer would sure work with fewer elements, as this would increase his earnings. Maybe there is a better design (from a better designer) with fewer lens elements, but we can assume that the original designer did the best he can …

How to improve the lens then? Improving a lens implies to change the directions the light beams travel. If there’s no need to change the directions, there’s no need to change the lens at all;)
Light travels straight (in optical homogeneous media) and changes directions only at the interface between air and glass or at the interface between two types of glass.
To change lights directions additional lens elements will be needed (for the assumed optimized lens). More glass means higher cost and a larger footprint of the lens (let alone a redesign of all other lenses in the system … you in general can’t just change ONE lens element but have to work holistic. A new lens element or a reshaped lens element influences all other lenses in the system.

As production costs are limited by constraints of the target price and the mechanics, there definitely are limits for the s-mount lenses and markets that will be c-mount markets for a long time or forever.

A general replacement is c-mount by s-mount is not possible for the same reasoning, however in my personal opinion 50% or more of the c-mount lenses can be replaced by s-mount lenses.

For the pros and cons of c-mount and s-mount lenses please also check …
Comparison: c-mount lenses vs. s-mount lenses (M12x0.5)

Comparison: c-mount lenses vs. s-mount lenses (M12x0.5)

Featurec-mounts-mount (M12x0.5)
Thread1″ 32 TPIM12x0.5
manual iris+generally not
IR-Cut filterin camerain general possible in the lens
Special filtersfront filters by thread in general, in the lensin the lens
Sensors1/10″ … 4/3″1/4″ 1/3″ (most used) 1/1.8″ (2/3″ very rare)
Filter changercamera sidecamera side or in the lens
F-NumbersF# 0.7 .. F#360F0.95, F1.0, F1.2 ..
focal length1.7mm .. 2000mm0.98mm .. 50mm  (others on request)
Total Track(27) 55mm .. 3000mm9.5mm .. 25mm
Availability Varios+++
Availability Zoom(very expensive)
Availability Fisheyes+ (expensive)+ (low price)
5 Mega+(expensive)+(low price)
Availability 10 Mega+(very expensive)+ (few types, but good price)
low distortionexpensivelow price
Irisadjustable– not adjustable + doesn’t move
mountingscrew in as far as possibleo Lockring recommended
focussingInner focussing by focus ringouter focussing using thread
Delivery times+ small volume, – large volume+ few + large amounts
custom design generally not possible (but ask us! )++ (could make sense for 50+ )
consumer market++
Distance Sensor-Lensgenerally 6-10mm0.5 (!) – 20mm
use in OEM projectsless and lessmore and more
use in handhelds++
use in mobile phones++
famous brands+
text on the lens+o in general not (on demand)
Optomechanical qualityin general pretty gooddepends on origin –(toy quality) to ++ (OEM)
Availability Telezentric++o few available (ask us 🙂 )
Availability Macro(expensive)o (using toolbox) low cost
Availiability Micro(expensive)
Availability scientific(very expensive )