Chapter 04: Dangerous Half Knowledge

There are a lot of statements sbout lenses and optics, which we are sure that they are true.
All the more more surprising is the length of the following list of statements that are not (completely) true

Half knowledge about light

The speed of light is constant and equal for all light colors

For correctness this sentence is missing the words’ .. in vacuum .. ‘ and the word color should be replaced by ‘wavelength’, because colors do not exist!

Optics the way we know it wouldn’t be possible if the speed of light were constant in all media and for all wavelength. Refraction and reflection rely on different speeds of light.

Light doesn’t change its direction within a medium

This is only valid within media of constant refractive index.
 Light is for example distracted by fast moving water!
There are materials, such as GRIN lenses which have a different focal length, depending on their length.
We all know exceptions like mirage or the smears above hot asphalt in summer:
Hot air has a lower refractive index than cooler air, so the sky reflects on the hot layer directly above the road or  desert.
Hot air just above a lighted candle caused streaks that deflect the light because hotter air has a lower refractive index.
The most common exception (though not “obviously” ;)) is the lens of the eye, the radial direction (= from the edge to center) has a varying refractive index.

Half knowledge about glass

Glass is transparent

Not always right: 
 Position a glass bottle between a telecentric lens and a telecentric light source that are precisely aligned. On the screen, the bottle then appears mostly deep black.
This particular setup allows only axis-parallel light to pass through. Light from the telicentric light source can directly pass if there is no glass in the light path. Glassrefracts light however, so the axis parallel light from the light source change their direction when they hit the bottle. It easy to understand if our eye takes the role of a camera sensor for a moment: A glass bottle distorts what’s behind behind it.

Each type of glass has a refractive index

The word ‘a’ is wrong here. Optical glasses have an individual  refractive index for each wavelength. Refractive indices of different wavelengt (eg, red / blue) have indeed in each individual type of glass a fixed ratio, That ratio is different from glass type to glass type however.

Half knowledge about lenses

Lenses are transparent

Lenses for thermal cameras are for example made of Germanium. Visible light does not pass through these lenses.

Lenses for large sensors can also be used on all smaller sensors.

A customer had received a lens from his colleagues that worked fine on 2/3″ sensors, but on the new 1/2″ sensor it supplies inferior image quality. This is possible if the smaller sensor has smaller pixels than the big ones.

Viewing angle and focal length are equivalent.

That would mean that larger focal length lenses always have a smaller field of view (FOV). But 
 this need not be so!

For example a 2 cm long f = 6mm lens at the same camera position can have a larger (!) viewing area than a 4 cm long f = 5.8mm lens.
Even two equally long lenses with the same angle of view can have different FOV!, Namely when the principal planes (?) Of the lenses have different positions!

Lenses with the same angle, the same length and focal length have the same FOV.

This is not necessarily true, namely, if the lenses have different main levels, the other dimensions may be the same and the lens still sees a different section

The larger the (apparent) “opening” of a lens, the better it collects light.
This only applies for lenses of the same focal length!
With wide angle lenses, the apparent opening is smaller than when using lenses with the same focal length is increased light sensitivity! This means that a lens with a smaller aperture can apparently be quite sensitive to light than one with a seemingly larger.

Lenses of the same focal length have the same viewing angle.

Some manufacturers call a lens with e.g. 8.3 mm focal length “8mm” lens. This allows lenses seemingly “same” focal length definitely have different opening angles. Lenses can also see differently with different distorsion much the same focal length but.

The smaller the focal length, the larger the depth of field.

This is wrong.

If for a camera FOV and the F# are constant, then the  DOF is constant too

The larger the aperture value, the larger the depth of field.

This is wrong, if you compare lenses of different focal lengths.

The DOF of an f = 25mm lens at 3m distance at F2 is about the same DOF of a 100mm lens at an aperture of F32!

It’s not valid also because there is a natural upper limit of the F-number of a lens. This number is called “critical aperture”.
At even smaller apertures, errors from diffraction are greater than the gain through smaller apertures

An objective can never be “too good”.

It is possible that the resolution of a lens is too good.
Then something known as Color moiré occurs.
Problems occur when the lens “sees more” so to speak than the sensor.
To avoid color moiré can be called OLPF (optical long pass filter) used to ensure a “minimum uncertainty”.

For C-mount lenses, the distance between the flange of the lens thread and the sensor (image plane) is 17.526mm

Just valid “in air”.
If there are are glass plates / filters between lens and sensor (most likely to be the case), then the actual distance increases by roughly 1/3 of the glass thickness.

For CS-mount lenses, the distance between the flange of the lens thread and the sensor (image plane) is 12:52mm

Just valid “in air”.
If there are are glass plates / filters between lens and sensor (most likely to be the case), then the actual distance increases by roughly 1/3 of the glass thickness.

The difference between the flange focal distance C-mount and CS-mount lens is 5mm

Only valid if rounded (17.526mm-12, 52mm = 5.006 mm).

Lenses change their focal length depending on the sensor size

Wrong. The lens changes nothing. Is meant that the visible object section depending on the sensor size is different.

Fixed focal length lenses have a constant magnification

Only true for telecentric lenses. Entocentric and pericentric lenses have a different magnification in for each working distance.

see also: 
 magnification, comparison entozentrisch – telecentric – perizentrisch

If object and the sensor have the same aspect ratio (eg 4:3), so has the resulting image on the sensor

This is true for very few lenses:

Situation: Using a 1/3 “sensor of 4.8 x 3.6mm 120x90cm a large poster to be recorded. Amazingly, the object has 4:3 format, the sensor also, but the image is not: right and left you can see too much.
Background of this phenomenon is distorsion. For a lens without distortion, as would also be the image in 4:3 format.

Since lenses in general, have a distorsion, generally one side of the image is longer than expected as distorsion in vertical and horizontal direction generally are different (and thus in the horizontal direction is longer). Since distorsions in the objective there on the market is usually negative, on right and left side of the image there’s more image information than desired.

Near objects have larger images than remote objects

While this is true for entocentric lenses, it is not valid for telecentric or pericentric lenses.

The physical distance between the object and the sensor determines the viewing angle.

Wrong, especially for short working distances: a f = 6mm lens and a (longer) f = 5mm lens can see the same image at different angle at 100mm distance between sensor and object.

Leave a Reply