Coated vs uncoated lens?

[petrochemist]

A single surface or density change will reflect 4% of the light. But multiple surfaces are NOT a cumulative total. Multiple surfaces will reflect between 0-16%. I know this sounds odd and counter intuitive but the maths that predicts it was worked out by Richard Feynman and it's been tested and found to be correct, in fact modern lens coatings are designed with it. The total of the reflected light through modern lenses can be less than 4%, and in uncoated lenses it can be as much as 16% (not including scratches ).

The reflection at a surface depends on the change in refractive index of the two mediums. Anti reflective coatings are made with a carefully tuned refractive index (between that of the the materials either side) and with precisely controlled thickness. This ensures the light reflected from the front surface of the coating destructively interferes with light reflected from the rear surface cancelling out both reflections. Modern multi coatings use a series of layers so that they can be fine tuned to the different wavelengths that might be significant. If you try infra red photography you will find modern coatings generally do very little to reduce reflections of infra red wavelengths.

 

[Tim Tucker]

A single surface or density change will reflect 4% of the light. But multiple surfaces are NOT a cumulative total. Multiple surfaces will reflect between 0-16%. I know this sounds odd and counter intuitive but the maths that predicts it was worked out by Richard Feynman and it's been tested and found to be correct, in fact modern lens coatings are designed with it. The total of the reflected light through modern lenses can be less than 4%, and in uncoated lenses it can be as much as 16% (not including scratches ).

The reflection at a surface depends on the change in refractive index of the two mediums. Anti reflective coatings are made with a carefully tuned refractive index (between that of the the materials either side) and with precisely controlled thickness. This ensures the light reflected from the front surface of the coating destructively interferes with light reflected from the rear surface cancelling out both reflections. Modern multi coatings use a series of layers so that they can be fine tuned to the different wavelengths that might be significant. If you try infra red photography you will find modern coatings generally do very little to reduce reflections of infra red wavelengths.

Exactly, well that's how the maths works. For light hitting a single surface (of glass) straight on the reflection is always the same, around 4%. Even with uncoated glass the reflection off the rear effectively "interferes" with the reflection off the front, (it is not actually the surface of the glass but the glass itself and nobody can really explain why). When light hits the second glass lens surface this does not "start again" and become a cumulative total but "interferes with the light reflecting off the first lens surface".

It's worth clarifying this as there's a lot of internet nonsense about modern multi-coated multi-element zoom lenses. I entirely agree with your observations on IR.

 

[petrochemist]

The reflections from both surfaces of uncoated glass will indeed interfere, but it won't be purely destructive interference. In some places the amplitudes will add to increase intensity.

I have studied the depth of penetration on total internal reflection somewhat for spectroscopy use. I don't remember the formula but I remember it related to both wavelength and refractive index (significantly complicating quantitative work using ATR rather than transmission cells) I suspect the physics of it will be similar for other surface reflections at less than the critical angle. I'm glad to hear I'm not alone in not understanding why it happens!

 

[jvansanten]

The reflections from both surfaces of uncoated glass will indeed interfere, but it won't be purely destructive interference. In some places the amplitudes will add to increase intensity.

I have studied the depth of penetration on total internal reflection somewhat for spectroscopy use. I don't remember the formula but I remember it related to both wavelength and refractive index (significantly complicating quantitative work using ATR rather than transmission cells) I suspect the physics of it will be similar for other surface reflections at less than the critical angle. I'm glad to hear I'm not alone in not understanding why it happens!

 

[IanG]

The reflections from both surfaces of uncoated glass will indeed interfere, but it won't be purely destructive interference. In some places the amplitudes will add to increase intensity.

I have studied the depth of penetration on total internal reflection somewhat for spectroscopy use. I don't remember the formula but I remember it related to both wavelength and refractive index (significantly complicating quantitative work using ATR rather than transmission cells) I suspect the physics of it will be similar for other surface reflections at less than the critical angle. I'm glad to hear I'm not alone in not understanding why it happens!

As someone else revived this thread

About 2 or 3 years ago I tested quite a few uncoated lenses, they results were quite predictable.

I started with a 1913 110mm f6.8 Dagor (CP Goerz Berlin) this was the highest contrast close to a modern MC lens as with only 2 internal air glass surfaces, next was a 165mm f5.3 CZJ Tessar an uncommon aperture, with 4 internal air glass surfaces there was a noticeable loss of contrast, next was an Ihagee-Goerz 135mm f6.8 dialyte lens with 6 internal air glass surfaces there was a significant drop in contrast making it a challenging lens to use. Finally a Meyer Weitwinkler WA lens a double Gauss design like a dialyte 6 internal air glass surfaces this had the lowest contrast of all. I have examples but I'm not in the UK and don't have access to them at the moment.

For my main work I only use coated lenses. they include a two Angulons a 65mm & 120mm, a few Super Angulons 65mm, 75mm (f8 & f5.6), 90mm f5.6, and a 165mm f8, plus a CZJ 150mm T coated Tessar, and some Xenars plus a Tele-Xenar. There's no difference between these and my Multi Coated LF lenses.

Ian