Analog (optical) filters

There's a story in

He looks quite human to me, even though he did his own thinking. What did he have to say about multilayer optical filters?

Jeroen Belleman

As I recall, someone suggested something to him in a hallway encounter. A few days later he delivered a complete theory of multilayer optical filters.

He was far more intelligent than humans can be. His hobby, as a kid, was seeking out difficult differential equation to solve.

At 13.5 nm, nothing refracts and nothing reflects... except extreme multilayer coatings. That makes for interesting mask reducing optics.

Multi-layer optical filters obey the same laws as consecutive pieces of coax alternating between two impedances, except that the working frequency is much lower, and the coax is just a single pixel.

Jeroen Belleman

And optical coatings have to work at different angles of incidence and two polarization states.

Cheers

Phil Hobbs

Analogue traditional dye and gelatine optical filters are still available to professional photographers today. eg

A simple fader is usefull in getting relative lux/power measurements for emitters, while keeping detector in linear range.

RL

Is there any reason to use a filter in digital photography? Imager chip pixels have extreme dynamic range and one can digitally post-process an image.

Yes. The most common reason would be to photograph in some narrow wavelength band to isolate particular elements. Scientific imaging tends to be done with monochrome CCDs and very well characterised glass filters in astronomy and fluorescence microscopy for example.

You can get a roughly plausible colour image from Halpha to R and OIII

Single shot colour imaging is popular with hobby astronomers, but even there many push the technique using narrowband filters (which helps cut out a lot of the light pollution).

In discussions about lights near observatories they offered to use yellow LEDs as an alternative to LPS - missing the point that it wasn't the yellow light so much as the very narrow waveband it pollutes (and sodium is so ubiquitous there is always some sky glow at sodium D lines.

I was thinking about the "traditional dye and gelatine optical filters" used by "professional photographers", not high-Q optical filters used by scientists. Wedding pictures and stock images can be post-processed for color and effects.

$40 for the "Fog" filter in 1985 dollars! Highway robbery!

Sure, many reasons, some of which are similar to why you use analog filters in electronics.

-- Eliminating large unwanted signals (e.g. the pump light in a "black light" image)

-- Glint elimination using linear or circular polarizers

-- Dynamic range reduction by emphasizing dimmer spectral regions.

"Extreme dynamic range", e.g. 50k electrons full-well. That's a shot noise limited SNR of 47 dB--enough for a good image, certainly, but not for a good image plus a lot of reserve for high contrast situations.

Postprocessing is fine, but it won't fix inadequate SNR or dynamic range.

Cheers

Phil Hobbs

Commercial photographers still use ND3 to ND5 filters even today to get the motion blur effects on waterfalls and sea scapes. It could be done by taking loads of shots and combining them in post processing.

There are some effects that are easier to do with a filter in front of the lens especially if they are using large cameras and sheet film.

The other one is graduated density filters for managing bright skies. You can do it with digital HDR but the image can look very artificial.