scanny77 wrote:
why is it only the interior of the car that is better in the 2nd picture but the rest of the picture is far poorer quality? even the headlights are better quality in the first picture and they are through glass too? surely the headlights and interior would both be better quality
Like I said, I didn't use the highest quality polariser;
everything in the photo is lower quality, even the tree - much of the detail was lost. The problem was amplified by the fact the camera was used at quite a zoom (about 14x optical). Had I used a better quality polariser the rest of the picture would have been of the same quality. Regardless, it proves my point even with my low quality filter.
jomukuk wrote:
Quite simply: Polarising filters work quite well [sometimes] when the object being viewed is stationary. When the object OR THE CAMERA viewing it is moving then the light reflection will change AND THE EFFECTIVE POLARITY of light will change.
Of course it will; however, I did ask you "
just how much do the angles change by?" - I think you'll find that if you answered that question you'll get an appreciation of the level of benefit, as well as understanding if the 'angle of the filter' actually needs to be adjusted in real-time.
The camera moving has no impact on the optimal angle of the filter; the angle of light bouncing off the incident surface to the camera is the only factor.
Don't forget, a polar filter won't suddenly pass all light when used slightly away from the optimum angle - there is a rolloff.....
As for me, I would set the angle of the polariser to somewhat less than half the typical windscreen angle knowing that I couldn't track a vehicle closer than 45 degrees of parallel to me, so still giving a significant overall benefit at any trackable angle - but what do I know........IR is a red herring. All video systems that don't utilise IR use an IR rejection filter otherwise the resulting image will be blurry, or the optics will be much more complicated/expensive to prevent further dispersion with the greater span of wavelengths. I speak as someone who has done a lot of IR work, and some of that not unrelated to the goings on of this site!
jomukuk wrote:
Most car windows now are made to reflect infrared light back, this also affects the camera. Since the camera would be using a ccd imaging system, it would also have problems with colour (most ccd imagers have an i/r filter in front of the semiconductor wafer). In any case, since different imaging systems have different response/s to different light frequencies I would certainly be surprised if a polarising system worked in a system where both target and camera are moving. And that does not even begin to take into account the light being reflected from other moving objects, at different angles of polarisation,
That's only relevant for objects that give specular reflection; skin covered faces and plastic phone cases bounce diffuse light (possibly more than the level of polarised light). Even then it's irrelevant because the objects inside the car are being illuminated from many different angles (front/sides) hence bounce many angles of polarisations, hence yes some light beams from those objects may well get filtered out, but many more others will get through - unlike the bounce on the surface of the windscreen which will all be at roughly the same angle of polarisation.
I think you'll find that many mobile speed camera operators already use polarising filters on their video systems - I wonder why!
Oh, and I took a shot of all the cars down my street (not posting that one for obvious reasons) all vehicles at different positions but the filter fixed on one position - the effect was the same on all of them. Like I said, I recommend you actually try it, or at least try to quantify the effects, instead of simply guessing at what would happen.
Login
Register
FAQ
Search
Members
Think I read something in a photo mag. Must look it up to refresh mind.
But at least our customers understand why they were pulled.
