Quote:
Originally Posted by cosmicma
you both have a point ( mattuk and jet_kit ) and i would say most filters can be electronically emulated apart from neutral density filters especially something like a 10 stop filter
the problem with lowering iso levels is below a certain threshold the image begins to lose dynamic range hence not seeing many cameras with iso levels below 100 and in Nikon's case 200 iso
i would imagine that when you turn the iso levels or gain down to the levels required to emulate a 10x neutral density filter things would probably be looking pretty poor
|
Welcome to the big debate.
I remain to be convinced. When you add it all up the pixel (or to give it the proper term 'Photosite') has an electronic cover over it. When you activate the shutter the cover comes off and loads of little photons drop into the pixel's bucket. When the shutter closes the software counts the number of photons in the bucket and that determines the exposure value for that pixel.
Now, if your ISO is set to 400 and your shutter speed is 1/100 you would get the same exposure value than if the ISO was 100 and the shutter speed 1/25. If you think about it the shutter is open for 4 times as long, the aperture hasn't changed so you should get 4 times as many photons in the bucket, but you don't - you get the same amount because, effectively, an electronic 2-stop ND4 filter has been placed in front of the pixel cutting down the number of photons that land in the bucket. If we take this to it's logical conclusion ISO 6,400 (or whatever the top rank is on your camera) could be considered unfiltered, ND2 is generated to give us 3,200, ND4 = 1,600 etc. Until we get to ND64 for ISO100. It isn't actually as simple as that, but the principal is sound. and it works exactly the same way as putting a dark bit of glass in front of the light gathering device (Lenses to you and me).
If the software was capable of doing this for another 10 stops (and it probably isn't -YET) the overall effect would be identical to the use of a Big Stopper. In fact, it would be slightly improved because you will not have added two further glass surfaces to the mix. The main point is that the number of photons left in your bucket is the same. The question of losses in the dynamic range is a physical issue to do with photons bouncing around and landing in the wrong bucket. This is just as likely to happen with a 10-stop filter as with the above scenario because it's related to the amount of time the shutter remains open.
Good fun this, isn't it?