Terrywoodenpic
Suspended / Banned
- Messages
- 8,172
- Name
- Terry
- Edit My Images
- Yes
Is there a need to things differently to get the most out of Digital Photography?
When digital cameras came on to the scene there was an obvious need to take advantage of the skills Film photographers already had.
There was a conscious effort at equivalence at every stage.
A sensor was equivalent to a film
Illumination and exposure were directly related to film speed and ISO
A Digital ISO setting was the equivalent of film speed.
Pixels densities were the equivalent of film grain size.
Film development and raw processing were equivalent.
Print sizes were related to pixel numbers.
Graininess and noise were related
For a while I thought all these things were true
when in fact none of them are.
It is now very difficult to start again thinking with a clean slate.
The most important consideration is that film and sensors are nothing like each other... they do not even do the same thing.
Film receives and stores an image formed by photons, these release within it self, atoms of silver on grains of sensitised halide. The sensitivity of that halide emulsion is predetermined in manufacture, and combined with further enhancement (development) grows those silver atoms into a visible image. Graininess is the spaces between those developed grains.
The film becomes the carrier of the image
A sensor is quite different, but it too captures both photons and and passes them on as signal of data and noise. Each discrete segment (pixel) of that sensor is covered by a coloured filter. In turn each pixel is affectively a photon counter and passes that data to a central processor which then either passes that partially processed data as a raw file on a data card. Or more completely processes it as a JPG.
A sensor has a fixed native sensitivity to photons, It does not have an ISO speed.
I used to suppose that the ISO setting was achieved by amplification of the signal and so the sensor became more sensitive... like using a faster film. But after reading a few Geoffrey Crawley and professor Newman's articles I realised that this is not the case.
The sensitivity remains fairly constant. The ISO setting simply changes the processing of that signal.
The result is that a raw capture is essentially always at native sensitivity and is proportional to both the illumination, exposure and Photons counted . It has no ISO value until a modifier is applied during processing.
A JPG also originates from the raw signal but is processed in camera, using predetermined criteria derived to produce an equivalent ISO speed. (it disposes of the unwanted data)
When we change an ISO setting we are pre selecting a particular slice of the raw signal data. We are not increasing sensitivity.
This has further implications....
End of first instalment...
When digital cameras came on to the scene there was an obvious need to take advantage of the skills Film photographers already had.
There was a conscious effort at equivalence at every stage.
A sensor was equivalent to a film
Illumination and exposure were directly related to film speed and ISO
A Digital ISO setting was the equivalent of film speed.
Pixels densities were the equivalent of film grain size.
Film development and raw processing were equivalent.
Print sizes were related to pixel numbers.
Graininess and noise were related
For a while I thought all these things were true
when in fact none of them are.
It is now very difficult to start again thinking with a clean slate.
The most important consideration is that film and sensors are nothing like each other... they do not even do the same thing.
Film receives and stores an image formed by photons, these release within it self, atoms of silver on grains of sensitised halide. The sensitivity of that halide emulsion is predetermined in manufacture, and combined with further enhancement (development) grows those silver atoms into a visible image. Graininess is the spaces between those developed grains.
The film becomes the carrier of the image
A sensor is quite different, but it too captures both photons and and passes them on as signal of data and noise. Each discrete segment (pixel) of that sensor is covered by a coloured filter. In turn each pixel is affectively a photon counter and passes that data to a central processor which then either passes that partially processed data as a raw file on a data card. Or more completely processes it as a JPG.
A sensor has a fixed native sensitivity to photons, It does not have an ISO speed.
I used to suppose that the ISO setting was achieved by amplification of the signal and so the sensor became more sensitive... like using a faster film. But after reading a few Geoffrey Crawley and professor Newman's articles I realised that this is not the case.
The sensitivity remains fairly constant. The ISO setting simply changes the processing of that signal.
The result is that a raw capture is essentially always at native sensitivity and is proportional to both the illumination, exposure and Photons counted . It has no ISO value until a modifier is applied during processing.
A JPG also originates from the raw signal but is processed in camera, using predetermined criteria derived to produce an equivalent ISO speed. (it disposes of the unwanted data)
When we change an ISO setting we are pre selecting a particular slice of the raw signal data. We are not increasing sensitivity.
This has further implications....
End of first instalment...
Last edited:
.
