Is ISO the new megapixel?

Consumers are beginning to see how megapixels don’t make a better camera, so camera manufacturers (or marketers) are looking for other numbers to slap on the front of the box. One number that many cameras have begun to tout is the ISO. In some cases, manufacturers claim “low-light” or “boost” modes.

In the days of film, ASA or ISO was a measure of how sensitive the film was to light. If you remember, 100 was meant for shots out in the sun, 200 was for cloudy days, and 400 was for indoors.

In digital cameras, the numbers mean pretty much the same thing, but with one crucial difference – you don’t actually make your sensor more sensitive when changing your ISO. All you are really doing is amplifying the signal from the sensor – and this is bad.

Why? Noise – that’s why.

The sensors on our cameras produce a small amount of noise – that’s a given. This noise shows up in the image as random dots of color. In a low ISO image, the image exposed is usually bright enough so that no noise is visible. However, when you raise the ISO, the signal from the sensor – noise and all – is amplified. This makes more of those irritating dots show up and covers your image in a multi-colored speckle.

So what can you do? Check the other numbers on the camera like the maximum aperture – a lower number (like f/1.8) lets more light into the camera and does more for you than a higher ISO. That, or consider getting a DSLR.
—Luis Cruz

Comments

[chris]

I hated the performance of p&s for indoor shots. When I got a dSLR I knew certain lenses (prime f/1.7 lenses for example) would be better. Still, I'd like to take shots with a kit lens (f/3.5 and above)...and the only way to really do it with out spending $$$? Slow shutter, and image stabilization. It's great...because maybe we need a tripod, maybe we don't.

[fred]

Fuji makes some decent high ISO p&s with the F series, even the Z series is pretty good.

[f sessink]

As technician, not as a photographer, I follow the discussion about ISO versus megapixel and I see a missing element in all these discussions. In 100 ASA film the noise is caused by the physical size of silver particles. Exposure time has no influence on this "quantization" effect; the total amount of light for a good photo depends on diaphragma and exposure time; the "quantization" noise itself is not affected and is always there. Higher ASA, or ISO, just gives larger-size silver particles and for the old film there is a direct relation between noise (crackle? We call it "korrel") and ISO sensitivity.

For CCD sensors there is a new parameter: averaging time (or exposure time for photographers). Noise is affected by the exposure time. When the total amount of light for a certain picture is the same then the noise is also the same.

The problem in specifying the ISO-sensitivity of a CCD is that this relation is never used or given; in fact unknown. Proper specification is sensitivity and averaging time. Only the amplification factor of the electrical signal is used to have that higher sensitivity. No manufacturer will give the degradation factor.

Conclusion: the signal to noise ratio for a specified light level on the sensor is a specification point and the rest is "noise amplification". This is not new; most comments give the same conclusion without explanation.

F Sessink
(The Netherlands)