Subscriber Discussion

What Is The "Effective ISO"?

Undisclosed Manufacturer #1
Oct 24, 2013

Customer request informaton about "Effective ISO at 0 gain".

What is the Effevtive ISO? If anyone knows about this, please explain.

Brian Rhodes
Oct 24, 2013
IPVMU Certified

I suspect this is a somewhat awkward cross reference to photography cameras, and ISO speed vs. gain to project some sort of wide-dynamic range ability.

[From this page:]

In digital cameras, typically photography still cameras, the ISO speed is 'simulated', since digital cameras use no film. In most cases, it is used as a method of adjusting contrast in the scene.

However, ISO speeds are not common with IP surveillance cameras. Factors like f stop and shutter speed are manipulated to influence light exposure, but I cannot recall any that use ISO.

John Honovich
Oct 24, 2013

ISO is to photography as AGC is to surveillance cameras. If the camera had 0 gain, it would be the equivalent of the lowest ISO setting on the camera. See our AGC test results.

Oct 24, 2013

There is the ISO 12232, which describes ISO sensitivity measurement of digital cameras. No mention of an "effective" ISO there, but could still be what is intended by the customer.

Matt Ion
Oct 24, 2013

There IS a correlation between ISO and other exposure factors that would allow one to provide an answer to this question. Let me see if I can articulate it here...

Going back to the photography realm, ISO is (was?) used as a rating of a film's sensitivity. ISO 200 film was twice as sensitive to light as ISO 100, and half as much as ISO 400. As with anything, there were also trade-offs: 200 would be "grainier" than 100, but not as much as 400, although there's not the same mathematical progression (ie. 200 isn't twice as grainy as 100).

Film "grain" corresponds to the size and density of the crystals of light-capturing compounds on the film, so it's roughly analogous to the pixels of a sensor - higher-ISO film has larger crystals to gather more light at the expense of "detail", like a sensor having larger pixels at the expense of resolution.

Also as with sensors, there was a wide variation in grain quality depending on the manufacturer, the particular formulation... and the cost. For example, I used to shoot sometimes with Kodak Ektar film, which was their particular line of higher-grade film. For nightclub work, I would use Ektar 1000, which had about the same level of grain (clarity) as most other ISO 400 films, but with one-and-a-third more stops of sensitivity - that is, I could shoot with a significantly faster shutter. I also paid about a 30-40% premium for this film over most standard lines. They also made an ISO 50 version of this film that was so incredibly fine-grained, on a shot of a radio tower silhouetted against the sunrise, I could make out the bolts attaching the guy wires on a 4x6 print, and could see magnified bumps in the paper before I could detect any grain.

So anyway, regarding coming up with this "equivalent" number... in photography, there's a rule-of-thumb called "Sunny 16", which states that in unobscured sunny conditions, with an f/16 aperture, your shutter should be 1/ISO... so with ISO 100 film, you'd shoot 1/100s; for ISO 400, you'd use 1/400s, and so on. This would generally give you "ideal" exposure. Now assuming your lens does a max of f/2 (like many CCTV lenses), Sunny 16 with ISO 100 would dictate a shutter of 1/6400s (f/16 -> f/2 = +6 stops, so your shutter needs to go -6 stops from 1/100 to 1/6400).

So in this example, if you set an f/2 lens wide-open, lock the shutter at 1/6400s, and come up with a more-or-less ideal exposure (no blown highlights, no lost shadows) in bright sunlight, you can report that the "equivalent ISO" is about 100. If that's too dark and you have to drop the shutter to 1/3200s, then your "equivalent ISO" would be closer to 50; if it's too bright and the shutter needs to go to 1/12,800s, your "equivalent ISO" would be 200.

Of course, Sunny 16 is just a rule-of-thumb and hardly a basis for scientific measurements, but hopefully you get the idea. With the proper lab equipment, you could measure the output of a sensor while feeding it a carefully controlled amount of light and get a much exact number, although even then, the ISO number of a digital sensor is still somewhat arbitrary, as it will provide an INDICATION of light sensitivity, but not a DEFINITIVE number... and as with film, it's little more than a vague suggestion as to the actual image quality produced. As Brian notes, the sensitivity of the sensor itself doesn't change, so when you "adjust" the ISO on a digital camera, what you're really adjusting is the gain of the sensor, electronically boosting the signal from that 0-gain setting, as John describes.

New discussion

Ask questions and get answers to your physical security questions from IPVM team members and fellow subscribers.

Newest discussions