Image Sensor AGC relation with Minimum Scene Illumination

[diraniyoussef]

How Camera AGC modification factor is used in adjusted "minimum scene illumination"?

AGC (Automatic Gain Control) switches on when light is reduced in a scene. When ligh is enough it switches off.
AGC is one of the factors to determine the "actual" minimum scene illumination, as the environmental conditions of the actual scene usually differs than the ones in the camera datasheet, so we need to calculate the actual minimum scene illumination, and not be satisfied by the value mentioned in the datasheet.
Modification Factor F = Ad/Aa,
where
Ad = AGC position in the data sheet
Aa = Actual AGC position

right from this article http://www.idealcctv.co.uk/lux-agc-s...e-illumination
they said: "If AGC off = 1, then AGC on = db figure from the data sheet"
but I still didn't understand Aa precisely, according to mentioned definition. According to the formula, when Aa gets higher, the minimum scene illumination must be lower, so the camera needs less light to obtain a useful image to a certain extent. Could Aa be a delta (difference) value that is in dB terms > 1, as difference between charges levels.

Anyway I'm guessing, is Aa the needed electricity charges in Coulomb produced right before AGC is on, at a certain level above the minimum scene illumination? if so, then the corrective factor should be Aa/Ad, and Ad is generally different than 1.

[Manfred M]

Welcome to CiC - this is primarily a site that deals with still camera photography, and while there are some members with varying levels of technical knowledge, this is not a site where I would expect a great deal of expertise regarding CCTV cameras.

As Ad is given in dB (a common measure of gain in the video world) and Factors (F) have no units, it is safe to say that Aa is the actual gain measured at the sensor output for a particular light level, also measured in dB. I'm not sure where a delta comes into play, as a delta assumes a difference, and the calculation is a division. Not quite sure where Coulombs come into this question; but obviously some level of current will be applied through the amplifier section of the sensor, but that is something for the circuit designer to worry about, not the end user.

[diraniyoussef]

Thank you.
Finally, this factor is interesting to photographer as well, since it'll help determining the minimum scene illumination. Although a photographer could do: trial & error, without taking care of any such issue.
I was curious scientifically.

Is is a dB light or dB electricity (charges)?
This factor should help on paper before trying. So how could a "curious" know the actual dB of the image sensor?
Thx

[Manfred M]

Most photographers really wouldn’t care less, as this is really not something that is measured the same way in photography. What is important to photographers is the sensitivity to light at a particular setting and this is defined by ISO, which is really an updated carryover from the film camera days and applies similar light sensitivity measurements to digital cameras. If you are interested:

http://en.wikipedia.org/wiki/Film_speed#Digital

CCTV cameras are generally unsophisticated instruments that in reality are little more that robust webcams. Some models can be moved remotely or zoomed remotely, but ultimately they provide a video signal at a given frame rate. I don’t know for sure, but I suspect that most will have a fixed aperture. Things like low noise, colour accuracy and other features that are important to photographers are of less value in CCTV. At a base level, it is possible that the only way to adjust light sensitivity is through manipulating the output gain of the video signal.

I expect that consumer level video cameras likely have a similar way of operating. Higher end pro / prosumer video cameras do provide gain control (mine has 0dB, 6dB and 12dB settings), over and above aperture control, shutter speed (rarely changed) and have built in neutral density filters to ensure that the camera operates in the sensor’s “sweet spot”.

dB is a power measurement and uses a logarithmic scale. Look at the following for a brief overview:

http://en.wikipedia.org/wiki/Decibel

If you want to know the performance specs of a specific sensor; if you know what it is and it is a commercially available sensor, try to find the manufacturer’s spec sheet on it. If you can’t, you would have to do your own test setup. That could be difficult because you would likely only be able to measure the final output, which means any intermediate signal processing would be part of your measurement.