[Leica] IMG: Another Comparison

[len-1 at comcast.net (Leonard Taupier)]

When I first read Brian's comment about an IR problem I thought no  
way. Then the more I thought about it the more I thought it might be  
possible. Now I don't know anything about photo sensors or amplifiers  
in the visible light range, but I do know about RF amplifiers,  
detectors  and amplifiers in general. Since the M8 sensor is allowing  
a much greater frequency bandwidth from Uv to IR, the power to the  
amplifier increases and creates an increase in distortion and makes  
the amplifier much easier to overload. In order to avoid overload the  
signal to the amplifier is reduced by gain reducing the circuit. The  
amount of gain reduction (AGC) is controlled in the design of the  
amplifier The design curve of this AGC system can be developed in the  
lab on a test bench by duplicating the actual operating conditions of  
the total system. However this is under ideal conditions using test  
generators of the input spectrum of concern, visible light. However  
if the incoming spectrum in the finished product is greater than what  
the design model was set up for, you can go into an overload  
condition. However it is not that easy to develop a perfect AGC  
system.  If you gain reduce the system too much, you introduce noise.  
That may be what the increased noise is from in the higher ISO  
ranges. The system may be in an over AGC'ed mode and still not fully  
getting out of overload. The amount of AGC is also probably  
controlled by the ISO setting. Dropping the ISO and then decreasing  
the exposure would certainly reduce noise, but also increase the  
likelyhood of overload.

Please forgive my ramblings. I'm still thinking like an engineer. It  
makes sense to me, though.

Len


On Dec 20, 2006, at 11:10 AM, Frank Filippone wrote:

> Then the 486 filter should fix the issue......it would kill the IR  
> components......easy to test for.....
>
> Frank Filippone
> red735i@earthlink.net
>
> Given that candles are a strong source of IR, I would guess that  
> this is almost entirely an IR problem.
>
>
>
> _______________________________________________
> Leica Users Group.
> See http://leica-users.org/mailman/listinfo/lug for more information