Paul - 25 June 2008 12:44 PM
cerupcat - 25 June 2008 12:21 PM
Well that’s probably because the unibrain is more sensitive in the 800nm range and 940nm is high for it. LEDs should really be catered to the camera spec. If you used a 850nm LED, you may see your results flipped.
Quite possible!
I’d love to see if someone with the 850nm LEDs can prove that point.
I’m stuck with the 940s because I spent money on a Thorlabs IR filter for that range.
Paul,
There is no need in proving anything. The Unibrain Fire-I camera uses the Sony ICX-098BQ CCD sensor.
I attached the spectral sensitivity curve for this sensor. As you can see from the chart, the light response clearly drops as the wavelength increases.
So while at 850nm the response is around 0.2 (1/5th of the total range), at 940nm this drops to about 0.75 which is more than 2 times smaller!
This is why you get low brightness blobs. Keep in mind that these camera CCDs are ‘tuned’ for a color spectrum of the visible light. Therefore deviating much from this in terms of the light wavelength will not only result in lower light response but also a higher noise.
The whole point of having the IR light in MT setups is so that it is not visible by human eye and so that there is no intereference with the image projected to the screen. Therefore using near IR is completely fine, and there is no need to go as high as 940nm (unless you are using more expensive specialized cameras sensitive in this wavelength range).
~Alex
Click thumbnail to see full-size image