Thread: Project NiteVision - NiteMax + HUD + carputer

Originally Posted by shotgunefx

Cool on the pot, I was wondering what that was for. BTW, with ALC, you short it to ground IIRC, not put a + voltage to it, maybe that's why?

You are correct about the pot being for the auto-iris sensitivity. There's two different types of auto-itis outputs. The one output is for "video" auto iris and I know the pot adjusts this output for sure. (Video iris is where the lens itself is smart enough to look at the video signal and decide how much to compensate.) The pot adjusts the amount of amplification applied to the video signal sent to the lens.

The feature inputs are digital and go directly to the DSP chip. They're pulled high with a pull-up resistor and you need to ground them externally to activate the feature.

Originally Posted by spaceskater

Update: I've been playing around with all possible futures of the CCD. Unfortunately the ALC has none to negative effect on the image. Basically with the auto-iris lens in operation the ALC has NO effect what so ever, in any lighting conditions. Without the auto-iris lens only when there is a lot of light in the whole image pressing the ALC will saturate the image. It looks like it deactivates the Automatic Level Control. If there is average to low lighting conditions again activating the ALC has NO effect on the output.

The ALC feature shortens the integration time (exposure time) of the CCD in bright conditions. When there's not a ton of light, the ALC function does nothing. With an auto-iris lens, your lens is making sure the amount of light entering the CCD is always moderate and therefore ALC shouldn't ever kick in.

Originally Posted by Genesisfactor

yeah, the nitemax has an automatic light control, so i would assume it had something like an autoiris...

Honestly, i'm still trying to figure out why autoiris is needed. Maybe its somethig i haven't observed yet. Maybe tonight or tomorrow night i'll have the thing in my car to see and play with. I'll record the drive.

I mean, if its the lights and stuff on a dark road, i think a software solution would be in order more than a hardware solution as a software solution could dim only the bright spots using a filter (Gaussian maybe? We breifly went over it in the computer vision section in robotics) while a hardware will dim the entire image.

A software solution doesn't know when the hardware is saturated and therefore the output has very poor dynamic range. With an auto-iris you're keeping the CCD sensor right in the operating range where it can capture the greatest detail. If you're shooting a Polar Bear in a snowstorm, your software solution won't be able to restore the detail lost by a saturated CCD.

The ALC of the CCD versus the auto-iris is a valid argument. I think they both do a similar job of keeping the image within a usable range. But...

Remember how a small aperature (like a pin-hole) eliminates the need to focus? This increases the depth-of-field because everything is in focus from up close to the horizon.

If nothing else, the auto-iris will provide a small aperature during daylight hours and provide a well focused image without regard to how far away objects are. At night, the auto-iris will cause a large aperature. This will capture more light and improve sensitivity too. However, the price you pay is less depth-of-field. Since up-close objects will be the only thing illuminated at night, the need for good fucus from up-close to the horizon is minimal -- you only need good focus up close. With the auto-iris, you get these good focus characteristics you want. The ALC alone can't do this!