Based on that, I'm concluding that nothing but external darkness can be relied on to eliminate light leak.
Even given that my "dark" condition was in a dark box, there's still led's and lcd screen on, and the box had a reflective bright white surface, so those could potentially show up in a picture.
You can turn off or block the LCD and mask the LEDs.
To be really convinced it's amp glow, I'll have to make sure the display mode turns the screen off, and see if there's a visible difference (showing that the lcd on/off can exposure light).
It has been verified by many people on many different cameras that this is an internal effect in the camera, not a light leak.
The purple makes it clear this is the case. If you look at the raw values, you will find the the red, green and blue pixels in the area dominated by amp glow all have essentially the same value. For virtually any real light source, you will see a checkerboard pattern due to differing response of the different filters.
For a light leak to be
exactly the right mix of colors to do this is extremely improbable. Doing so across multiple cameras is basically impossible. The amp glow is typically purple because the green pixels are most sensitive and thus get turned down in the color matrix.
Attached is a crop from a 256 second dark frame from my d10 (taken with the shutter closed under artificial light). You can see that there is no evidence of the bayer pattern. Also attached is a crop from a neutral image, which appears uniform grey in the canon jpeg. The bayer pattern is readily apparent.
An external light light is also ruled out for other reasons. I have dark frames taken with the shutter closed under bright artificial light, which correctly subtract the "amp glow" from exposures taken under dark skies. In the dark sky shots, the brightest "amp glow" value exceeds the average value of the exposed image. There isn't enough ambient light available to cause that kind of leak.
