VR Light Leaks with Nikon 24-85mm AF-S

There are two main advantages to taking star trail photographs during the winter: 

1. The air is colder, which can result in clearer skies. 
2. The sun sets much earlier. Often as early as 4:00pm for me. Taking photographs is a far easier endeavour in comparison to the summer when the sun sets at 9:00pm and good exposures are only possible after 12:00am.

There are two main disadvantages to taking star trail photos during the winter:

1. The air is colder and it can get down right freezing!
2. Winter often brings more precipitation, which results in more clouds, which results in less stars.

During a recent trip to Harrison Hot Springs, I was lucky enough have clear skies. So I took the opportunity to photograph the stars. My setup was a Nikon D600 with the Nikkor 24-85mm AF-S lens. After taking some initial test shots, I noticed a light leak at the bottom and lefthand side of the frame: 

Harrison Lake, facing North (24mm, f/3.5, 30sec)

With my DK-5 finder on, I initially thought that the leaks might be the artificial lights that were to my left and right.  I placed my lens hood on, but still observed the same amount of light leaking into the 24-85mm. I decided to recomposed, just to see if something would change:

Harrison Lake, facing NNW (24mm, f/3.5, 15sec)

Same amount of light leak, and in the same location in the frame. I tried various things, including using my toque to cover up various parts of the camera and lens.  After about an hour, clouds were beginning to approach, so I decided to use my 35mm 1.8 AF-S (DX, not the new FX version) to salvage what I could of the evening. 

When I moved to a warmer location, I spent some time experimenting with the Nikkor 25-85mm to see what was going on. I had used this lens for star trail photography before, so these light leaks were a bit of a surprise. I tried various experiments, including placing the lens cap on the lens and taking the exposure in a light-tight box, but I would continually get something like the following:

Light leak with lens cap on (f/3.5, 30sec)

This was very odd, considering that the surrounding light was changing during my tests, but the light leak remained the same. I suspected that the focus window on the lens was leaking light, and I even used a flashlight to help track down the leak on the camera/lens, but again, the exposure remained the same. The only thing that caused the light leak to change was zooming the focal length to 85mm:

Light leak at 85mm (f/4.5, 30sec)

Since the light leak always remained the same, all I could think of was that the camera itself was generating some kind of light internally (the view screen? the status LED? the LCD screen?). After doing some in-depth searching on the web, I came across a post on Luminous Landscape where someone theorized that the VR in lenses might be using lasers. I turned off the VR and voila, the light leak was gone:

Light leak gone! (24mm, f/3.5, 30sec)

I felt silly that my VR turned on the entire time, considering that I shouldn't have had VR turned on when my camera in mounted on a tripod! Time to create a preparatory "checklist" for my star trail photography. For those that are interested, here's a photograph taken with the 35mm 1.8 AF-S on the D600:

Nikon D600 with the 35mm f/1.8 AF-S (f/8, 30min)

Cheers!