One trick that sometimes helps when you're trying figure out something like this is to first get a better white balance, and then greatly increase the saturation and local contrast. I did that for this photo (see attached), but, to be honest, it didn't point in any directions that haven't already been suggested. Specifically, since there is obviously lots of light bouncing around in that lens (flashing dashed black lines in the areas of obvious lens flare), that means that we can't exclude lens flare from being the cause of the weird smear of light by the door.
Of course, a lens hood *always* is a good idea, and if the sun is just outside the frame, sometimes one has to supplement the lens hood with your hand or a card held at arms-length in such a position that the lens is in its shadow.
As also pointed out, another source of anomalies like this are OOF bright objects (eg, specs of dust, insects, glints off of cars, etc.) anywhere from on the front of the lens/filter to many feet in front of that.
However, my current favorite theory is one that hasn't yet been mentioned: Ambient light enters the eyepiece (ie, in the reverse direction that it usually travels), and the foam or felt that is supposed to provide a light seal around the mirror when it's in the "up" position is old and crumbly, so a tiny amount of light gets past this point. If this happens, light enters the mirror box, makes a bounce or two and winds up hitting the sensor in some strange pattern. This is one of the reasons that higher end cameras often have an eyepiece shutter that you can use in such situations:
http://static.photo.net/attachments/bboard/00V/00VWcm-210925584.jpg
This problem occurs much more often when using strong ND filters, so the light entering the usual path (through the lens) is much weaker than normal, but the ambient light level is quite bright. This theory also explains the relatively sharp v-shaped structure inside of the smear -- a small pin-hole sized light leak will cast relatively sharp shadows.
BTW, a pinhole light leak in the bellows of view cameras can produce all sorts of weird patterns on the image, some of which look quite similar to the pattern in Hershy's photo.
Go to Google Images and search for "light leak bellows", and you will find images like these:
http://www.instantoptions.com/faqs/images/filmproblems/integral-lightleak2.jpg
http://4.bp.blogspot.com/_vI0kN05tr...t5c/s1600/01Nov+-+800Z+-+Ikonta+520+A+011.jpg
Just my $0.02,
Tom M
