Depending on the type of
sign involved a photograph will not capture flat, accurate results due to the effects of perspective.
A photo is no big deal if you're just taking a reference photo of some flat panel
sign whose graphics can be easily reproduced. If it's a more complicated repair job then a photo won't do much good. For instance if you have a channel letter
sign with some complicated joined script and the trim-capped face is broken it will take one of two sources to make a new face. One source is the original art files. The other is making a pattern and then using that to manually cut a new face. A photograph of the damaged face won't yield anything that registers correctly with the rest of the
sign. The replacement face won't fit properly.
Post processing a
sign photo to remove effects of barrel distortion and perspective will work only so well. You can get really close by using a good DSLR mounted on a tripod (and using the right combo of lens and shooting distance). But nothing will yield a perfect fit thanks to the foreshortening effects of perspective.
It would be really cool if there was a kind of device that could draw/digitize large real world graphical objects and capture them into the digital realm without using a tablet surface. As far as I know nothing like that exists. I think a tablet-free pen would at least need to be connected to some home base device that accurately tracked the pen's location in 2D/3D space as well as record input data when one or more buttons are pressed.
