You may have explained your process already elmarko (in which case I need to find it), but are you using a lightbox/grid system? Your creations look deadly accurate.
I put it a few pages back, it's just a case of isolating the layers using photo-shop (by playing with contrast brightness) - then drawing those layers onto paper (taking regular measurements in the key areas such as the eye to eye distance, eye to top of page, bottom/middle chin to bottom of page etc) - pencil sketching the black section after you have measured out the key marks (Gestural lines first (two layers of gestural) then fine detail lines (erasing the extra lines), block shading the black, then doing the other shades pretty roughly.
As I only use black & 4 or 5 shades of grey - it's easier to isolate the shades so you can just add them on one layer at a time.
The only part I'm meticulous with is the black layer, the rest is adlibbed - but no, I'm not using a light-box (not really much point as you don't learn anything doing that) - if you look at the hair sections close it's pretty obvious (I suck at doing hair).