[Steven Sweeney]

Julia, I'm the former reigning prince of verbosity here, but I've semi-abdicated, so I'm going to use some images to replace at least a couple thousand words. I presumed permission to do so as an educational tool, not to "improve" your drawing. I trust that the presumption will not be annoying.

I think that the stiffness and tired look comes not from the model but from an overmodeling of the head and the features as a whole, which then produces an angularity and harshness that translates into fatigue.

By overmodeling, I mean that the nuances within a large value shape are too exaggerated. The dark values become too dark in too-discrete forms, and the light values as well. For example, there would be some reflected light along the jaw line in the shadow side of the face, but not as light as you've drawn it. However -- if it WAS that light, then as the artist in charge, your job would have been to tone it down, to keep that area within the overall value range of the much larger shape of which it was a part. My instructor used to admonish me not to "look too hard into" the dark areas, trying to find some light to pull out. Areas of reflected light are very often the trouble spots in this regard.

So here are a couple of images, to illustrate. I've set side by side your drawing, and an adjustment that doesn't change a single line of your drawing, but only (1) reduces the magnitude of value changes within the large value shapes you'd see by squinting at the model, and (2) softens some edges, both at the contour of the form itself and at the areas where two different value shapes join. My "re-valuing" is heavy-handed and I'm not trying to present a finished result, just an idea.

The second image is posted to illustrate what I mean by "large value shape." I don't mean to say that the value within that shape is identical throughout, but that it's in the same value "class" on the form, with relatively minor adjustments needed (or allowed). So, in the red-outlined shape, no light value is lighter than the light values outside that shape, and no dark values are darker than the dark values outside that shape. That's an oversimplification and exceptions immediately present themselves, but that's the idea. Get the LARGE value areas working for you, and save the accents (dark touches) and highlights (light touches) for the end, for punch, rather than compromise the large value shape by breaking it up. (By that, I'm referring to the too-light reflected light along the jawline and the lower cheek, and some of the darks on the ear -- ears are notorious for wanting to be overmodeled.)

Our mantra was "Squint" when in doubt, and those very large value shapes will reveal themselves. You want to try to keep those shapes intact, without introducing stray darks and lights that really belong in other value shapes. The goal is to have all the intact, large value shapes describe form, rather than break it up. First, simplify -- and then only as necessary, add detail.

I went overboard on the softening comparison image. You'd have to go back in now and put in some dark accents and a perhaps a very few highlights, especially around the eyes, to punch up the value contrast there and draw the viewer's eye to that area. A little of each (accent and highlight) goes a long way. Both are compressed, powerful effects.

As for the makeup. I'd have tried to ignore it as much as possible. A life model for pencil or charcoal work ought to be invited to dispense with that distracting effort. This isn't a glamour photo shoot, but an inexperienced model might be confused about that. If they do it anyway, finesse it diplomatically and try to draw "through" it, to the anatomy, to the form.

I softened or lightened a couple of areas that were too extreme in value differentiation, such as the sharp line at the top of the "bulb" of her chin, and some of the dark lines describing her nose. I darkened some of the reflected light at the base of the nose. Think of the base of the nose as a plane. It is turned away from this particular light. It should be pretty much one value, without high reflected lights or dark shadow edges.

Okay, I'll post the images and you can ask about them if I have been insufficiently clear, or sufficiently verbose. Incidentally, I swiped some darks around the fabric of the garment, just guessing at the flow of light. I always try to "see" the light source and watch how that light would play around and describe the form.