|
Dan,
I would not assume that the value relationships in your painted portrait are as "incorrect" as they appear to be when you compare them to those in the photo. I've done color sketches of people from life, then taken a photo, and the photo looks quite different.
My advice would be to try to work the life study up to a point where you consider it just about finished, then only use the photo to "finish" the non-essential details and/or make a few minor adjustments. It is probably not going to be helpful to simply start a portrait from life and switch to a photo at this stage. You will get such different information from the photo that you basically have to start over.
I think Mischa's advice about looking for the large areas of light and shadow first, then working towards smaller areas of color within those larger areas, is excellent. It appears that you are rushing into the small areas before you have established the value relationships in larger areas. A good example of this is in the eye area. The entire eye socket is in shadow. The best way to approach the eyes is to paint the shadow (as in post # 8) and then start looking for variations within it. You make a "bed' for the eye and then the eye just falls into place. The value range within the shadow of the eye socket is really very small--whereas in your portrait you have made the irises bright, light blue as if in direct light.
Don't be too discouraged. You are well on your way to achieving the proper value relationships. Rather than relying on books, the best way is to try different mixtures and see what looks better. Achieving transparent shadow areas is not easy. Many color mixtures make mud! I have had success mixing blues and violets with cadmium yellow deep and cadmium orange, which greys the color (using complementaries) and also adds warmth and coolness to shadows. These can be mixed in with a rosy tone like cad red, or permanent rose, to achieve a good skin tone. But that's only one way--there are countless others.
|