To review the exercise we did in class:
Working from a photo, make a study of just the middle values.
Begin by mixing up a puddle of a single color that is halfway between black and white in value. Try to estimate how much paint it will take to do the job, and mix a little more than you think you need. Be sure to use a color that can get dark enough to represent the strong darks you’ll be adding later.
Draw the profiles of the shapes that are very light, and paint everything else with your middle value wash. You will be painting around the lights, leaving the white paper to represent them. It will probably be necessary to round the middle values up or down to make them all the same. The idea here is to over-simplify the subject deliberately, to get an idea of which information is essential and which is optional.
What you have done, basically, is to reveal the pattern that the lightest lights make, which allows you to stand back, and assess the extent to which the lights alone realize an illusion of believable light, space and substance.
When the middle value layer is dry, paint in the strong darks with the darkest version you can make of your color.
Where would you like to see more subtlety or specificity? Where is this oversimplified treatment surprisingly satisfying?
In class, everyone got to the stage where the lights and middle values were visible. Finish this study by adding the darks, and then make another one from an image of your choice.