A still life setup that teaches value needs one directional light source, not the overhead light already in your room, and three to five objects with different surfaces placed close enough together to cast shadows on each other. Build the light first, choose the objects second, draw last. Skip the light and you are copying outlines no matter how long you practice.
Ambient light removes the value information the drawing needs
Most home still life setups fail before the pencil touches paper, because the room light is ambient. Overhead ceiling light and daylight from a window on both sides wrap every object evenly. There is no clear side that is lit and no clear side that is not. This is so, but the consequence is bigger than a flat-looking drawing: it removes the actual information a still life is supposed to supply.
A still life is not practice in drawing objects. It is practice in reading value, the light side, the core shadow where the form turns away from the light, the cast shadow the object throws onto the table, and the reflected light bouncing back into the shadow side from the surface below. Under ambient light none of those four exist in a readable way, so you cannot fail the check and you cannot pass it either. You end up inventing tone from memory, usually the same three grey values in the same places you always put them, regardless of what is actually sitting in front of you.
That is the pattern behind a stall a lot of self-taught artists describe without naming it: hours of practice on autopilot, where you cannot say afterward why any particular shadow went where it did. The setup taught nothing because the setup could not teach anything. One directional light source changes that immediately. Once every object has one lit side, one shadow side, and a cast shadow with a clear edge, the same hour of drawing becomes value practice instead of outline copying, because now there is something specific to get right or get wrong.
The setup goes together in a fixed order, light before objects
42 hours of pencil drawing training, lifetime access
The video course I point plateaued students to — full lessons on portraits, hands, and realistic shading, not another short tutorial.
- Place one light source to the side, not overhead. A desk lamp, a clip light, or a single window with the blinds closed on the other side works. Nothing is on the table yet. What matters at this stage is that the light comes from one direction only, roughly 45 degrees to the side and slightly above the eventual object height.
- Set the distance and angle. Move the light close enough that the shadow side of a test object (your hand works) reads clearly darker than the lit side, and far enough that the cast shadow has a defined edge rather than a soft smear across the whole table. Too close and the shadow edge goes soft and long. Too far and the light-to-shadow contrast flattens back toward ambient.
- Choose three to five objects with different surfaces. Not five mugs. Pick objects that behave differently under the same light: something matte (an apple, a ball of paper), something with a hard edge (a box, a book), something curved (a bottle, a jar), and one object with some shine or reflection if you have it. Different surfaces hold the light-to-shadow transition differently, which is the actual skill you are training.
- Set a plain backdrop. A single sheet of paper or cloth behind and under the objects, one value, no pattern. A patterned backdrop competes with the objects for attention and makes it harder to see where the cast shadow actually falls.
- Arrange the objects so they touch or overlap slightly. Objects spaced far apart on the table each cast an isolated shadow that tells you little. Objects close enough to shadow each other show you how one object’s cast shadow changes the value on the next object’s surface, which is where the real value information lives.
- Block in the big shapes first. Light pencil, no shading yet. Get the outer silhouette and the position of each object relative to the others. Nothing is dark on the page at this point, only shape and placement.
- Add the shadow shapes as shapes, not as shading. Look at the core shadow, the cast shadow, and the shape of the dark side as flat shapes with edges, the same way you drew the outer silhouette. Draw the edge of the shadow before you fill it in. This is the step that separates value practice from outline copying: you are drawing what the light is doing, not what the object looks like from memory.
- Fill in value from darkest to lightest. Establish the darkest shadow shape first, then the mid-tones, and leave the lightest lit areas closest to bare paper until last. Working dark to light keeps you honest about how much contrast the light source is actually producing, instead of guessing at a middle grey everywhere.
Four checks catch a flat drawing while the paper is still fixable
Squint at the setup and squint at the drawing side by side. Squinting removes detail and leaves only value masses. If the real objects show three or four distinct value masses when you squint and your drawing shows one grey soup, the shadow shapes did not get separated from each other on the page. Fix it before adding any more detail.
Cover the cast shadows with your hand and look at what is left. If the objects still read as three-dimensional with the cast shadows hidden, the core shadow and light side are doing their job. If the objects go flat the moment the cast shadow disappears, the form shading itself is too weak and needs more contrast, not more shadow on the table.
Hold the drawing at arm’s length under the same light as the setup. A drawing that looks correct up close under studio light can go flat under a different light. Checking both the setup and the drawing under one consistent light source is what makes the comparison mean anything.
Ask where the darkest dark and the lightest light are, on both the setup and the drawing. If you cannot point to one spot on the real objects that is darkest and one that is lightest, your light source is still too ambient and the setup needs fixing before the drawing does. If you can point to both on the objects but the drawing has neither, the value range on the paper is too narrow.
Four setup mistakes undo the practice before the drawing starts
42 hours of pencil drawing training, lifetime access
The video course I point plateaued students to — full lessons on portraits, hands, and realistic shading, not another short tutorial.
Drawing under the room’s overhead light because moving a lamp feels like a hassle. This is the mistake underneath most of the others. Overhead light produces the ambient, shadowless result described above no matter how carefully you draw afterward. There is no shading fix for a setup with no shadow information in it.
Using too many objects. A crowded table looks more like a real still life painting, but it is harder to read and it turns the drawing into an exercise in fitting things onto the page rather than reading value. Three to five objects with different surfaces gives more usable information per object than eight or ten similar ones.
Starting with white objects on white cloth. It looks clean and controlled, and it is the hardest possible first setup. All the value information has to come from a narrow range near the top of the scale, which is difficult to judge accurately before you have practiced on a setup with a full range of values. Save the white-on-white setup for later, once you can already place a core shadow and a cast shadow correctly on objects that make it obvious.
Photographing the setup and drawing from the photo instead of the objects. Here is the honest one: a camera compresses the value range your eye actually sees and flattens the exact contrast the setup exists to teach. Drawing from a photo of your own still life is a real fallback when you truly cannot keep the objects out for the session, but treat it as a fallback, not the default. The whole reason to build the setup is to see the light directly.
This setup is not the right fix for everyone. If your drawing already shows a clear light side, core shadow, and cast shadow and your actual stall is in proportion or perspective, a lighting change will not move the needle, that is a different problem with a different fix.
One session tests the setup, a month of sessions tests the habit
Tonight, build the setup as described above, thirty to forty-five minutes total. Ten minutes to set the light and objects, twenty to thirty to draw. Do not skip the shadow-shape step to save time, it is the part doing the actual teaching. One drawing is enough for the first session. The point is not volume, it is confirming the setup produces a light side, a core shadow, and a cast shadow you can actually see and draw.
Over a month, keep the light source consistent from session to session and change only the objects. This isolates what you are learning: with the light variable held steady, any improvement in how the shadows read on the page comes from your handling of value, not from a different setup doing different work each time. After a month of sessions built this way, the squint check should start passing on the first pass more often than it fails, and the shadow shapes should start going down on the page as shapes before they get filled in, without you having to remind yourself to do it that way.
Common questions about building the setup
What objects should I use?
Three to five, with different surfaces: one matte, one with a hard edge, one curved, and one with some shine if you have it. Avoid five objects of the same material and finish, they will all take the light the same way and you will only be practicing one kind of transition.
What light should I use?
One source, coming from the side rather than overhead, at roughly 45 degrees and close enough to produce a clearly darker shadow side. A desk lamp or a single window with other light blocked both work. Two light sources fighting each other from different directions puts you back in the ambient problem this setup exists to fix.
How long should one still life take?
Thirty to sixty minutes for a session built around this setup, with the block-in and shadow shapes done before any detail work. Longer sessions do not add value information, they add rendering time on top of a structure that is already decided in the first fifteen minutes.
Can I draw from a photo of my setup?
As a fallback, yes, but not as the default. A photo compresses the value range you would otherwise see with your own eye, which works against the exact thing the setup is built to teach. Build the setup and draw from it directly whenever the session allows it.
How do I know if the practice is working?
Run the squint check and the cover-the-cast-shadow check from this article on the same setup every few sessions. When the shadow shapes start appearing on the page as clean shapes before you fill them in, and the squint check passes without extra correction, the setup is doing its job.




