By device
Projector test
A projector is the one display where the room is part of the picture. The patterns are the same and what they mean is not.
Last updated
Before you start
- Let the projector warm up for at least twenty minutes. Lamps and lasers both shift in brightness and colour as they reach temperature, and a cold projector will send you chasing faults that disappear.
- Get the room as dark as you can. A projector has no control over black at all: whatever light is in the room lands on the screen and becomes your black level.
- Set the projector to its native resolution from the source. A projector scaling a mismatched signal softens everything, and no focus adjustment recovers it.
- Turn off keystone correction and any digital geometry adjustment before judging sharpness. Both work by resampling the image, which costs detail everywhere.
- Turn off eco or dynamic lamp modes while testing, because they change output in response to content.
The run, in order
- 01 Sharpness test
Focus first, and focus at the corners rather than the centre. A projector that is sharp in the middle and soft at the edges has a lens or alignment problem that no setting fixes.
- 02 Screen uniformity test
Brightness falloff towards the corners is normal on a projector and much larger than on a panel. This tells you how much you have and where.
- 03 Full screen colour test
Full-field primaries, for convergence. On a three-panel projector a misaligned panel shows as coloured fringing at the edges of the frame.
- 04 Contrast test
The checkerboard tells you how much light is bouncing around the room and back onto the screen, which is what actually limits projector contrast.
- 05 Black level test
The near-black ladder in a dark room. Projectors crush shadows more than panels do, and the setting that fixes it is usually on the source rather than the projector.
- 06 Gradient test
Banding, which projectors show readily because the image is large and the eye has room to see the steps.
- 07 Motion blur test
Last, because single-chip projectors introduce their own motion artefacts that have nothing to do with the source.
What matters on this hardware
Focus at the corners, not the centre
Almost every projector is sharpest in the middle. The useful question is how soft the corners are by comparison, and whether the softness is symmetrical. Asymmetrical softness usually means the projector is not square to the screen.
Keystone correction costs detail
Digital keystone squares the image by resampling it, which throws away resolution across the whole frame. Moving the projector until the image is square without correction is always the better answer.
Black level is the room
A projector cannot make black. It can only fail to add light, and every surface in the room reflects some of what it does add back onto the screen. Dark walls and ceiling do more for contrast than any projector setting.
Rainbow artefacts on single-chip models
A single-chip projector shows colours in sequence through a wheel. Some people see brief colour flashes on high-contrast moving edges and some never do. It is inherent to the technology rather than a fault.
Convergence on three-chip models
Three separate panels must line up exactly. When they do not, edges show coloured fringes that grow towards the corners. Some projectors offer a convergence adjustment and many do not.
The screen surface matters
Gain, colour and texture of the surface change the image as much as many projector settings do. A white wall is not a neutral reference, and a textured one adds its own pattern to every flat field you look at.
Questions people actually ask
Can these tests really be used on a projector?
The patterns can, because they are exact colour and geometry fields and it does not matter what displays them. What changes is the interpretation: uniformity, black level and sharpness all behave differently when the image is light thrown at a surface rather than pixels emitting on their own.
Why does my projector black look grey?
Because it is. A projector makes black by sending as little light as possible, and everything it does send bounces off the walls, ceiling and screen and comes back. In a room with light surfaces the black level is set by the room rather than the projector.
Should I use keystone correction?
Only as a last resort. It squares a skewed image by resampling, which softens the whole frame. Physically moving or tilting the projector, or using lens shift where it exists, keeps the resolution you paid for.
How do I check convergence?
Show a white field with fine detail, or the sharpness patterns, and look at the edges of high-contrast features near the corners. Coloured fringing that changes colour on opposite sides of the frame is a convergence problem rather than a focus one.
Does the pixel-exact test work on a projector?
Only if the source is running at the projector native resolution and the device pixel ratio is a whole number. Even then the optics blur single pixels, so the patterns behave more like a focus check than a gamma reference. The page will say when it cannot give an honest reading.