How to Calculate Light Output Through Diffusion

You can’t. You must test. Here are some test results.

Recently I was looking for a formula that would allow me to calculate how much light output I would get from a given size HMI once I punched it through a 12×12 diffusion frame. Calculating light output for an un-modified light source is pretty straightforward; tools like the iOS app Pocket LD offer a catalog of heads, and tell you what their photometrics are at a given distance. But diffusion is a much different animal…there are so many variables, really the only way to know how a light source output will be affected is by experience and testing…

Thankfully, two DPs shared some diffusion and bounce testing, and provide an incredible resource for gaffers and other DoPs with their Diffusion & Bounce Rag Tests. These tests were shot by Stephen Murphy and Ed Moore, and offer a collection of still frames, video, and exposure data from a series of tests that show how myriad types of diffusion and bounce affect a 10K source.

It's a very thorough roundup of materials…they cover Full Silk, Half China Silk, Quarter Silk, Rosco Soft, Half Soft Frost, Rosco Hi-Lite, Rosco Half Hi-Lite, Bleached Muslin, Unbleached Muslin, Full Grid, Full Silent Grid, Half Grid, and Quarter Grid. They they round out the tests with each of these materials in combination with an 8×8 Ultrabounce, in a book-light configuration. For each setup, the only parameter that changes is the diffusion material and the camera's f-stop to compensate for proper exposure. Using this data, you can determine what kind of firepower you may need to get through a large diffusion panel and still acheive the stop you need at subject. You can watch the test video below, or browse still framegrabs at The accompanying blog post includes more data on the setup, and offers overhead diagrams as well. 



Matthew Jeppsen

Matthew Jeppsen

Matt Jeppsen is a freelance cinematographer with over a decade of experience in commercials, music videos, and documentary films.