Diffusion Geekery

, , Leave a comment



I cleared up some questions I’ve had about diffusion and tested some different diffusion material against the usual suspects.

I used the Desisti 1K in a 16 foot space, with my friend here in the middle. Using the Cinemeter II app on my phone and the Luxi incident bubble, I set it at 800 ISO, 24 fps, shutter speed 24. The image above metered at f11. I tested 250, 216, Tracing Paper, Opal and then two kinds of fluorescent light (ceiling fixtures) polystyrene panels I picked up very cheaply at Lowes. I’ve been interested in those for a while because they seem to spread the light more evenly, diffuse just as well as 250 and lose less light in the process. As it turned out, I was right, but I also found out a few other interesting things.

Here’s the big comparison. For exposure, the images are not as important as the meter readings, so look at the numbers. The other thing to pay attention to is the evenness and spread of the light on the wall. It changes quite a bit.



There are a number of interesting points here. First, the 250 diffusion and Panel 1 (details about the panels at the end of this post) reveals that the panel is a full stop brighter. In addition, the spread of the light is more even with the panel.


The 250 and the tracing paper are also very close, but tracing paper is a lot cheaper and it also creates a more even spread.


Unsurprisingly, the Opal does very little by itself. I personally find it works best to take a little more edge from an already diffused light.


And there is a strong similarity between Opal and Panel 2, although once again, the panel creates a more evenly spread light.


The panels themselves are just hard plastic sheets used to diffuse light from overhead fluorescent fixtures. All the big box stores carry them and you can often find damaged sheets for a dollar or two. I cut these with a table saw into 24″ squares. You can cut them with a utility knife but I don’t recommend it. They damage easily because they are very brittle, especially around the edges. I put a small wrap of gaff around the edges to protect them, carry them in a portfolio and have not had any breakage problems. To deploy I just bite them with a duckbill or sheet metal clamp. They also work wonders with LED panels, which I’ll get to in another post. I shouldn’t need to say this, but don’t put this material in front of an HMI unless you can put several feet in between, and even then I’d be cautious.


These panels are very similar in design and effect to the lenses for Par lights like HMIs. The smaller the pattern, the wider the diffusion. Panel 1 (below) is a very small, irregular pattern and you can see the effect in the illustrations above.


Panel 1 detail:


Panel 2 is a slightly larger pattern with not as much diffusion effect:


Panel 2 detail:



Leave a Reply