Spotmeter test - Sekonic 778

[Lars Åke Vinberg]

I finally got around to test my Sekonic L-778 light meter this weekend.

UV: Test conditions were clear skies at noon, with and without UV filter in fron t of the lens, and then same test at night indoors. Ther was no measurable diffe rence between outdoor and indoor measurements - in both cases the UV filter incr eased exposure by 0.1 stops.

Linearity: Using a B+W 8x ND filter, I mesasured at different light conditions f rom brightr sunny skies to lowest measurable light. The ND filter consistently r educed exposure by 3.0 stops.

Infrared: My kitchen stove has heaters that glow when heated. This affected expo sure by 2-3 stops. My conclusion is that the meter is oversensitive to the near infrared spectrum. However, a hot but not glowing surface (such as a stove heate r turned off after glowing) does not affect exposure. My conclusion is that the meter is not sensitive to the far infrared spectrum.

Beast regards, ?ke

[David Kaufman]

I also have a Sekonic L-778 which I purchased for its flash spot ability. But I have tested it against a Zone VI modified Pentax digital meter, which I also own, and the Sekonic is far less linear in its response. Using a Macbeth colour chart in bright sunlight and T-Max film, the Sekonic was linear in its response for objects in the red, brown and flesh tone ranges, but showed reading much too low (as much as a stop and a half) for subjects with intense blue components, and too high (by a stop) for subjects with lots of green and some yellow components. The Pentax meter was linear across the board but a little bit inaccurate when reading very brights subjects, tending to produce readings that are too low by about a stop for the white or bright yellow patches. The Sekonic was more accurate in this respect. To sum up, the Sekonic L-778 is not linear at all compared to the Pentax meter and should be used with caution by Zone system practitioners.

[Lars Åke Vinberg]

Didn't think of using a MacBeth chart for checking linearity - that is of course a more challenging test. I do have one somewhere, so I'll try it out. /?ke