April 25, 2020
From the other entries, you can see that I’ve been stabbing away at measuring and getting predictable results with artificial lighting. As time has gone on, I have added flashes from other vendors (Godox, Profoto, Nissin) as well as third party triggers (Godox, Profoto, Nissin, Pocket Wizard). These additions have not made the objective of getting predicable results any easier.
For this exercise, I wanted to start at the ground floor and work up. The simplest configuration is on-camera flash and TTL. If the flash is close enough to the subject, I would expect consistent, correct, repeatbale reults. So, here goes.
I tested three, on-camera flashes:
Nikon SB5000
Godox V1
Profoto A1
The camera / flash combo was placed approx. 68in from the subject. The exposure was set to manual 1/160 sec and the aperture was varied.
The subject contained a ColorChecker calibration card from which I could determine exposure and white balance.
Exposure was measured from the the card on the gray swatch third from the right. Exposure was adjusted in LR to approx. 50%,50%,50% after white balance correction. The exposure corrections are listed in the table below.
Here’s an example of the target:
And, here are the reults:
You will notice that, other than the V1 w/o diffuser, all flashes were able to expose the target very accurately (to within +/- 0.2 stop). Even the V1 was less than 1-stop out and measured more accurate when the diffuser dome was added. These results suggest that TTL is working fairly consistently between all these flashes when mounted directly on-camera. The flashes ran out of gas around f/11 - f/16 in this configuration (shown in red).
When you remove the flash from the hotshoe and operate it in remote mode, things change a bit.
The bottom part of the table is data for off-camera flash where the flash is off to the left, about the same height as the camera. Not a lot of shift.
You can see that the SB5000 tends to underexpose by about 1-stop. The Godox and Profoto underexpose around 1/2 stop on average. Subsequent tests at different flash angles and positions confirm this general shift in exposure. I have come across this result in other experiments with Nikon flashes where the main subject is under-exposed so this is no surprise. I suppose that nailing the exposure on one swatch of a colorchecker card while light is coming in from different directions, bouncing of different surfaces while relying on the integrity of the pre-flashes is not a situation where you can rely on accuracy. Getting it to within +/- 1 stop is probably as good as it gets. I’m also guessing that Nikon, in particular, is being conservative about exposure so as not to over-expose any parts of the image which may have highlights from the flash. All this is worth noting when using single flashes, on or off camera.
The reason I’m dwelling on this analysis is that the main use for TTL flash is when you have a single flash, on or off camera, and need to get consistent results over a range of images, near & far, light & dark. It would be prohibitive to manually adjust the flash/camera combination each time you wanted to take a shot so TTL is essential. You will also notice that the power of these flashes is pretty limited. Even as close as 6ft to the subject, the min aperture is around f/11. This will likely drop at least 1 - 2 stops if the flash head is tilted, then losing another stop or so if the flash is inside a softbox or bouncing off an umbrella. Pretty soon f/4 is where things are max’d out (at 6ft) so getting really close is important (or up the ISO if the background can stand it)