I have two cameras that I've put some money and love into to get them working really well: an Olympus OM-1 and a Canon FTb. Both have recent CLAs, fresh seals, known-accurate shutter timings, sexy aftermarket leatherettes, and a fair amount of sentimental value. Unfortunately, both also have built-in meters that want PX645 1.35v mercury batteries. Lately I've been digging deep into the many and varied solutions to this problem:
One video I watched suggested buying a Neewer LED panel (a video light basically) that had a relatively high CRI (97+), and fine control over both color temperature and brightness. Then set 5500k to match daylight conditions, and use a trusted light meter to find the right brightness setting for EV12. That can then be the baseline to which the camera meter is zeroed (filling the entire frame with the light source). Repeat the exercise as desired at other EVs.
When trying this approach with recommended video light and fresh zinc air batteries (good voltage confirmed with my multimeter at the time of testing), I found that my OM-1 meter's readings were off by roughly a stop compared to my trusted spotmeter. My FTb meter's readings were off by about 2/3 stop, but in the opposite direction.
But here's where it gets frustrating: if I do this same test with a fairly dark, but evenly-lit indoor white wall (EV9, and what light is falling on it is primarly warm LED household bulbs at around ~3000k), then the error between the trusted spot meter and the two camera meters changes by more than 1/3 of a stop for each camera. And if I go outside on my shady front porch and try again on evenly-lit concrete, I get a third set of deltas that doesn't match either of the first two.
The majority of old forum threads, Reddit posts, etc. I have read about 35mm camera built-in meter accuracy seem to have an electrician or two chiming in at some point with words to the effect that "these old meter circuits are pretty crude instruments to begin with, I'd just try to get it more or less within 1/3 of a stop or so and call it good, film is forgiving." And that seems to be where the discussion ends. There are so many possible confounding factors (meter spectral sensitivity compared to film spectral sensitivity, age and condition of CdS cells, differences in nominal vs actual voltage, etc. etc.) that it's starting to feel impossible to get a perfect meter reading from a built-in meter.
And by "perfect" I really mean something very pragmatic: I just want to be able to trust that I can get an exactly correct exposure on slide film every time. An error of 0.1EV is no big deal. But an error of 1/3 of a stop? That could matter, given the right scene. Even as someone who primarily photographs nature scenes in diffuse lighting.
I feel like I'm chasing my tail here. Is my best bet to just pick one of the options for a known-brightness calibration target, zero the camera meter to that, and call it good? Or is there something I can do that would a) result in more consistent and accurate exposures and b) not require me to spend hundreds or thousands of dollars on old calibration equipment?
- Use a 1.55v silver oxide battery and accept the metering errors and nonlinearity. Non starter for me. I want to shoot slide film.
- Use a 1.55v silver oxide battery and offset your ASA dial to try and compensate for the excess voltage. Nope - causes nonlinear readings at the high and low ends of the EV scale.
- Use 1.4v, 675-size zinc air batteries with an obnoxiously short activated lifespan. This is my current solution, and I'm unhappy with how many batteries I go through and how often a battery is dead when I've had the camera on the shelf for a few weeks.
- Adjust the meter (or pay a professional to do so) using the calibration pots or equivalent so that it's accurate with 1.55v input. Apparently doesn't always work (some OM-1s and FTbs have enough "room" to be calibrated to 1.55v this way, some don't), and I never could get a straight answer on whether this will still result in linear readings at low/high EVs.
- Shell out $40 USD for a genuine MR-9 adapter that knocks SR44 voltage down to 1.35. Potentially open to this.
- Solder a diode into the meter circuit to accomplish the same thing as the MR-9. This is where I'm leaning. Schottky diodes are super inexpensive, and I foresee no need to ever move away from 1.55v silver oxide batteries for these cameras.
One video I watched suggested buying a Neewer LED panel (a video light basically) that had a relatively high CRI (97+), and fine control over both color temperature and brightness. Then set 5500k to match daylight conditions, and use a trusted light meter to find the right brightness setting for EV12. That can then be the baseline to which the camera meter is zeroed (filling the entire frame with the light source). Repeat the exercise as desired at other EVs.
When trying this approach with recommended video light and fresh zinc air batteries (good voltage confirmed with my multimeter at the time of testing), I found that my OM-1 meter's readings were off by roughly a stop compared to my trusted spotmeter. My FTb meter's readings were off by about 2/3 stop, but in the opposite direction.
But here's where it gets frustrating: if I do this same test with a fairly dark, but evenly-lit indoor white wall (EV9, and what light is falling on it is primarly warm LED household bulbs at around ~3000k), then the error between the trusted spot meter and the two camera meters changes by more than 1/3 of a stop for each camera. And if I go outside on my shady front porch and try again on evenly-lit concrete, I get a third set of deltas that doesn't match either of the first two.
The majority of old forum threads, Reddit posts, etc. I have read about 35mm camera built-in meter accuracy seem to have an electrician or two chiming in at some point with words to the effect that "these old meter circuits are pretty crude instruments to begin with, I'd just try to get it more or less within 1/3 of a stop or so and call it good, film is forgiving." And that seems to be where the discussion ends. There are so many possible confounding factors (meter spectral sensitivity compared to film spectral sensitivity, age and condition of CdS cells, differences in nominal vs actual voltage, etc. etc.) that it's starting to feel impossible to get a perfect meter reading from a built-in meter.
And by "perfect" I really mean something very pragmatic: I just want to be able to trust that I can get an exactly correct exposure on slide film every time. An error of 0.1EV is no big deal. But an error of 1/3 of a stop? That could matter, given the right scene. Even as someone who primarily photographs nature scenes in diffuse lighting.
I feel like I'm chasing my tail here. Is my best bet to just pick one of the options for a known-brightness calibration target, zero the camera meter to that, and call it good? Or is there something I can do that would a) result in more consistent and accurate exposures and b) not require me to spend hundreds or thousands of dollars on old calibration equipment?
