Limitations
The light meter is absolute—you can count on its readings to mean what they say, regardless of circumstance, since they’re calculated from the camera’s reported brightness value. Once you calibrate Cine Meter II to your reference standard, it should always give you correct readings.
However, the visualization displays—the picture, false-color picture, and waveform monitor—are only relative. They show scene brightness values relative to other levels in the scene, but the levels of those images and waveforms depend on how the camera sets its exposure, which often differs from the brightness value the camera reports to the light meter, as described in How It Works.
For this reason, you can’t directly compare light meter readings with waveform or false-color levels!
You can’t preset an exposure level and then use the displays to show you absolute levels. Cine Meter II doesn’t let you manually preset exposure, so you can’t set the scopes for, say, ISO 800 @ 1/48 sec @ f/4.0. Since the iPhone camera has a tonal response different from that of your production camera—and different models of iPhone use cameras with different responses—there’s no reliable way to use what you see on Cine Meter II’s picture and WFM to predict how your production camera will respond to those same exposure settings.
The only way to set a particular exposure level is to trick the camera by showing it the “correct” light (for example, using a gray card) and then locking its settings. Once locked, you can then look at the pictorial displays to see how tones and colors in a scene render relative to the locked settings.
That’s the key to getting the most out of Cine Meter II’s visualization tools: treat it like a point-and-shoot camera with exposure and white-balance locks (which is really what it is, only with fancier readouts), and you’ll be able to “fool” it into doing what you need.
When Cine Meter II isn’t actively running, your iDevice releases the camera for other apps to use, and when Cine Meter II regains control, the previous exposure settings will have been lost. Therefore, Cine Meter II always unlocks the camera when it launches or wakes up. If you want to keep your locked-in exposure, don’t turn your iDevice off, let it go to sleep, or close Cine Meter II. It’s OK to display the settings pages; just don’t exit the app or let your iDevice shut off.
On iOS 8 or higher, Cine Meter II will remember and restore the previous white balance setting, and transfer it between the front and back cameras. Front and back cameras may show slightly different temperature/tint values for the same locked setting; this is normal.
The waveform monitor scale runs between 0% and 100%; there are no “superblacks” or “superwhites” shown. Cine Meter II gets an 8-bit RGB image from the camera, and the 0%-100% range seemed like the best fit for the camera’s exposure tendencies.
Tips and Tricks
If you want to “dial in” an exposure, shoot a gray card (or your hand, or
something in the scene) while watching the waveform monitor and meter readings. Tilt the
card or shade the subject as needed to get the exposure you desire, then use 
to lock it into place when you get the value you need. Often it’s easier to dial in an
exposure using the spot meter than the matrix meter mode.
Similarly, when you need to preset a white balance, the best way is to use 
while shooting a white or gray card under the desired lighting conditions. Once locked in,
you can then aim the camera at other parts of the scene to judge the color of the lighting.
Starting with version 1.11 on iOS 8 or higher, you can also set the camera’s white balance
to 3200 K or 5600 K, to preview scene color using standard white-balance presets.
I often set an exposure on Cine Meter II by progressively covering the lens with my finger until the exposure opens up to where I want it, or aiming my iDevice more and more towards a light source until the picture gets as dark as I want it.
A gray card is a handy thing for setting exposure and white balance, but
there’s never a gray card around when you need one. In a pinch, use a white card or paper:
set white balance lock on it, then tilt or shade it until exposure drops about two stops
before setting auto exposure lock.
iDevice cameras use fixed-aperture lenses, and vary shutter speed and gain (ISO) to set
exposure. They also record the scene’s “brightness value” in camera metadata, and it’s that
brightness value that Cine Meter II uses to calculate exposure. When you change ISO and
shutter speed settings in Cine Meter II’s light meter, you’re changing the inputs for the
standard APEX equation that calculates the displayed aperture value, but you’re not
affecting how the iDevice camera itself sees the scene.
By the same token, setting Meter Compensation only affects Cine Meter II’s exposure readout; it has no effect on the camera’s own determination of proper exposure, or how its picture is captured or displayed.
When Cine Meter II is running, it’s flogging the camera hardware, the CPU, and the GPU mercilessly, so that you can see images updating in a satisfyingly real-time manner. It will, however, run down your battery in the process, so you’ll probably want to shut it off when you’re not using it.
Just hitting the Home button on your iDevice, or turning it off, will put Cine Meter II to
sleep.
Just be aware that turning off your iDevice or moving Cine Meter II into the background causes the camera to forget any locked settings (as described in limitations), so you’ll either want to keep Cine Meter II awake and displayed if you have locked settings you need to preserve, or be prepared to re-acquire those locked settings when you start Cine Meter II up again.
If you’re worried about battery drain but need to hold your locked settings, tap 
to display the Help & Settings page. The app still “owns” the camera, so your locked
exposure is preserved, but the camera isn't running when the picture isn't shown.
How It Works
iDevices have automatic cameras in them: you can lock and unlock exposure, focus, and white balance, or (starting with iOS 8) set them to specified values, but you can’t control the programming the cameras use to determine their actual exposures, so the image you see and the waveform on the waveform monitor may not correspond exactly to the camera settings and brightness value the camera reports.
The cameras have fixed apertures and vary both gain and shutter speed to control exposure.
The cameras record scene brightness values (Bv) in their EXIF
metadata. It’s that brightness value that Cine Meter II uses to compute its light
meter readings, and (once you’ve calibrated Cine Meter II for the camera in your particular
iDevice) those readings match those of external meters very closely.
So far, so good. However, the camera’s actual exposures do not track Bv
precisely.
When the camera reaches its low-light or bright-light limit, or when exposure is locked, Bv will still change (you can verify this yourself with Cine Meter II’s onscreen data display; tap and hold the camera picture for four seconds to see it). My iPod touch 4G’s camera will report Bv, when locked, to about 2.5 stops of overexposure and roughly 5 stops of underexposure.
Even with the camera unlocked, Bv will vary even when the camera’s exposure
settings don’t. The camera’s internal auto-exposure programming has some hysteresis in it;
it’ll hold a setting until exposure differs by a third of a stop or so before it changes.
It won’t chase minor variations in light level, instead holding a more visually-pleasing
constant exposure as long as possible. Practically speaking, that means the light level can
vary by as much as two-thirds of a stop before you’ll see the camera change its own
exposure settings. But all the while, Bv is changing in response to the scene.
Confusing enough? Wait, there’s more: even when the camera’s reported exposure setting
isn’t changing, the brightness levels in the picture may vary. I’ve seen the waveform
levels suddenly jump 5% on my iPod touch 4G, even though the camera’s metadata swears
that ISO, shutter, and aperture remained constant. My iPad 3’s camera will vary picture
levels considerably, with midtones changing from 45% to 55% in several discrete steps, even
though the camera claims to be fixed at ISO 1600, 1/24 sec, f/2.4. Does the camera have an
additional, undocumented gain stage in it? Does it implement dynamic range optimization or
dynamic contrast control? Whatever the mechanism is, the camera can’t even be trusted to
report how it has come up with its final exposure, not with any consistency and
repeatability.
Because the camera runs its own exposure program and doesn’t give us full and accurate
metadata to relate that back to Bv, you can’t directly compare levels in the
displayed images to levels reported by the light meter. And since the camera’s metadata
doesn’t give us the complete picture (pun intended) of how the camera produces its image,
you can’t even relate levels in the picture to the camera’s own stated exposure.
In short, there’s simply no way to look at the camera’s picture, its false-color version, or the waveform monitor levels, and relate them to any absolute measure of exposure.
Cine Meter II is really two separate instruments in one package: (1) a light meter, and
(2) a picture monitor with fancy display modes—and sadly enough, they don’t always agree.
They’ll usually be within about a stop of each other, but that’s the best you can
hope for.
Observations
The colorimetry of iDevice cameras is not likely to match those of other cameras exactly, especially cine cameras with extended red responses. Cine Meter II’s RGB waveform monitor (WFM) is a very useful quick-and-dirty color-checking tool, but it’s not intended to replace the view through the lens of your production camera.
You may sometimes observe that the camera picture’s exposure varies from what the meter readings indicate. It appears that at least some iDevice cameras have an additional processing stage, separate from the explicit gain and shutter settings: something like a Dynamic Contrast Control or Dynamic Range Optimizer, perhaps. If you move the camera rapidly between light and dark scenes, you’ll often see the exposure reading stabilize within a second, but the waveform and the image brightness may pump up or down over the course of three seconds or so (during which time the camera itself reports no changes in gain or shutter settings!).
iDevice cameras tend to be on the contrasty side, with some S-curving of highlights and shadows, typical of still cameras. Their gamma curves are not matched to Rec.709 broadcast specs, nor to any other cine camera curve. The WFM and False Color displays you see in Cine Meter II are not direct matches for those you’ll see from your production camera (unless, of course, you’re shooting your show on the very same iDevice). That’s why you can adjust the False Color levels yourself: you can set your own trigger levels based on your needs and experience, and compensate for the difference between your iDevice’s camera and your production camera.
You might sometimes see “sticky texture” and fixed-pattern noise in the False
Color view: patterns of red and yellow or blue and purple that remain even as the picture
changes. It appears that some iDevice cameras use recursive (“3D”) noise reduction, and
False Color makes those sticky textures visible.
The False Color view is only reporting what it sees; it gets a fresh frame to work with every time, and has no memory of previous frames. The static patterns are visible in False Color because False Color exaggerates the contrast between the pixels just below a false-color level and those just above it; in normal view these tiny differences (as little as 0.4%) aren’t distinct enough for the static patterning to be visible.
Not all iDevice cameras behave the same way. An iPod touch 4G’s camera shows a
lot of sticky texture and fixed-pattern noise, while an iPad 3’s camera and the iPhone 4S’s
and iPhone 5’s cameras do not.
On the other hand, the lowly iPod touch will auto-white-balance cleanly in dim
tungsten light and under “warm white” fluorescents, while the fancier devices insist on
rendering the same scene as a dirty yellow. Sometimes you win; sometimes you lose...
© 2015–2018 Adam J. Wilt. Last updated 2018.05.24