The Camera and Imaging Products Association (CIPA) an industry origination for digital camera makers produced a standard for digital still camera (DSC) battery consumption testing. This standard is the familiar “number of shots” standard that we see in camera marketing literature, reviews, and camera specifications.
This standard was published in 2003 when the digital camera landscape was radically different than it currently is. As digital camera technology has advanced and markets have changed this 2003 standard either doesn’t include useful, and arguably necessary, metrics or simply has become otherwise outdated.
This RFC presents an expanded and revised set of power consumption tests that aim to provide a more comprehensive an useful look at the battery life that modern digital camera users should expect under various conditions.
This document itself is intended to serve as the start for a broader discussion (hence the RFC label), as well as an outline for testing that camera reviewers can begin to do now.
When shopping for, or reading reviews, of a new camera, we’ve all seen the estimates for the number of possible shots in the marketing literature. Almost always there’s note indicating that this was done according to the CIPA standards for battery consumption.
The CIPA is an industry organization that’s made up of a number of camera manufactures and other photography and imaging companies. They create standards and provide guidance as well as collate statistics for their members.
CIPA DC-002: The Current Standard for Battery Consumption Testing
The CIPA DC-002 standard is the standard that is currently used by most digital camera makers for battery consumption testing. An English translation can be found here.
Reading and understanding DC-002 is at least partially necessary to understand this RFC, as this RFC builds on DC-002 in many places.
DC-002 establishes pre conditions, a test methodology, and end-of-tech criteria for testing a camera’s battery life in terms of the number of frames that can be made per fully charged battery.
The CIPA standard was published in 2003, when the digital camera camera landscape was substantially different form today. When the standard was finalized, it would still be 4 years until the release of the original iPhone and the idea of people using phones as their primary camera wasn’t yet a real consideration.
In the years since the standard was release, a number of notable changes have happened to the digital camera landscape.
To start with, mobile phones, especially smartphones, have advanced substantially in their imaging capabilities. These devices, instead of dedicated low-end cameras, have become the mainstream way for people to take pictures.
A direct result of this shift in technology, markets have shifted. Over time, standalone digital camera sales has shifted towards higher end, or at least more serious or sophisticated users. One consequence of this shift is a change in what constitutes the average camera, and a subsequently what the average use age patterns will be.
For example, more sophisticated users are less likely to leave their camera’s flash on auto (when one is even built in) where it may be activated in an appreciable number of their images. Instead they are more likely to use a dedicated external flash unit in their photography (if their camera supports one), or at least control the use of the flash to their photographic needs. Consequently, the standard’s 50% of frames being shot with a full power flash may no longer (if it ever was) be an accurate model of use.
Another major shift in digital camera technology is the increased prevalence of video recording being done with what previously would have been a still photography only camera. The CIPA battery consumption standard makes no provision for testing video recording times for what it describes as digital still cameras (DSCs). Moreover, there doesn’t appear to be a current CIPA standard for this, or for that matter any other standard that I could find.
Finally, looking toward the future, at the high end the industry is moving towards mirrorless interchangeable lens cameras instead of, or in addition to, DSLRs. These cameras replace the traditional optical viewfinder with a live “video” stream from the sensor to a display in the electronic viewfinder (a process that’s similar to what compact digital cameras have always done).
The most notable consequence of the move from optical to electronic viewfinders is that cameras now consume power at all times when the viewfinder is active, and not just when its recording an image.
For example, a wildlife photographer using a mirrorless ILC and a super telephoto lens as a spotting scope is consuming power even though they’re not making images, where a DSLR would not. For such a photographer, especially one working in the wilderness away from easy charging, understanding the rate at which they’ll consume power doing this may be invaluable.
On top of the above issues, I believe the CIPA standard has a few deficiencies.
One overreaching issue is that it attempts to compare all cameras in one standardized way. Though there is certainly utility in the having an industry wide baseline, in practice it’s unlikely that people will compare cameras across widely separate tiers (such as comparing a full size pro DSLR to a compact fixed lens digital camera) in this way.
Moreover, because of the design of the standard’s test procedure, there are inconsistences between how the standard simulates usage versus how the camera would be used in actual market segments. This discrepancy can have significant and hard to predict deviations from the standard.
As one example of this, consider this example using an EOS-1D mark III professional DSLR that I previously owned. Under the CIPA standard, this camera is rated at 2200 shots at 73°F (23°C). In practical use, shooting an airshow with the camera doing servo tracking AF, running a stabilized lens and shooting at 10 FPS, I shot more than 4500 frames on only 55% of the camera’s battery. Extrapolating to the full battery under those conditions, I could have expected nearly 8,200 frames on a charge; more than 3.7 times the CIPA standard testing suggested.
Finally, history has demonstrated that cameras and batteries are not a static systems. Both Canon and Nikon have updated battery designs over time from the original batteries that were tested with the camera, while still maintaining compatibility.
For example Canon’s EOS-1D series of cameras have maintained a compatible battery format form the 2007 vintage EOS-1D mark III and EOS-1Ds mark III, all the way through to the current EOS-1D X. Over this time, the batteries have gone from the LP-E4 at 11.1 V, 2300 mAh, 25.53 wh. To the LP-E4N at 11.1 V, 2450 mAh, 27.20 wH. To finally the LP-E19 at 10.8 V, 2700 mAh, 29.16 wH. An increase of more than 14% in capacity.
While it would seem like a reasonable assumption that battery life should scale proportionally to the battery’s capacity. Changes in battery chemistry will change the way a battery reacts under load.
For example, Canon’s 1DX mark II when paired with the older LP-E4 series batteries is limited to only 12 FPS (form it’s peak of 16 FPS), due to the batteries chemistry and the required discharge current.
It’s my opinion that a standard such as this should serve two purposes:
- Provide a standardized mechanism for users to compare cameras across brands and models.
- Provide end users useful information to enable them to predict power consumption for their usage.
Even though I’ve argued that different classes of users and different tiers of cameras are going to be used differently, there is still value in providing a standardized metric that can be compared across all brands. However, I will contend that such a cross brand and platform comparison tests need to be done in such a way that it puts all cameras on as level of a playing field as possible. In an effort to do everything in one test, this is something the CIPA standard doesn’t do.
Secondly, consumers should be able to make useful predictive decisions from the information that’s provided. In practice I’ve found that CIPA standard testing provides numbers that can vary wildly from the actual use I get out of a camera — with errors varying wildly in both directions.
This kind of variability makes it hard to make any kind of predictive assessments, even with gear that I’m familiar with. Instead, it almost always becomes a case of just taking all the batteries, just to make sure I don’t run out of power.
Providing useful predictive information means using either an accurate workload simulation, or testing specific usages that shows consumption under sufficiently narrow but useful conditions that a consumer could use them as a basis for a more accurate analysis.
Finally, it’s my belief that wherever possible, standards like this should be able to be independently verifiable. This places 2 basic requirements on the standard as a whole:
- The standard must be sufficiently simple to understand and deal with that interested end users and reviewers can preform the same test procedure and verify the manufacturer provided values.
- The standard must avoid the use of expensive, complicated, and especially proprietary test equipment.
To satisfy these conditions, all tests should be capable of being carried out using only the to be tested camera, a battery, and the user interfaces provided by the camera.
Alternative, in the that external testing equipment should be used, this equipment should not be so onerously expensive that an average hobbyist can afford to buy or produce the equipment. E.g., a $50,000 optical test bench is an unreasonable demand, but a $200 oscilloscope probably isn’t.
Any good testing standard should be based on sound scientific principals. Principally tests should be designed to control as many variables as possible to eliminate uncontrolled externalities. However, it’s also recognized that with a scope as wide as all digital cameras, there is a very large degree of variability.
Interchangeable Lens Variability
Interchangeable lens cameras create problems for testing camera power consumption, as they are parasitic power sinks. Specifically lens functions, such as autofocus motors and image stabilization servos, are powered by the camera’s battery. However, these loads are not uniform or well defined. E.g., one lens’s AF motor may require more power to drive than another.
Moreover, because the lenses are interchangeable the power consumed by the lens can range from none, in the case of a fully manual lens, to potentially as much as the camera is capable of providing.
Finally, because all interchangeable lenses aren’t compatible across all platforms, there is no one lens that can be adopted as the standard test lens for all tests.
Unfortunately, the CIPA standard provides no specific guidance on how ILCs should be tested or what lens settings should be used. The closest it comes to offering specific guidance is the following quote, “By specifying the standard for high power-consuming functions…zoom and retracing lens movement.”
Moreover, while the CIPA standard states that AF behavior should be left to the factory default settings, which almost always are enabled and a one shot type mode. The power consumption form a lens focusing on an already in focus target is going to be different than the power used to focus at different points in the focus range.
There are clearly two lines of thought here. On one hand the lens is a parasitic power draw from the camera’s battery, and therefore things like focusing and image stabilization consume power and therefore reduce the shot count.
The other is that because the lens on an ILC can be changed, and with it the lens’s power usage, and there is no practical standard candle that can be used on all platforms, then the lens’s power consumption should be mitigated as a contributing factor.
Since a standard like this should produce a conservative value that underestimates real world performance. I’m inclined to suggest that AF and in lens stabilization should be left on for ILCs and a lens that’s representative of a the average lens used for that platform should be tested. In the case of cameras sold as kits (i.e., a body and a lens) the kit lens would be a reasonably lens to choose for testing that camera model.
Fundamentally all the measurements that are being made in this test suite are a measure of watts. That is the power used to complete some task.
However, for a publicly presented result the output format should be formatted in a way that the typical user can understand. Saying, “The camera uses an average 20 mW per frame,” is not suitable for most consumers.
As such presentation should be in a form of per battery; though with the per battery assumed. E.g., 10 hours, 20 frames, etc.
The battery used should be the model of battery that ships with the camera.
Form this point forward the key words “MUST”, “MUST NOT”, “REQUIRED”, “SHALL”, “SHALL NOT”, “SHOULD”, “SHOULD NOT”, “RECOMMENDED”, “MAY”, and “OPTIONAL” in this document are to be interpreted as described in RFC 2119.
This RFC builds on the DC-002 test conditions. The following section outlines conditions that deviate from those prescribed by DC-002. If a condition is not specified here, it should be set in accordance to DC-002, or the recommendations of DC-002.
For cameras that use primary (non-rechargeable) batteries the batteries used MUST be standard cells. Battery examples include alkaline AA cells.
For cameras that use standard sized secondary (rechargeable) batteries, the capacity of the rechargeable battery MUST be specified in the test results. Battery examples include NiMH AA cells or standard sized (e.g., 18650) Li-Ion cells.
For cameras that use a proprietary secondary (rechargeable) battery, the battery used MUST be the same model that ships with the camera.
Interchangeable Lens Cameras
Since lens power will vary from model to model and brand to brand, the lens used MUST be recorded and included in the test report.
The specific lens to be used is left to the discretion of the tester, however the following guidelines SHOULD be followed.
- The lens SHOULD provide the widest range of power sinking features that are available on the platform (e.g., autofocus and image stabilization).
- The lens SHOULD be representative of the typical lenses that would be used on the camera. E.g., while a manufacturer may offer a 400mm f/2.8 prime that’s compatible with their entry level ILC platform, it’s much more likely that a normal range zoom (e.g., 18-55mm for APS-C, or 24-70mm or 24-105mm for full frame) will be more representative of the lenses that are typically used by most users.
During testing, the lens’s autofocus and image stabilization modes MUST be enabled.
Low Temperature Environment
Low temperature testing is optional. It’s intended to reflect the performance of the battery system under cold conditions such as winter climates.
Low temperature testing SHOULD be done for professional and high end interchangeable lens cameras, even though it’s optional.
If low temperature testing is done, measurements MUST be conducted at 0°C ± 2°C, with a relative humidity of 50% ± 20%.
Alternate Non-Battery Power Sources
Some tests described in this document provide for an alternate test procedure that measures current, voltage, and energy used via an external DC power source. These procedures are provided for tests that may have unreasonably long run times with indications provided by the camera.
If these procedures are used, the tester MUST note that used these procedures in the test results.
Exact external instrumentation and the connection configuration is up to the discretion of the tester. External instrumentation configurations MUST be documented by the tester and made available with their results.
If using external power monitoring instrumentation, the tester MUST measure the power on the DC side of the power supply between the power supply and the camera’s battery terminals. It is not acceptable to plug a external power adapter (e.g, Canon ACK-E6, or Nikon EH-5B + EP-5C) into a kill-a-watt style power meter and report those values.
This document proposes 4 power consumption tests; 1 replicates the DC-002 standard, and 3 are new.
- Legacy Battery Consumption Test (replicate DC-002)
- Baseline Cross Platform Battery Consumption Test
- EVF Observation Battery Consumption Test
- Video Recording Battery Consumption Test
Legacy Battery Consumption Test
The legacy battery consumption test is defined in CIPA DC-002. This test is retained to provide an uninterrupted line of comparison to cameras tested under the previous standard.
See CIPA DC-002.
Baseline Cross Platform Battery Consumption Test
The intent of this test is to provide a more level comparative point between cameras. Its built on the same methodology as the Legacy Battery Consumption Test, but the flash is powered off on all images for all cameras that supports disabling it.
The DC-002 standard that the Legacy Battery Consumption Test directly inherits, intends for the use of all power consuming aspects of the camera such as the extension and retraction of lenses on power up and own, the operation of a camera powered zoom, and the use of the built in flash, among others. However, of the identified power consumers, the flash is most often photographically dictated, meaning it’s not an operational requirement of the camera. That is, a lens that the camera collapses for storage must be extended to even use the camera, and a powered zoom must be actuated to change the lens’s focal length. The flash, however, is a compositionally specific element that’s use, out side of purely auto operation, is dictated by the scene and the photographer’s needs.
By removing the flash from the Test 0 test case, it provides a number that can be directly compared to cameras without flashes.
See CIPA DC-002, with the modification that the camera’s built in flash MUST be turned off or disabled for the entirety of the testing procedure.
Exceptions: For cameras that cannot have their internal flash disabled, the use of the flash is no longer considered a photographic prerogative and is considered to be an operational aspect of the camera. For these cameras this test may be ignored as the results will be identical to the Legacy Better Consumption Test
Electronic Viewfinder Observation Power Consumption Test
This test only applies to electronic viewfinder cameras.
Many cameras today (both compact, integral-lens, cameras and high-end, interchangeable lens, cameras) utilize electronic viewfinders (EVFs) instead of traditional optical viewfinders (OVFs).
Unlike an OVF, for an EVF to function (in other words for the photographer to view the scene through the lens) the camera must be powered up and operating. This includes:
- The sensor being energized and read out at a sufficiently high rate that the image in the viewfinder is smooth
- The onboard processor preforming image processing to convert the sensor data into a usable video signal for the EVF
- The generation of a visual signal in the EVF (e.g. powering a backlight and LCD matrix, or powering an OLED matrix)
- Periodic ancillary functions that must run for the camera to remain in a state where the EVF is active (e.g., autofocus, image stabilization, and metering computations as a result of half-pressing the shutter release to keep the camera from sleeping due to inactivity)
The ultimate intent of this test is to provide photographers with an estimate of the battery life that will be consumed using their EVF based camera for prolonged periods as an observational tool (e.g., a replacement for binoculars or spotting scopes) before or between making images.
Example Use Cases: An example of this kind of use is a wildlife photographer waiting in a condition that they’re prepared to shoot immediately as their subject matter makes a dramatic move. E.g., photographing a bird taking off from a perch or a predator initialing a pursuit. They cannot use a secondary means of observation (such as binoculars) as the delay in switching to the camera would result in missing the critical action.
Two procedures are provided for this test: the first relies only on the indications provided by the camera itself, and a second procedure that uses an external electrical sensors to determine power consumption.
Procedure 1: Self Contained Testing
The time based test methodology is intended to provide any user the ability to make these measurements without any external support equipment.
The test timer starts when the camera is brought up to a state where the EVF is active and a scene can be observed though it.
For the duration of the test, the the camera is to be kept in a state where the EVF is active. This is to be done using the default configuration settings and based on the capabilities of the camera.
- If the cameras use a facial proximity sensor to detect presence and activity and this sensor can, under the default settings, keep the EVF active for the duration of the test, this may be used.
- If the camera requires a manual user input, such as half-pressing the shutter release, on a periodic schedule this may be done at the longest possible interval the camera’s settings allow.
- If the camera can be kept awake by the actuation of a button other than the shutter release, this button may be used if:
- It’s available under the default configuration.
- The manual clearly calls out that pressing this button may be used to keep the camera awake and it’s benefits or limitations (e.g., doesn’t activate the AF system).
- If pressing this button uses less power (e.g., doesn’t activate the AF system) than it may be used instead of the shutter release if it can be accessed easily with the photographers hands in the normal, natural, shooting position
The end condition occurs when the battery indicator has changed from the initial full charge indication to the next lower state.
E.g., On a Canon 5D mark IV, the manual documents the battery indicator as having 6 states as shown in the table below.
|100 – 70 %||Solid, outline with 4 internal bars|
|69 – 50 %||Solid, outline with 3 internal bars|
|49 – 20 %||Solid, outline with 2 internal bars|
|19 – 10 %||Solid, outline with 1 internal bar|
|9 – 1 %||Blinking, outline with 1 internal bar|
|0 %||Blinking, outline only|
Since the test starts with a fully charged battery, the test concludes when the battery indicator drops to the second state in the table and the test’s run time is recorded.
Procedure 2: Power Based Measurements
Camera manufactures, or testers, may wish to generate necessary data more quickly (though potentially less accurately) than waiting minutes or potentially hours for the battery to be discharged as prescribed in Procedure 1 above. Since power draw can be measured directly, and reasonably extrapolated to battery capacity this can be done to accelerate testing.
This method is intended to provide similar results as the previous method but based on direct measurements of the power consumed, instead of trying to calculate the energy use.
This test follows the same operational procedure as the time based test. However the test can be conducted over a fixed time period (suggested 10 minutes) with the camera hooked up to power monitoring equipment.
The average energy used (in watts), directly taken form the measurements, is noted and converted to the final published value.
Use of this methodology MUST be noted in the test results.
The results should be converted to minute per battery based on basic extrapolation.
LCD Observation Test (Optional)
As a companion to the EVF observation test an LCD base test may also be optionally provided. This test uses the same methodology as the EVF test, however it uses the large LCD display instead of the EVF.
This test was marked as optional as the use case is not completely clear. While it would potentially be of use to some DSLR users, it closely overlaps the LCD video recording test. Though a recording would require more power, using the rear display recording would provide a more conservative value to the consumer using this for planning purposes. Finally, while some compact cameras only have a rear LCD, it’s unlike that they would be used for long term observation of something in the manner this test is intending to replicate.
See the Electronic Viewfinder Observation Power Consumption Test, with the alteration that the large rear LCD is used instead of the EVF.
Internal Video Recording
The intent of this test is produce the most conservative run time for video recording operations.
For this test the most power intensive recording bitrate, image size, and frame rate MUST be used.
Note to testers: This will likely be the large size recording at the highest frame rate with the highest bit rate settings. However, in cameras, such as the Canon EOS 1DX mark II, where 4K video is recorded in MJPEG, and 1080p video is recorded in h.264, it may be necessary to test multiple modes to determine the most power demanding setting. Only the results from the most demanding setting should be reported.
This test follows the overall standards established in DC-002 with the following exceptions.
- Video recording MUST be done.
- Audio recording MUST be done using the camera’s internal microphone, if available, using the factory default settings.
- The video resolution, frame rate, and bitrate MUST be recorded and provided, and the combined video settings MUST be the highest power options provided by the camera.
- For cameras with interchangeable lenses, autofocus and in-lens image stabilization MUST be disabled.
- In body image stabilization MUST be set to the factory default state.
The test starts when the record button is pressed and the camera starts recording.
Many cameras primarily aimed at still photography but offer video recording features have limited recording times. When testing cameras that have limited recording times, the tester MUST restart the recording as soon as the camera allows a new recording to commence.
The test end criteria is when the camera powers off or is no longer able to record video.
Example Report Formats
To be added.
Arguably there’s room to expand the testing to cover more and more edge cases. However, at some point expanding the test cases becomes counter productive as the sheer volume of numbers becomes overwhelming.
With the exception of the Baseline Cross Platform Battery Consumption Test and the Electronic Viewfinder Observation Power Consumption LCD Observation Test these tests are already being done by manufactures in some form or another. In this case, the video recording tests simply aim to standardize the video recording testing already being done to a comparable worst case standard.
- Initial Publication: 2018-08-24
For the sake of all tests the standard still photography shooting position is defined as follows:
- The right hand should grip the camera by the camera’s grip
- The right index finger is to be located near or on the shutter release.
- The right thumb shoudl be located on the camera’s back side to the right of the viewfinder and LCD in the place where it would naturally fall.
- The left hand supports the lens from below at:
- The position of the zoom ring if the lens is a zoom
- The position of the foucs ring if the lens is not a zoom.