How to synchronize camera emitter patterns

A shared trigger can align camera exposures without aligning the alternating illumination states attached to those exposures. This distinction matters when you combine active-illumination cameras, compare illuminated and unilluminated images, or subtract a background frame. Two images can represent the same acquisition instant yet belong to opposite emitter states.

To synchronize camera emitter patterns reliably, you need documented control of both exposure timing and sequence phase. The practical task is to measure those separately, preserve frame identity through acquisition, and reject arrangements that depend on a fortunate startup order.

Separate exposure timing from emitter phase

An alternating illumination system has several related but distinct events:

• A trigger arrives at the camera input.
• The sensor begins integrating light.
• The emitter enters its intended state.
• The camera assigns metadata and transfers the image.

A timing connection might control only the first event or the relationship between the first two. It does not automatically reset an internal illumination sequence, select its first state, or expose that state in metadata.

For example, both cameras could expose together while one follows ON, OFF, ON, OFF and the other follows OFF, ON, OFF, ON. Their exposure clocks are aligned, but their pattern phases differ by one acquisition.

There is also a separate optical question: does ON mean that camera's own emitter is active, or that the scene receives light from any emitter? A camera with its own emitter OFF may still see a neighboring camera's projected pattern.

The guide to shared-trigger exposure synchronization explains the exposure-timing layer. The additional requirement here is a defined relationship between each exposure and each illumination state.

Calculate acquisition rate without assuming sequence behavior

A device that produces several acquisitions for each accepted trigger has a different timing model from a device that produces one acquisition. Establish which model applies before calculating rates.

These relationships describe counts and average rates. They do not establish the spacing between acquisitions inside a burst, the order of emitter states, or whether the device accepts every incoming trigger.

Consider a hypothetical documented burst sequence with these application requirements:

• Accepted trigger rate: 25 Hz.
• Acquisitions per trigger: 2.
• First exposure starts at the trigger event.
• Second exposure starts 15 ms later.
• Each exposure lasts 2 ms.

The arithmetic is:

The acquisition intervals alternate between 15 ms and 25 ms. An average rate of 50 frames/s therefore does not imply evenly spaced exposures every 20 ms.

Now suppose both cameras expose at the same scheduled times. Camera A labels the first acquisition ON and the second OFF; camera B does the reverse. The timing arithmetic remains correct, but the illumination pairing is wrong.

Do not infer that configuring a nominal frame rate automatically establishes this burst behavior. Confirm supported trigger rates, burst completion time, readout constraints, and behavior when another trigger arrives too soon. The explanation of external pulses and camera frame rate covers the distinction between incoming pulses and accepted acquisitions.

Verify three layers independently

Use a staged test rather than judging synchronization from two paused preview windows. Displays can present buffered images from different acquisition events.

First, verify exposure timing

Record the common timing event and each camera's documented exposure-active output, where available. Check the output's definition: a signal describing readout or frame transfer is not necessarily an exposure-active signal.

Measure the relative exposure-start offset over a representative run. Repeat after startup and after an acquisition restart. If electrical timing outputs are unavailable, a suitably controlled optical timing target can provide evidence, but rolling-shutter effects and measurement resolution must be considered.

The guide to verifying multi-camera exposure timing describes why matching displayed images is weaker evidence than measuring the acquisition event.

Second, verify illumination state

Prefer documented emitter-state metadata tied to each acquisition. Check whether the field describes a commanded state or a confirmed optical event.

An optical measurement can provide an independent check. A suitable photodetector and measurement instrument can show when light actually reaches the monitored location. Detector bandwidth and wavelength sensitivity must suit the illumination being measured.

When measuring each source separately, arrange the test so that neighboring illumination cannot masquerade as that source's output. Follow the manufacturer's operating instructions throughout. Do not disconnect internal emitter wiring or modify protective functions to obtain a measurement.

Image brightness alone is not a dependable state indicator. Scene reflectance, automatic exposure, saturation, ambient light, and another emitter can all change brightness without changing the camera's own emitter state.

Third, verify frame identity

Log a record for every image containing the available acquisition timestamp, frame counter, trigger-event identifier, sequence position, and emitter-state field. Mark unavailable fields explicitly rather than inventing replacements.

Different cameras may use different timestamp origins. Even within one device, a timestamp might describe trigger receipt, exposure start, or another event. Confirm its meaning before comparing values.

Never assume that two images in a burst share a timestamp, or that equal timestamps guarantee equal emitter states. The distinction between trigger calls and exposure time is especially relevant when host software initiates acquisition.

Interpret the evidence before changing settings

Use the observed combination of timing, optical state, and metadata to choose the next test.

A missing image does not necessarily mean the camera's physical sequence changed. For example, losing one image from an alternating stream can reverse the apparent parity of every later image if software labels states from received-image order.

Maintain acquisition-based identity instead. If identity becomes ambiguous, reject the affected pair and re-establish a documented sequence boundary. Do not silently relabel later images to make the pattern appear correct.

Choose a supported illumination architecture

There are several legitimate ways forward, but each solves a different problem.

Use documented emitter-sequence synchronization

This is the closest match when the application requires each camera's integrated emitter. The documentation must specify more than a trigger mode name.

Look for an explicit definition of the initial emitter state, state order, sequence reset, restart behavior, and the relationship between trigger events and acquisitions. Check supported combinations of hardware, firmware, and acquisition software using manufacturer-provided documentation and downloads.

The guide to shared timing events for multiple depth cameras addresses the common timing foundation. Emitter-phase guarantees remain an additional requirement.

Drive supported camera sequences from an external timing source

An external timing source can provide a common reference when every camera documents the necessary trigger and sequence behavior. It cannot force an undocumented internal state machine to follow the desired phase.

Verify electrical levels, polarity, pulse duration, accepted event rate, and readiness requirements before connecting equipment. Establish how every camera recovers after a missed event. A controller that generates accurate pulses does not, by itself, establish sequence-state agreement.

Use separate illumination for an independent inspection channel

For ordinary intensity imaging, external illumination may let you place cameras and lights under a common documented timing design. This can be useful when the task is appearance inspection rather than depth reconstruction.

However, an external floodlight is not a substitute for a calibrated structured-light projector. It may improve brightness without providing the spatial pattern required by the depth algorithm. Confirm that disabling integrated illumination is supported and that the remaining imaging method can still perform the task.

Separate emitters in time

If simultaneous active illumination causes interference, non-overlapping illumination slots may be preferable to matching ON states. Each camera can then acquire during an assigned slot, provided the equipment supports the timing arrangement.

This changes the acquisition time between views. Evaluate motion and registration errors before accepting it. Synchronized emitters and interference-free measurements are not necessarily the same objective.

Parts for a separate area-scan inspection channel

The following catalog parts illustrate a separate intensity-imaging architecture, not a solution for controlling an integrated depth-camera emitter. They are candidates for evaluation, not a verified matched kit.

The area scan cameras category and camera finder help narrow sensor and interface choices. The MV-CU020-90GC is a global-shutter color camera with a GigE interface. Confirm its documented trigger and exposure-output capabilities before selecting it for the timing design.

The light controllers category and controller finder cover illumination-control options. The LS-CTL-0070 lists a strobe feature and constant voltage 24 V output. Its catalog listing alone does not establish pulse timing, trigger compatibility, or suitability for a particular light.

The special lights category and lighting finder include segmented and sequencing illumination options. The LS-SPC-0086 is listed as a multi-channel sequencing light source with 8 zones and 24 V operation. Verify zone access, sequencing controls, and the required driver arrangement from its documentation.

Matching nominal voltage is only one compatibility check. Also confirm current demand, channel arrangement, connector wiring, permitted operating modes, and the division of timing responsibility between camera, controller, and light.

For a supported external lighting design, documented duty-cycle limits and measured optical rise time matter as well as trigger alignment. Neither a strobe label nor a sequencing label establishes compatibility with an integrated camera emitter.

Commission against explicit acceptance criteria

Define success before testing. A commissioning record should state the required exposure offset, allowed illumination transitions, pairing rules, and recovery behavior.

For illustration, an application might require exposure-start mismatch below 100 microseconds, correct emitter-state pairing on every accepted pair, and explicit invalidation after a missing acquisition. These are application requirements, not specifications for the catalog parts above.

Use this checklist:

• Document the sequence. Record the expected state and acquisition count for every timing event.
• Lock the test conditions. Use documented manual exposure and gain settings where supported, so automatic adjustments do not obscure timing evidence.
• Measure optical overlap. Confirm that useful illumination falls inside the required integration interval, including startup and shutdown transitions.
• Test normal restarts. Repeat the supported startup sequence and acquisition stop/start procedure.
• Test interruption recovery. Check documented behavior after an acquisition interruption, without altering machine safety functions.
• Detect missing images. Confirm that a delivery failure cannot silently shift ON/OFF labels.
• Record configuration. Save supported settings, firmware versions, timestamp definitions, and measurement evidence.
• Define rejection behavior. Stop using ambiguous pairs until the documented synchronization state is restored.

A short test with no failures is not a guarantee of indefinite operation. Choose test duration and restart coverage according to the consequences of an incorrectly paired measurement.

When matching ON/OFF states is the wrong target

Matching emitter states is useful only if it improves the measurement. Simultaneous ON states may increase optical interference, while simultaneous OFF states may remove the signal needed for depth estimation.

Similarly, subtracting an OFF image from an ON image assumes that scene appearance remains sufficiently stable between acquisitions. Moving parts, changing exposure settings, or illumination from another camera can violate that assumption even when sequence phase is correct.

Repeatedly restarting cameras until their images appear aligned is not a production synchronization strategy. It neither proves the initial state nor defines recovery after an interruption. Changing undocumented firmware behavior or directly switching internal emitter power is not an acceptable substitute for supported timing control.

Start with a written timing requirement and ask for a documented architecture that meets it. Use the controller configuration step for a separate lighting channel, or contact applications support with your exposure diagram, emitter-state requirements, and acquisition logs before selecting parts.