LED Light Bar PWM Dimming What It Means for Camera Flicker and Banding

Nov 20, 2026 - 09:47
 0  114
LED Light Bar PWM Dimming What It Means for Camera Flicker and Banding

Shenzhen Aurora Technology Limited is a Shenzhen-based LED lighting manufacturer established in 2011, operating a 35,000 sqm factory with IATF 16949 certification, supplying 4WDKING-branded off-road and work lighting for OEM and private-label programmes.

Introduction

Shenzhen Aurora Technology Limited manufactures LED light bars, and LED light bar PWM dimming switches those emitters on and off rapidly. A camera with a short exposure or a rolling shutter records only slices of each cycle. The result is uneven frames or striped images. These artifacts reflect a sampling mismatch, not a light bar defect. PWM frequency, exposure time, and shutter architecture make the outcome predictable and correctable.

Key Takeaways

· Raise the PWM frequency above the camera's exposure threshold to reduce flicker and banding.

· Lengthen the camera exposure time to span multiple PWM cycles for stable brightness.

· Synchronize the light's clock with the camera trigger to eliminate artifacts in fixed setups.

How LED Light Bar PWM Dimming Produces Flicker and Banding

How LED Light Bar PWM Dimming Produces Flicker and Banding

Shenzhen Aurora Technology Limited builds light bars around switching drivers, and the behavior of those drivers explains most camera artifacts. LED light bar PWM dimming controls brightness by chopping a constant current into rapid on-off pulses. A camera never sees the full pulse train. It sees a series of samples, and the mismatch between pulse timing and sample timing creates the visible problem.

 

PWM in Brief: Switching, Duty Cycle, Frequency

A PWM driver holds the LED current at a fixed level and switches it on and off at a set rate. Three quantities describe the waveform completely. The switching frequency sets how many on-off cycles occur each second. The duty cycle sets the fraction of each cycle that the LED stays on. The amplitude sets the current during the on portion.

Duty cycle maps directly to perceived brightness. A 50 percent duty cycle delivers half the average optical power of a full-on drive. A 10 percent duty cycle delivers one tenth. The human eye integrates these pulses above roughly 100 Hz and reads them as a steady, dimmed output. The eye is a slow, forgiving sensor. A camera is not.

Frequency and duty cycle interact in a way that matters for imaging. At a fixed frequency, lowering the duty cycle shortens the on-time window within each cycle. A 1 kHz signal with a 10 percent duty cycle keeps the LED on for 100 microseconds per cycle. That window is short enough to fall entirely between camera samples. The same 1 kHz signal at 90 percent duty cycle keeps the LED on for 900 microseconds, and most exposures will catch at least part of the on-time. Dimming therefore changes artifact risk even when the frequency stays constant.

Designers also face a tradeoff between frequency and efficiency. Higher switching frequencies push the ripple out of the visible artifact range. They also raise switching losses in the driver and can increase electromagnetic interference. Many light bars settle in the low-kilohertz region as a compromise. That region sits squarely inside the exposure windows of common cameras, which is why LED light bar PWM dimming so often shows up in footage.

What the Camera Actually Samples

A camera does not record a continuous image. It accumulates light over an exposure window and then reads out the sensor. Each pixel reports the total light energy it collected during that window. The shutter architecture determines how those windows line up across the frame.

A global shutter camera exposes every pixel at the same instant. The whole frame integrates one window of time. A rolling shutter camera exposes rows sequentially, from top to bottom. Each row starts and ends its exposure at a slightly different moment. The readout takes milliseconds, and the row timing offset is fixed by the sensor design.

Both architectures sample the PWM waveform, and both can produce artifacts. The mechanism differs. A global shutter camera compares one exposure window against the pulse train. A rolling shutter camera compares many staggered windows against the same pulse train. The second case produces spatial patterns, and the first produces temporal ones.

Exposure time relative to the PWM period decides what the camera captures. Consider a 1 kHz PWM signal with a 1 millisecond period. An exposure of 10 milliseconds spans ten full cycles. The camera integrates the same on-time fraction in every frame, and the average brightness stays stable. An exposure of 0.3 milliseconds spans less than one third of a cycle. The camera may catch the on portion in one frame and the off portion in the next. Frame brightness then swings wildly.

The ratio of exposure time to PWM period is the controlling variable. When exposure spans many cycles, the camera averages the waveform and the artifact disappears. When exposure approaches or falls below one cycle, the camera aliases the waveform and the artifact appears. This relationship holds for any LED light bar PWM dimming scheme, regardless of driver design.

From Flicker to Banding

Flicker and banding share one root cause and differ in how the artifact distributes across the image. Flicker is a frame-to-frame brightness change. Banding is a brightness change within a single frame.

Flicker appears when the exposure window and the PWM phase drift relative to each other. The camera captures a different slice of the waveform in each frame. Successive frames come out brighter or darker in a repeating pattern. The viewer sees pulsing or strobing. The effect is most obvious with short exposures and low PWM frequencies, where each frame lands on a distinctly different part of the cycle.

Banding appears with rolling shutters. Each row integrates its own exposure window, and adjacent rows start at slightly different times. If the PWM period is comparable to the row-to-row timing, different rows catch different parts of the waveform. Bright rows and dark rows alternate down the frame. The result is horizontal stripes that can scroll, hold steady, or shift with camera movement.

The band spacing follows a predictable relationship. It depends on the ratio of PWM frequency to the row readout rate. A high PWM frequency relative to the readout rate produces fine, closely spaced bands. A low PWM frequency produces wide bands. When the PWM frequency matches a multiple of the frame rate, the bands can appear to freeze in place. When the two drift apart, the bands crawl across the image.

Duty cycle shapes the band profile. A low duty cycle concentrates light into a narrow on-time window. Rows that catch the window look bright, and rows that miss it look dark. The contrast between bands is high. A high duty cycle spreads the on-time across most of the cycle. Every row catches some light, and the band contrast drops. This is why dimmed light bars often band more severely than bars running near full output.

The two artifacts can appear together. A rolling shutter camera with a short exposure can show both frame-to-frame flicker and in-frame bands. The flicker comes from phase drift between frames. The bands come from row timing within each frame. Diagnosing which artifact dominates helps narrow the fix. A global shutter camera removes banding entirely and leaves only flicker. A long exposure removes flicker and can soften banding. Neither change alters the underlying LED light bar PWM dimming behavior. Both change how the camera samples it.

Key Parameters and Fixes for Clean Footage

Key Parameters and Fixes for Clean Footage

Shenzhen Aurora Technology Limited builds light bars for vehicles that increasingly carry cameras. Dash cams, action cams, and machine-vision systems all sample the light output. Clean footage therefore depends on matching the light's switching behavior to the camera's sampling behavior. Four parameters control the outcome: PWM frequency, exposure time, duty cycle, and clock synchronization. Each parameter has a measurable target and a practical fix.

PWM Frequency and the IEEE 1789 Reference

PWM frequency sets the fundamental timing of the light output. A higher frequency shortens each on-off cycle. A shorter cycle makes the camera more likely to average several cycles within one exposure. The averaging effect suppresses both flicker and banding.

IEEE 1789 provides a widely cited framework for evaluating flicker risk. The standard defines modulation depth and frequency limits for human observers. It classifies flicker risk into no-effect, low-risk, and high-risk zones. The low-risk boundary rises steeply below roughly 90 Hz and flattens above a few hundred hertz. A light bar that operates in the low-risk zone for human vision still may not satisfy a camera. Cameras sample faster than the eye, and a camera can resolve modulation that a person cannot perceive.

The practical lesson is directional. Designers should push PWM frequency upward until the period falls well below the shortest exposure the camera will use. A common engineering target is a PWM period at least ten times shorter than the minimum exposure time. A camera with a 1/1000 second minimum exposure needs a PWM period below 100 microseconds, which corresponds to a frequency above 10 kHz. A camera limited to 1/120 second exposures needs far less, and a frequency of 1 kHz or 2 kHz can suffice.

Higher frequency carries costs. Switching losses rise, driver efficiency drops, and electromagnetic interference becomes harder to contain. Shenzhen Aurora Technology Limited balances these factors per model, and the approved model datasheet states the actual driver behavior. Buyers who plan to shoot video should confirm the PWM frequency of the specific model before purchase. Some models can be supplied to customer requirements, subject to the individual model datasheet.

A useful reference table summarizes the relationship between camera exposure and the PWM frequency needed to avoid artifacts:

Minimum exposure time

PWM period target (≤ 1/10 exposure)

Minimum PWM frequency

1/60 s (16.7 ms)

≤ 1.67 ms

≥ 600 Hz

1/250 s (4 ms)

≤ 0.4 ms

≥ 2.5 kHz

1/1000 s (1 ms)

≤ 0.1 ms

≥ 10 kHz

1/8000 s (0.125 ms)

≤ 12.5 µs

≥ 80 kHz

The table shows why high-frame-rate and high-shutter-speed work is demanding. A camera shooting at 1/8000 second needs a light source switching above 80 kHz, and few LED drivers reach that region without added cost. Designers who cannot raise frequency must instead control exposure or synchronize clocks.

Exposure Time and Rolling Shutter Interaction

Exposure time is the single most powerful lever a camera operator holds. The ratio of exposure time to PWM period predicts whether artifacts appear. A long exposure integrates many PWM cycles and produces a stable average. A short exposure integrates a fraction of one cycle and produces a volatile sample.

The fix is straightforward in principle. Operators should lengthen exposure until it spans at least ten PWM cycles. A 1 kHz light needs exposures of 10 milliseconds or longer, which corresponds to 1/100 second or slower. A 10 kHz light needs only 1 millisecond, which opens the door to 1/1000 second shooting. The frequency and the exposure must be chosen together, and neither parameter works in isolation.

Rolling shutter adds a spatial dimension. Each row of the sensor starts its exposure at a slightly different time. The row-to-row delay is fixed by the sensor, and typical values range from a few microseconds to tens of microseconds. When the PWM period approaches the row delay, adjacent rows sample different parts of the waveform. Bright and dark bands then appear across the frame.

The band spacing follows a simple ratio. Band spacing in rows equals the PWM period divided by the row delay. A 1 kHz light with a 10 microsecond row delay produces bands every 100 rows. A 10 kHz light with the same row delay produces bands every 10 rows, which reads as fine texture rather than obvious stripes. Raising frequency therefore compresses bands into a less objectionable pattern even when it cannot eliminate them.

Camera settings offer additional control. Many cameras include an anti-flicker mode that locks exposure to a multiple of the local mains frequency. This mode helps with mains-powered lighting and does little for a DC-driven light bar. A better approach for LED light bar PWM dimming is to set exposure manually and test at the actual shooting distance and angle. Automatic exposure can hunt between frames and amplify flicker, so manual lock is the safer choice.

Global shutter cameras remove banding entirely. Every pixel integrates the same window, so the frame shows one uniform brightness level. Flicker can still occur if the exposure is short relative to the PWM period. The fix remains the same: lengthen exposure or raise frequency. A global shutter camera paired with a high-frequency driver is the most robust combination for demanding footage.

Duty Cycle, Multiplexing, and Sync Effects

Duty cycle changes the shape of the artifact. A low duty cycle concentrates light into a narrow window. Rows or frames that catch the window look bright, and the rest look dark. Contrast between bands is high. A high duty cycle spreads light across most of the cycle, and every sample catches some output. Band contrast drops, and the artifact becomes subtler.

This behavior creates a counterintuitive result. A dimmed light bar often bands more severely than a bar at full output. The dimming action itself raises artifact risk. Operators who shoot at low brightness should compensate with a longer exposure or a higher PWM frequency. Shenzhen Aurora Technology Limited designs dimming circuits with this tradeoff in mind, and the approved model datasheet documents the dimming method for each model.

Multiplexing introduces a second timing layer. Some multi-emitter designs drive groups of LEDs in sequence rather than all at once. Each group switches at its own phase within a shared frame period. A camera can catch one group on and another group off, which produces color shifts or moving patterns across the beam. Multiplexing is common in RGB products and in designs that share driver channels across many emitters.

Synchronization solves the problem at the system level. A camera can output a trigger signal, and the light driver can lock its PWM phase to that trigger. The camera then samples the same part of every cycle, and the artifact disappears regardless of frequency or duty cycle. Sync requires a hardware link between camera and light, and it suits fixed installations such as work trucks with permanent camera systems. Sync also enables short exposures at modest PWM frequencies, which saves driver cost.

Three fixes therefore cover most situations. Raise the PWM frequency above the exposure-derived threshold. Lengthen the exposure until it spans many cycles. Synchronize the light clock to the camera clock. Each fix solves the problem in a different application, and the best choice depends on the camera system and the shooting conditions.

Test Methods for Flicker and Banding

Measurement turns a subjective problem into an engineering one. Three test methods cover most needs, and each answers a different question.

The first method uses a photodiode and an oscilloscope. A photodiode converts light output into a voltage signal with microsecond response. The oscilloscope displays the raw PWM waveform, and the engineer reads frequency, duty cycle, and rise time directly. This method verifies the driver behavior and confirms the datasheet values. It does not show what the camera sees.

The second method uses the camera itself. The operator points the camera at a uniform white surface lit by the light bar and records a short clip at the intended exposure and frame rate. Frame-by-frame analysis reveals flicker as brightness variation over time. A single-frame inspection reveals banding as horizontal stripes. This method captures the real sampling interaction and is the most relevant test for a specific camera and lens combination.

The third method uses a flicker meter or a dedicated test pattern. A spinning disc with alternating dark and light sectors produces a stroboscopic pattern under pulsed light. The pattern's appearance and stability reveal frequency and modulation depth at a glance. This method suits production-line checks and quick field verification.

A practical test protocol combines all three. The engineer first confirms the driver waveform with the photodiode. The engineer then records footage with the target camera at the shortest planned exposure. The engineer finally repeats the recording at the longest planned exposure and at the dimmest planned setting. Artifacts often appear only at the extremes, and testing the extremes catches problems that a mid-range test misses.

Design Checklist to Remove Banding

A structured checklist keeps the design review focused. The following items cover the decisions that most often determine whether footage comes out clean.

· Confirm the camera's minimum exposure time and frame rate before selecting a driver frequency.

· Set the PWM period to at least one tenth of the minimum exposure time.

· Verify the duty cycle range and test at the lowest dimming level, where band contrast peaks.

· Check whether the design multiplexes emitters, and confirm the phase relationship between groups.

· Specify a global shutter camera where the budget allows, and reserve rolling shutter cameras for longer exposures.

· Provide a sync input on the driver when the camera system supports a trigger output.

· Document the PWM frequency and dimming method on the model datasheet.

· Test with the actual camera, lens, and shooting distance rather than a bench reference.

The checklist works best as a gate. A design that fails any item goes back for review before release. This discipline prevents the common pattern where a light performs well on a bench and fails in the field.

Common Mistakes to Avoid

Several mistakes recur across projects. Each one has a clear correction.

The first mistake is treating frequency as the only variable. A high frequency helps, but a short exposure can still alias a fast waveform. Frequency and exposure must be evaluated together.

The second mistake is testing only at full brightness. Dimming raises band contrast, and a light that looks clean at 100 percent output can band badly at 20 percent. Test at the dimmest setting the application will use.

The third mistake is trusting the human eye. A light that appears steady to a person can still flicker on camera. The eye integrates slowly, and the camera does not. Always verify with footage.

The fourth mistake is ignoring the rolling shutter readout time. Two cameras with the same frame rate can have very different row delays. The band pattern changes accordingly, and a fix that works for one camera may fail for another.

The fifth mistake is assuming that anti-flicker camera modes solve the problem. Those modes target mains-powered lighting. A DC-driven light bar needs a frequency, exposure, or sync fix instead.

The sixth mistake is skipping documentation. A clean result on the bench means little without a recorded frequency, duty cycle, and test exposure. Documentation lets the next engineer reproduce the result and defend it to a customer.

LED light bar PWM dimming rewards a systematic approach. The parameters are measurable, the fixes are known, and the outcome is predictable. A team that tests early and documents carefully ships footage-ready lighting with confidence.


About the manufacturer

Shenzhen Aurora Technology Limited treats PWM artifacts as a predictable sampling interaction between a switching light source and a camera's exposure. They are not a random failure. Raising PWM frequency above the camera's exposure-derived threshold, shortening exposure, adopting DC dimming where the dimming range allows, or synchronizing light and camera clocks each solve the problem in different applications. Matching the dimming approach to the camera system and use case remains the most reliable path to clean footage.

FAQ

Does camera flicker mean the LED light bar is defective?

No. Flicker and banding reflect a sampling mismatch between the light's switching rate and the camera's exposure. Shenzhen Aurora Technology Limited builds these drivers to specification. The artifact comes from timing, not from a faulty unit.

How does a buyer pick a PWM frequency for a specific camera?

The buyer checks the camera's minimum exposure time first. The PWM period should run at least ten times shorter than that exposure. A 1/1000 second exposure therefore calls for a frequency above 10 kHz.

When does DC dimming or clock sync beat a frequency fix?

DC dimming removes switching artifacts entirely, and it suits applications that need only a modest dimming range. Clock sync locks the light to the camera trigger. Both approaches outperform a frequency-only fix in fixed installations.

What's Your Reaction?

Like Like 0
Dislike Dislike 0
Love Love 0
Funny Funny 0
Angry Angry 0
Sad Sad 0
Wow Wow 0
\