DJI Flip ND16/64/256 Exposure and Validation Guide

An ND filter is an exposure control, not a cinematic preset

The randomly selected product for this guide is the DJI Flip ND Filter Set (ND16/64/256). The current listing contains one ND16, one ND64 and one ND256 filter, gives an approximate mass of 0.34 g per filter, and identifies DJI Flip and DJI Lito X1 compatibility. DJI describes the set as a way to control shutter speed in strong light and to support long-exposure timelapse work.

Neutral density reduces the light reaching the sensor without intentionally changing scene color. It cannot recover clipped highlights, create dynamic range, correct white balance, sharpen a soft lens, remove propeller shadows or guarantee pleasing motion blur. Its engineering role is narrower and more useful: after frame rate, color mode, aperture and ISO have been selected, the filter moves the required shutter time into a desired range.

That distinction matters on the DJI Flip because the camera has a fixed f/1.7 aperture. Unlike a camera with an adjustable iris, it cannot stop down to reduce light while retaining ISO and shutter speed. DJI's specifications list a 1/1.3-inch sensor, 24 mm-equivalent field of view, video ISO ranges of 100–6400 in Normal and 100–1600 in D-Log M, and recording modes up to 4K at 100 fps. The ND filter is therefore the main physical exposure-control variable available after the shooting format and ISO are fixed.

1. Translate ND16, ND64 and ND256 into stops

The number is the light-reduction factor. ND16 transmits one sixteenth of the incident light, or 6.25%; ND64 transmits one sixty-fourth, or 1.5625%; ND256 transmits one two-hundred-and-fifty-sixth, or 0.390625%. Because each photographic stop halves the light, the stop reduction is the base-two logarithm of the factor. ND16 is four stops, ND64 is six stops and ND256 is eight stops.

At the same aperture and ISO, a four-stop filter allows an exposure time sixteen times longer. An unfiltered 1/800 s meter reading becomes approximately 1/50 s with ND16. The same reading becomes 1/12.5 s with ND64 and about 1/3.1 s with ND256. These are exposure-equivalent calculations, not guarantees that the aircraft can hold the camera still for those durations.

The set jumps by two stops between filters. There is no five-stop ND32 or seven-stop ND128 in this package. In some light, none of the three produces the exact target shutter while ISO remains at its minimum. The correct response is to choose the closest useful compromise, change frame rate or creative intent, or wait for different light—not to force the camera into a high ISO that cancels the optical reduction.

Filter labels should be verified as a set during acceptance. Photograph or meter a stable, evenly illuminated target from a fixed position at manual exposure. Relative to no filter, ND16 should reduce exposure by about four stops, ND64 by six and ND256 by eight. Consumer metering, flicker and changing daylight can obscure small errors, so conduct a controlled comparison and repeat it. A large departure, uneven density or unexpected color shift is a reason to inspect the filter and test method.

2. Calculate a 180-degree shutter instead of memorizing one speed

Shutter angle expresses exposure time relative to the frame interval. For a 180-degree shutter, exposure time is half of one frame period: t = 1 ÷ (2 × frame rate). At 24 fps the target is about 1/48 s; at 25 fps it is 1/50 s; at 30 fps, 1/60 s; at 48 fps, 1/96 s; at 50 fps, 1/100 s; at 60 fps, 1/120 s; and at 100 fps, 1/200 s. DJI uses the 24 fps and 1/48 s example in its product explanation.

A 180-degree target is a conventional motion-rendering choice, not a law. Faster shutter produces crisper individual frames and more staccato motion. Slower shutter increases blur and can smear detail. Aerial footage contains camera translation, yaw, gimbal movement, vibration, moving subjects and parallax at different distances, so two shots at the same shutter can look different. The target should follow the final delivery and movement, not an internet slogan.

For inspection, photogrammetry, frame extraction or subjects that must be measured sharply, a fast shutter may be more important than cinematic blur. Adding ND simply to force 1/60 s could destroy useful detail. Conversely, a low-altitude lateral tracking shot at 30 fps and 1/1000 s may appear unnaturally crisp because ground texture changes position substantially between frames without blur connecting the samples. Choose the output purpose first.

High frame rate changes the requirement. If 4K/100 fps will be conformed to 25 fps for quarter-speed playback, a 180-degree capture target is approximately 1/200 s, not 1/50 s. Motion blur is created at capture and remains when footage is slowed. Setting 1/50 s at 100 fps would require an exposure longer than the frame interval and is not a coherent 180-degree plan.

3. Select the filter from an unfiltered meter reading

Set the actual recording mode before measuring: resolution, frame rate, Normal or D-Log M, white balance and ISO. Set the desired shutter. With no filter installed, point the camera at the representative brightest composition—not directly at the sun—and note the shutter speed that would produce the chosen exposure at the same ISO and f/1.7 aperture. The required filter factor is the unfiltered shutter denominator divided by the target shutter denominator.

Example one: at 30 fps the 180-degree target is 1/60 s. If the unfiltered reading is 1/4000 s, the required factor is 4000 ÷ 60 = 66.7. The stop requirement is log2(66.7), about 6.06 stops. ND64 is the closest match. After installation, re-meter and confirm the histogram and highlight warnings rather than assuming the arithmetic perfectly predicts transmission.

Example two: at 24 fps the target is 1/48 s. If the unfiltered reading is 1/800 s, 800 ÷ 48 = 16.7, or about 4.06 stops. ND16 is appropriate. Example three: at 100 fps the target is 1/200 s. An unfiltered 1/3200 s reading requires a factor of 16, exactly four stops, so ND16 is the logical starting point.

Consider a brighter 30 fps scene metering 1/8000 s. The ideal reduction to 1/60 s is a factor of about 133, just over seven stops. ND64 would yield approximately 1/125 s; ND256 would yield roughly 1/31 s at the same ISO. Neither is exact. ND64 preserves highlights with a faster-than-180-degree shutter, while ND256 would require a faster shutter than 1/31 s to prevent overexposure. The desired motion, highlight protection and available exposure controls decide the compromise.

Do not select a filter only from labels such as “sunny,” “cloudy” or “snow.” Scene luminance changes with latitude, season, altitude, ground reflectance, haze, sun angle and composition. Water, pale roofs and snow can differ from vegetation under the same sky. Meter the intended shot and reevaluate when the aircraft orientation or route changes.

4. Protect highlights before protecting the rule

Digital highlight clipping cannot be repaired by a smoother shutter angle. Use the live histogram, overexposure warning or zebra display available in the current DJI Fly workflow. Check the brightest important surfaces while the aircraft is at a safe position, and leave margin for turns that bring specular water, sunlit roofs or cloud edges into view. The tiny remote-controller preview is not a reliable clipping detector by eye.

Keep ISO at the intended baseline when the objective is maximum image quality, but do not treat ISO 100 as more important than correct exposure. If the selected filter is too strong and automatic exposure raises ISO, the filter may have traded light for noise and reduced highlight flexibility without delivering the planned shutter. Manual exposure makes that trade visible. When lighting changes rapidly, a controlled faster shutter with low ISO can be cleaner than an exact 180-degree shutter achieved through high gain.

D-Log M and Normal are not interchangeable exposure workflows. DJI lists video ISO up to 1600 for D-Log M and up to 6400 for Normal. Log footage needs the appropriate monitoring and post-production transform, and exposure practices should be validated for the chosen profile. Do not judge log clipping or midtones solely from the flat-looking display. Record short tests, apply DJI's current D-Log M to Rec.709 transform where appropriate, and inspect scopes on the intended editing system.

DJI states that HDR is enabled by default at 60 fps or below in relevant camera modes. HDR processing does not make exposure unlimited and does not eliminate the need to protect important highlights. Test the exact frame rate, profile, firmware and editing pipeline because automatic image processing can affect tone and the way motion artifacts appear.

5. Separate video motion blur from still-photo long exposure

For video, the common objective is consistent temporal rendering across shots. Lock frame rate, shutter, ISO strategy and white balance for a sequence, then choose the filter that makes that exposure feasible. A filter change between shots is preferable to allowing automatic shutter to move from 1/60 to 1/2000 without creative intent. However, never delay a safe flight decision to preserve camera continuity.

For stills, ND can blur water, traffic or people, but the camera itself is airborne. Wind correction, GPS position changes, gimbal movement and vibration can blur the entire frame. The DJI Flip specification lists 12 MP exposures to two seconds plus a simulated long-exposure range of 2.5–8 seconds; it also supports DNG/RAW. A simulated long exposure is computational and should not be assumed equivalent to a single uninterrupted optical exposure.

A practical still test starts on the ground with motors off where allowed by the manual. At ISO 100 and f/1.7, note the unfiltered shutter, calculate the filtered equivalent, and inspect a raw file for sharp stationary edges, color and vignetting. Then conduct a lawful airborne test in calm conditions at a conservative height. Compare the center and corners at 100% view. If the background is soft as well as the moving subject, the exposure exceeded the platform's stable capture capability.

For timelapse, distinguish interval from exposure. The interval must exceed the exposure plus processing and storage overhead. A strong filter can create an exposure so long that the requested interval is impossible or leaves insufficient time for gimbal or position settling. Test the complete sequence duration, battery reserve and file-writing behavior rather than verifying a single frame.

6. Control aircraft movement as part of the exposure

At 1/60 s, an aircraft moving 10 m/s travels about 0.167 m during the exposure. That does not translate directly into image blur because distance, viewing angle and focal length matter, but it shows why nearby ground can streak while a distant horizon remains sharp. Lower altitude and sideways movement create strong angular motion. Plan speed, path and gimbal angle together with shutter.

Yaw is especially visible. An abrupt heading correction moves the entire image across the sensor, and longer shutter spreads each frame over a larger angle. Use smooth control inputs and a repeatable flight path. A mechanical three-axis gimbal stabilizes orientation, but it cannot remove intentional camera translation or every high-frequency disturbance. ND should never be used to hide poor propeller balance, vibration or aggressive piloting.

Propeller shadows and flicker are separate phenomena. When sunlight passes through a rotating propeller before reaching the lens, bands can appear. Changing shutter or flight orientation can alter their visibility, but an ND filter is not a guaranteed cure. Avoid compositions and sun angles that place rotor modulation across the lens, and review test footage at normal and frame-by-frame playback.

Artificial lighting can flicker at mains-related or driver-specific frequencies. A 180-degree shutter may conflict with local 50 Hz or 60 Hz lighting. In a 50 Hz environment, 1/50 or 1/100 s often offers a useful starting point; in a 60 Hz environment, 1/60 or 1/120 s may. LED drivers can use other modulation. Run a short ground recording and inspect for bands before the mission. Choose flicker control over a theoretical shutter angle when both cannot be satisfied.

7. Validate optical neutrality rather than assuming it

A filter introduces two air-glass interfaces and a mechanical frame. Acceptance should check transmission, color, sharpness, flare, ghosting, vignetting, reflections and seating. Use the same manual white balance for filtered and unfiltered comparisons. Photograph a neutral target and a detailed flat field under stable light, then compare RGB values and corner detail. Auto white balance can conceal a color cast by changing the correction between frames.

Test strong point sources near and inside the frame. Extra reflections may create ghosts or lower contrast. Rotate or tilt the camera through representative sun angles on the ground. Dust, fingerprints and cleaning streaks become especially damaging in backlight. Inspect both surfaces with appropriate illumination and use only a lens-safe blower, brush or clean microfiber procedure recommended for optical coatings.

Evaluate sharpness with raw or minimally processed files. Digital sharpening can hide small losses, while motion blur can be mistaken for optical softness. Use a high-contrast detailed target, stable camera, low ISO and several repeated frames. Compare center and corners at the same effective exposure. If only one filter shows asymmetric softness, inspect seating and glass; if every configuration shows the same directional smear, investigate camera movement or focus distance.

Check vignetting in a bright uniform field and in real scenes. Mechanical misalignment may darken a corner. Uneven density can create a gradient that follows the filter rather than the sky. Do not “fix” a poorly seated filter by pressing on the gimbal or forcing the frame. Remove it, inspect the mount and confirm the exact installation procedure.

8. Treat 0.34 g as a gimbal-interface specification

One filter is listed at approximately 0.34 g. Relative to the Flip's sub-249 g standard aircraft mass, that is small, but the relevant mechanism is the camera gimbal, not total aircraft lift. Mass ahead of the gimbal axes changes balance and motor load. The official accessory is designed for the compatible camera; stacking filters, adding adapters or using a damaged frame changes that condition.

Install or remove the filter only with the aircraft powered off and secured, following current DJI instructions. Avoid touching or twisting the gimbal mechanism. Confirm that the filter is fully seated and cannot detach. Remove the gimbal protector before power-up as instructed, keep the aircraft clear, and observe the startup calibration. A gimbal overload, abnormal noise, limited travel or visible horizon error is a stop condition.

After installation, move the powered-off gimbal only as permitted to inspect clearances, then perform a low-risk camera check before takeoff. Verify that the filter frame does not enter the image at extreme controllable tilt. After a hard landing, filter impact or transport event, inspect the mount, glass and gimbal before flying. A cracked filter can shed fragments near the lens and moving components.

9. Use a repeatable field decision sequence

First, define delivery. Choose frame rate, resolution, aspect orientation, color mode and whether the shot will play at normal speed or be conformed. Write the target shutter and ISO strategy. A 4K/30 Normal shot, 4K/60 HDR shot and 4K/100 slow-motion shot are three different exposure tasks.

Second, meter without the filter. At f/1.7 and the chosen ISO, point at the representative brightest view and determine the shutter needed to protect important highlights. Calculate required stops from the target. Select ND16 for approximately four stops, ND64 for six or ND256 for eight.

Third, install and verify. Power down, fit the filter correctly, clean it and run the gimbal startup check. Restore manual exposure and white balance. Recheck histogram and highlight warning because actual scene light and filter transmission can differ from the planning estimate.

Fourth, record a short test. Include camera movement, aircraft motion and the subject behavior expected in the real shot. Review exposure, clipping, motion cadence, vibration, corner sharpness, flare and flicker. A static preview cannot reveal every temporal problem.

Fifth, monitor change. Cloud cover, heading and ground reflectance can shift exposure by several stops during a route. If the filter no longer fits, land safely and change it. Do not attempt to handle the camera on an armed aircraft. When continuity matters, record an exposure note for every shot so post-production can identify intentional differences.

10. Diagnose common failures from evidence

The image is dark and noisy. Check whether auto ISO increased because the filter was too strong. Compare manual exposure with and without the filter. A stronger ND does not improve quality when the camera compensates electronically.

Footage looks smeared. Separate subject motion, aircraft translation, yaw, vibration and focus. Increase shutter for a test without changing the route. If smear disappears, the chosen motion blur was too long for the movement. If high-frequency softness remains, inspect propulsion balance and gimbal behavior.

Color shifts between clips. Confirm that manual white balance and color mode were constant. Shoot a neutral reference with each filter. A shift that follows one filter suggests transmission tint; a shift that occurs without changing filters may be auto white balance or changing illumination.

One corner is soft or dark. Reseat the filter and inspect the frame and glass. Compare with an unfiltered ground image at the same exposure. Do not assume software lens correction will hide a mechanical seating problem.

Gimbal startup fails. Power down immediately. Confirm compatibility, correct orientation, full seating and absence of stacked accessories or debris. Do not repeatedly force calibration against an obstruction. Remove the filter and verify the bare camera before deciding whether the accessory or gimbal requires service.

Bright bands appear under LEDs. Test shutter values synchronized to the lighting frequency and check at the exact location. Changing ND alone does not change flicker frequency; it only changes the exposure combination available.

11. Build a small acceptance matrix before client work

Create controlled samples for no filter, ND16, ND64 and ND256 at the same white balance, ISO and equivalent exposure. Include a detailed daylight target, a neutral card, sky gradient, backlit point source, moving subject, lateral flight and yaw. Record firmware, temperature, frame rate, shutter, profile and filter identity. Preserve original files rather than judging only exported social-media copies.

For every filter, verify approximately correct stop reduction, consistent color, acceptable center and corner sharpness, no mechanical vignetting, stable gimbal calibration and secure retention. Then define the creative operating matrix: which filter is the first candidate for each measured unfiltered shutter and target frame rate, and what alternate shutter is acceptable when the two-stop gaps prevent an exact match.

Repeat critical checks after firmware changes, camera service or filter damage. A field kit should keep filters separated and labelled, with clean storage and an inspection light. Do not place an exposed filter in a pocket where grit can damage the coating. Log cleaning or impact events if the footage has contractual image-quality requirements.

The imaging decision

The ND16/64/256 set provides a practical four-, six- and eight-stop range for a fixed-aperture DJI Flip camera. It is especially useful when bright light would otherwise force a shutter far above the motion-rendering target, and when deliberate long-exposure or timelapse work needs more exposure time. Its value comes from measurement and repeatability, not from leaving one filter installed all day.

Choose the format and motion intent, meter the real scene at the chosen ISO, calculate the required reduction, install the closest filter, and verify highlights and image quality. Accept that the two-stop gaps sometimes require a shutter compromise. For inspection or measurement, prioritize sharp evidence over cinematic convention. For creative motion, control aircraft movement as carefully as the exposure.

A well-run ND workflow makes footage easier to match because the operator can explain every exposure choice. It also reveals when no filter is the correct tool. That is the professional standard: use filtration only when it solves a defined optical problem, validate the accessory as part of the gimbal-camera system, and retain enough metadata to reproduce the result.