Deflicker Free Tool

Smooths exposure flicker out of a timelapse JPEG sequence, directly in your browser. Nothing uploads anywhere — frames never leave this tab.

1 Load sequence
2 Tune correction
3 Export

Select or drop the JPEG/PNG frames of one sequence, in order. RAW files can't be decoded in a browser — export a JPEG proxy set first if you're shooting RAW.

Drop frames here, or click to choose files
Frames are read locally — nothing is sent anywhere
Frames are sorted by filename. Name them so the sequence sorts correctly (e.g. frame_0001.jpgframe_0842.jpg).
Analyzing frame 0 / 0…
Edit Template
Free Resource
Get the Free Timelapse Field Guide.
Settings cheat sheets, interval reference, ND filter guide and pre-shoot checklist — all in one place. Instant delivery.
Spam check: =
✓ You're almost there — check your inbox (and spam folder) to confirm and get your free Field Guide.
Edit Template

Deflicker is a free browser-based tool that removes frame-to-frame exposure flicker from a timelapse JPEG or PNG sequence without ever sending your files anywhere. It reads every frame locally, measures its median luminance, converts that to an EV (exposure value) offset relative to the sequence average, and plots the result as a raw exposure curve. From there it computes a rolling median across a user-set window of neighbouring frames — this becomes the smoothed “target” curve your footage should have followed if exposure had been perfectly stable. The gap between each frame’s raw EV and its target EV is the correction that frame needs.

Two controls shape how that correction gets applied. Correction strength scales how aggressively the tool closes the gap between raw and target — 100% fully corrects to the smoothed curve, lower values blend a softer partial correction. Max correction per frame clamps how much gain any single frame can receive, so a genuinely bad exposure spike doesn’t get amplified into a visible pump. Every frame that hits this ceiling is counted and shown, because a high clamp-hit rate is a signal the flicker is too large for correction alone to fix cleanly.

Once you’re happy with the curve, the tool builds a per-frame gain lookup table in linear light (converting sRGB to linear, applying the gain, converting back), applies it to a scrubbable before/after preview, then — on export — re-renders every full-resolution frame through the same pipeline and packages the result as a ZIP. A parallel CSV export gives you the exact per-frame EV offset for every frame, so you can hand the same correction curve to Lightroom’s exposure sliders, an After Effects keyframe rig, or any tool that accepts a numeric exposure track, instead of trusting the JPEG re-render alone.

Why Timelapse Sequences Flicker in the First Place

Flicker rarely comes from the sensor itself — it comes from something in the capture chain making small decisions between frames that should have been locked. The most common cause on drones and mirrorless cameras with electronically-controlled aperture is aperture hunting: the lens re-settles to a very slightly different f-stop on each frame as it responds to metering, and because depth of field and light transmission both shift with aperture, the result reads as a rhythmic pulse rather than random noise. This is why fixed-aperture drones and manual, de-clicked lenses are structurally less prone to it than variable-aperture bodies shooting in anything other than full manual.

Auto-ISO and auto-exposure metering cause a related but distinguishable problem: instead of a small mechanical settling error, the camera is actively re-metering the scene every frame, which shows up as a slower drifting exposure rather than a sharp flicker — usually the result of clouds, changing shadows, or a scene with mixed bright and dark areas confusing average metering. This kind of drift responds well to a wide smoothing window and full correction strength, because the “true” signal you want to preserve is the actual light change, and the flicker you want to remove is the camera’s noisy attempt to track it.

A third cause — often confused with the first two — is genuine banding from AC-powered lighting sources (streetlights, indoor fixtures) running at a different frequency to your shutter speed. This looks like flicker in the histogram but has a different signature: it correlates with shutter speed and mains frequency, not with aperture or metering, and no amount of median-based exposure smoothing fully removes it, because the banding is baked into individual frames rather than varying frame-to-frame brightness alone. If your clamp-hit rate stays high no matter how you adjust the sliders, and the sequence was shot under artificial light, this is the more likely explanation — the fix is a shutter speed matched to the mains frequency next time, not more aggressive correction now.

How to Use Deflicker

  • Step 1 — Load your frame sequence

    Drop your JPEG or PNG frames onto the dropzone, or click to choose files, in the order they were shot. Name your files so they sort correctly by filename (frame_0001.jpg … frame_0842.jpg) — the tool sorts naturally by number, not by upload order. RAW files can't be decoded in a browser, so export a JPEG proxy set first if you're shooting RAW.

  • Step 2 — Let the tool analyse your sequence

    Each frame is read locally, resized to a small thumbnail, and measured for median luminance — nothing leaves your browser tab. A progress bar tracks frame-by-frame analysis. Once complete, the raw exposure curve, smoothed target curve, and corrected curve all render on the chart.

  • Step 3 — Adjust the smoothing window

    The smoothing window sets how many neighbouring frames feed into each point's target exposure. A narrow window (5–9 frames) tracks fast flicker closely; a wide window (25+ frames) is better for slower drift like cloud cover, since it won't chase genuine scene changes as if they were flicker.

  • Step 4 — Tune correction strength and the per-frame clamp

    Correction strength controls how fully each frame is pulled toward the smoothed target — start at 100% and back off if the correction looks over-applied. The clamp sets a ceiling on how much any single frame can be pushed, protecting against one bad frame getting over-corrected. Watch the "frames hitting the clamp" stat: if it stays high, the flicker is likely too large to fully fix here — see the Fundamentals section above on banding and aperture causes.

  • Step 5 — Scrub the before/after preview

    Drag the scrubber to step through your sequence frame by frame, watching the before and after panes update live alongside the exact EV correction applied to that frame. This is the fastest way to sanity-check the curve on individual problem frames before committing to a full export.

  • Step 6 — Export your corrected frames or the correction curve

    Choose an export resolution cap if you don't need full-resolution output, then download the corrected sequence as a ZIP of re-rendered JPEGs, or download the CSV correction curve to apply the same per-frame EV offsets manually in Lightroom, After Effects, or any tool that accepts a keyframed exposure track.

FAQ — Timelapse Deflicker Tool

What causes flicker in a timelapse video?

Timelapse flicker almost always comes from something in the capture chain making small, inconsistent decisions between frames — most commonly an electronically-controlled aperture re-settling slightly differently on each shot, or auto-exposure metering reacting to small scene changes frame by frame. True random sensor noise is rarely the cause.

The standard approach is to measure the brightness of every frame, build a smoothed target curve across neighbouring frames, and apply a per-frame exposure correction to pull each frame back toward that curve — which is exactly what this tool automates in the browser, with an optional CSV export if you’d rather apply the same correction manually in Lightroom or After Effects.

Yes — this tool runs entirely in your browser with no sign-up, no upload, and no watermark on the exported frames. Your files are never sent to a server; all analysis and correction happens locally in the tab.

Aperture flicker happens when a lens with electronically-controlled aperture re-settles to a very slightly different f-stop on each frame, changing both exposure and depth of field by a small but visible amount. It shows up as a rhythmic pulse rather than random noise, and is most common on variable-aperture cameras shooting outside full manual mode.

Not directly in a browser — browsers can’t decode RAW formats. Export a JPEG proxy sequence from your RAW files first (most editing software can batch-export proxies), run the proxy sequence through this tool, then apply the resulting CSV correction curve to your original RAW files in your editor of choice.

Bulb ramping adjusts exposure settings live during capture to smoothly transition across a lighting change (like day-to-night), preventing large exposure jumps from ever being recorded. Deflickering is a post-production fix for small, unwanted frame-to-frame variation that happens despite locked settings — the two solve different problems and are often used together on the same sequence.

If the “frames hitting the clamp” count stays high across most of your sequence no matter how you adjust the sliders, the underlying flicker is likely too large for smoothing alone to fix cleanly — over-correcting at that point can introduce a visible pumping effect of its own. A high clamp-hit rate is usually a sign to address the capture-side cause next time rather than push correction strength further.

Not reliably. Lighting-frequency banding is baked into individual frames as a mismatch between shutter speed and mains frequency, not as frame-to-frame brightness variation — median-based exposure smoothing doesn’t remove it. The fix is matching shutter speed to local mains frequency at capture time.

Yes — this tool never uploads your frames anywhere. All analysis, correction, and export happen locally using your browser’s own processing, and closing the tab discards everything; nothing is stored or transmitted.

JPEG and PNG sequences are supported directly. RAW files need to be exported as a JPEG proxy sequence first, since RAW decoding isn’t available in a browser environment.

Need help capturing professional drone footage for your next project but not ready to invest in the gear yet? Check out our Drone Videography service to see how Aero Timelapse Studio can help elevate your production.