MJPEG to HEVC Converter

MJPEG to HEVC Converter

Motion JPEG is video in the loosest sense: a run of complete, independent JPEG frames with no compression between them, which is why an hour from an IP camera, a microscope or a capture card takes so much disk for the picture it delivers. HEVC — ITU-T Rec. H.265 / ISO/IEC 23008-2, first edition approved in April 2013 and still actively maintained — replaces that redundancy with inter-frame prediction, so the same footage lands dramatically smaller. The one thing to understand before you convert: .hevc names a codec, not a container, so our pipeline writes the H.265 stream into a Matroska (MKV) container and gives it the .hevc extension you asked for. Rename it to .mkv and every MKV-aware tool opens it normally.

How to Convert MJPEG to HEVC

1. Upload Your MJPEG File: Drag your .mjpeg or .mjpg onto the page, or click "Add Files" to browse. Security-camera exports, webcam captures, machine-vision and microscope acquisitions and old digital-camera clips all work as input, and several can be queued together.
2. Set the Quality Under File Compression: The Video Codec group is pinned to H.265 for this target. File Compression opens on Quality Preset, whose dropdown runs Highest to Lowest with Very High (Recommended) preselected; Constant Quality exposes an explicit CRF slider on the familiar 16-51 scale, preset to 23.
3. Adjust Video Resolution or Trim (Optional): Video resolution starts on Keep original, which is what you want for evidence and scientific footage. Preset Resolutions sets a target height and lets width follow the source aspect; Width x Height takes exact pixels. Trim exports only a time range.
4. Convert and Download: Click Convert and save the file, renaming it to .mkv if a player refuses it. Uploads travel over an encrypted connection, are processed on our servers, and are deleted automatically after a few hours. No sign-up, no watermark.

MJPEG Format at a Glance

What the .hevc File Actually Is

Choosing a Rate-Control Mode

H.265 is a genuine CRF codec, so every mode in the File Compression group behaves the way its name suggests — which is not true of the older MPEG-4 targets on this site, where the quality field degrades into a quantiser scale.

• Quality Preset is the one-click route. It maps your chosen preset to a CRF value that also accounts for the output resolution, so a downscaled clip is not compressed twice over.
• Constant Quality hands you the CRF directly on a 16-51 scale, preset to 23. Lower means better quality and a bigger file; each step of about 6 roughly doubles or halves the bitrate. This is the mode to use when comparing settings across several clips, because it targets a perceptual level rather than a size.
• Constraint Quality adds a Max bitrate ceiling on top of the CRF, so quality floats but the stream never exceeds your cap — useful when a downstream pipeline has a hard bandwidth limit.
• Constant Bitrate and Variable Bitrate take explicit numbers in bps, Kbps or Mbps.
• Specific file size encodes toward an exact deliverable size. Pairing an aggressive size target with a heavy downscale is where visible damage appears first; change one variable at a time.

The server always re-encodes. There is no stream-copy path from JPEG frames into H.265, because the two share nothing at the bitstream level — this is a full decode-and-encode pass in every mode.

Common Errors and How to Fix Them

• "My phone, QuickTime or TV won't open the file" — Stock players do not read Matroska, and many older ones lack HEVC decoding besides. Rename to .mkv for MKV-capable apps, remux locally without re-encoding (ffmpeg -i video.hevc -c copy video.mp4), or convert to MP4 from the start.
• "VLC plays it but my editor rejects it" — The file has normal container timestamps, duration and working seek; what trips software up is the unfamiliar extension. Renaming to .mkv usually settles it.
• "There is no sound" — Expected. A bare Motion JPEG stream carries no audio track, so nothing is encoded. The Matroska file would carry a soundtrack fine if the source had one.
• "Playback speed or duration looks wrong" — Raw MJPEG stores no frame timing, so decoders fall back to an assumed rate. A capture recorded at 15 fps will not report the duration you expect.
• "The output still looks soft" — The source frames were already JPEG-compressed at capture. This is a lossy-to-lossy re-encode; it can preserve detail but never restore it.

Frequently Asked Questions

Why did I get a Matroska file when I asked for HEVC?

Because HEVC is a codec and a file needs a container. There is a raw H.265 elementary-stream format, but almost no ordinary player will open one, and it carries no duration or seek information — a genuinely raw .hevc would be less useful than what you get here. Our pipeline maps H.265 output to Matroska, which handles the codec cleanly and stores proper timestamps, so the download is a real MKV wearing a codec-style extension. Renaming it re-encodes nothing.

Is HEVC really about half the size of the equivalent H.264?

Roughly, at matched visual quality, and the standard was designed with a 25-50% bitrate reduction over H.264/AVC as its target — with larger savings at 4K than at low resolutions, since HEVC's bigger coding-tree blocks pay off most on large frames. Against an MJPEG source the reduction is far larger than either figure, because MJPEG has no inter-frame compression at all to start from. Real numbers depend heavily on content: static camera views save the most, handheld or noisy night footage the least.

Why is the converted file silent?

Because the source is. A bare .mjpeg or .mjpg file is an elementary stream — a sequence of JPEG images with no audio substream — so there is nothing for the encoder to compress, whatever the Audio Codec dropdown is set to. That dropdown is shared across every video target on the site and simply has no input here. If your footage does have sound, it is sitting inside a real container such as AVI or MOV; convert that file with MJPEG to MP4 or MJPEG to MKV instead.

What can actually play the file?

More than the extension suggests. Because it is an ordinary Matroska file, FFmpeg-based players such as VLC and mpv open it directly with correct duration and seeking, and MKV tooling like MKVToolNix ingests it happily. Stock consumer software generally cannot — phones, smart TVs, QuickTime and browsers mostly do not read Matroska, and hardware HEVC decoding is patchy on pre-2017 devices. Rename to .mkv for extension-driven software, or pick an MP4 target if the file is headed somewhere you do not control.

Should I choose .hevc, HEVC inside MP4, or AV1?

The .hevc output is already HEVC inside a container, so the real question is packaging and reach. Keep it (or rename to .mkv) when your tools handle Matroska. Choose MP4 or MKV with H.265 selected when the file is going to stock apps. Choose AV1 when you want the smallest result and are willing to wait longer for the encode — AV1 is royalty-free and slightly more efficient, but slower and less widely hardware-decoded than H.265.

Why does HEVC take so much longer than the source took to record?

Because the encoder is doing work MJPEG never did. Motion JPEG compresses each frame in isolation, which is cheap enough to run on camera hardware from twenty years ago. H.265 searches for motion between frames, evaluates variable-size coding-tree partitions and tests multiple prediction modes per block. That analysis is exactly what buys the size reduction, and it costs time — long or high-resolution captures take real minutes. Trimming to the section you need before converting is the fastest way to cut it down.

Does this converter deinterlace or tone-map my footage?

No to both. If your source is interlaced, the combing stays baked into the output — there is no deinterlacing filter in this pipeline, so an interlaced capture should be handled before it reaches here. Likewise there is no HDR tone-mapping: HDR input receives a plain colour conversion and may look washed out. Neither is common in MJPEG captures, which are overwhelmingly progressive SDR, but it is worth knowing before feeding in something unusual.

How long do you keep my file?

Not long. The upload travels over an encrypted connection, the conversion runs on our servers, and both the source and the output are deleted automatically after a few hours. Nothing is shared or made public, there is no account requirement and no watermark on the result. Because processing is server-side, the real constraint on a very large capture is upload time — use the Trim control, or the video cutter, to scope a long recording first. To shrink an HEVC file you already have, see the HEVC compressor.