AVCHD to HEIF Converter

AVCHD to HEIF: Pulling a Photo Out of Camcorder Footage

This is a frame grab, not a video conversion. The server decodes your AVCHD clip, takes the frame at a timestamp you choose, and saves that single picture as a .heif. Motion and sound are gone by definition — a HEIF written here is one still image. HEIF is a sensible target for this job because it is HEVC-compressed, so the same frame lands noticeably smaller than the equivalent JPEG. What follows is the practical part: how to land on a frame worth keeping, what the quality ladder actually does, and the one property of AVCHD footage that ruins frame grabs if you ignore it.

How to Convert AVCHD to HEIF

1. Upload Your AVCHD File: Drag the file onto the page or click "+ Add Files". This page accepts the .avchd extension specifically — if your camcorder clips are named .mts or .m2ts, which is what most cameras actually write, use MTS to HEIF or M2TS to HEIF instead of renaming them.
2. Set Frame Selection and Time (seconds): Frame Selection opens on Specific Frame with Time (seconds) at 0. Type the moment you want — decimals are accepted, so 12.5 is twelve and a half seconds in. Switch to Multiple Screenshots and pick a Capture Rate (opening at "1 second per frame") to sample the whole clip instead.
3. Set Quality Preset and Image resolution: Image Compression opens on Quality Preset at "Very High (Recommended)", with Specific file size as the alternative. Image resolution opens on Keep original; Resolution Percentage, Preset Resolutions, Width, Height and Width x Height are the other choices.
4. Convert and Download: Click Convert and save the .heif. Files are uploaded over an encrypted connection, processed on our servers, and deleted automatically after a few hours — no sign-up, no watermark.

The Interlacing Problem, and How to Work Around It

AVCHD was introduced by Sony and Panasonic in 2006, and a very large share of the footage shot in its first decade is 1080i — interlaced. In an interlaced frame the odd and even scan lines were captured a fraction of a second apart, so anywhere the picture is moving they do not line up, and the still you extract shows combing: thin horizontal teeth along moving edges.

There is no deinterlacing step in this pipeline, so combing in the source arrives intact in the HEIF. Nothing here removes it. What actually helps:

• Pick a still moment. A locked-off shot, a pause between movements, or a frame where the subject is stationary has little or no comb because the two fields captured nearly the same picture. This is the single most effective fix and it costs nothing.
• Sample first, choose second. Switch Frame Selection to Multiple Screenshots at one frame per second, look through the results, and convert again at the exact timestamp of the best one. Guessing at a number and hoping is how people end up with a combed frame.
• Scale down if you must. Halving the height through Image resolution softens comb teeth because the resampler averages adjacent lines. It is a blur, not a repair, but at thumbnail sizes it is often enough.
• Deinterlace properly elsewhere if the frame matters and the motion is unavoidable — a video editor's deinterlacer reconstructs a full frame from one field rather than blending both.

The other classic mistake is leaving Time (seconds) at 0. Frame zero of camcorder footage is very often the worst frame in the file: the lens still hunting for focus, an auto-exposure ramp, or the operator's hand still on the body. Treat the zero as a placeholder, not a default worth keeping.

What the Quality Preset Ladder Costs

Each preset hands a fixed quality value to the HEIF encoder. We measured the whole ladder on one 640 × 480 frame extracted from a test clip, so the sizes below are the real relationship, not an estimate:

"Highest" and "Very High" produced the same file size on our test frame, so reaching for the top of the list costs you nothing and gains you nothing. Note also that the picture data inside our HEIF is 8-bit HEVC at 4:2:0 chroma — HEIF as a format can carry more than that, but this pipeline does not emit it.

HEIF Compared With the Alternatives

On the same extracted frame, HEIF at the default preset came out about a third smaller than JPEG at the matching quality. That is the format's whole argument. The counter-argument is compatibility — if the still is going into an email, a document, or a website, AVCHD to JPG will save you a support conversation.

Frequently Asked Questions

Why is my HEIF just one picture instead of the whole clip?

Because HEIF is written here as a single still image. This page decodes one frame from the video and encodes that frame; there is no mode that turns an AVCHD clip into an animated HEIF. If you want several stills, switch Frame Selection to Multiple Screenshots and pick a Capture Rate — you get a separate .heif per sampled frame. If you want something that moves, convert to a video or animated format instead.

My camcorder files are .mts, not .avchd — can I use this page?

This page accepts the .avchd extension. Camcorders normally write .mts onto their own memory cards and the same stream is named .m2ts after import, so most people arriving here actually have one of those. Renaming a .mts to .avchd is not reliable — use MTS to HEIF, which accepts the extension your camera produced.

What happened to the audio from the clip?

It is discarded during frame extraction. A HEIF still has nowhere to put a soundtrack, so the AC-3 or linear PCM track in an AVCHD clip is decoded and dropped. If you need the audio as well, run the same file through an audio conversion separately — that is a different job from a frame grab and needs its own pass.

Why does my extracted frame have horizontal lines through moving objects?

That is interlacing comb, and it was in the source. AVCHD footage is frequently 1080i, meaning each stored frame interleaves two fields captured a moment apart; where the subject moved between them the lines do not align. There is no deinterlacing anywhere in this pipeline, so nothing removes it here. Grab a frame from a still moment instead, or deinterlace the clip in a video editor first.

Is "Highest" worth choosing over the default?

On our measurements, no. We ran the same extracted frame through the whole preset ladder: "Highest" (quality 100) and "Very High" (quality 95) produced files of exactly the same size, 38,255 bytes. The step that actually changes anything is "High" at quality 90. Leave the default alone unless you are deliberately trading quality for size, in which case "Medium" is the first setting most people find acceptable.

Will the still be full resolution?

Yes, by default. Image resolution opens on Keep original, so a 1920 × 1080 AVCHD clip yields a 1920 × 1080 still. The Preset Resolutions list sets the output height and lets the width follow the source aspect ratio, and it runs all the way from 4320p down to 16P for small thumbnails. Use Width x Height if you need an exact pixel pair. Upscaling past the source frame size only interpolates — the frame carries no more detail than the camera recorded.

Can I open a .heif on Windows?

Not always out of the box. macOS, iOS and iPadOS handle HEIF natively; Windows needs the HEIF image extension installed from the Microsoft Store, and Android support depends on the version and the gallery app. Browser support is limited enough that a .heif is a poor choice for a web page. If the file has to open somewhere you do not control, AVCHD to JPG is the safer target.

Does HEIF keep the camera metadata from the clip?

Video timecode, clip-level camera settings and the AVCHD index information do not survive a frame grab — the pipeline decodes the picture to a lossless intermediate and encodes that, so what reaches the HEIF is the image, not the camcorder's bookkeeping. Treat the output as a photograph of a frame rather than an archival record of the shoot, and keep the original clip if the metadata matters.

Is the file I upload kept anywhere?

No. It travels over an encrypted connection, is decoded on our servers, and both the uploaded clip and the resulting .heif are deleted automatically after a few hours. Files are never shared or made public, there is no watermark, and no sign-up or email address is needed. Because the work happens on our servers rather than your machine, the practical constraint is upload size and time rather than your own hardware.