Initializing... drag & drop files here
Supports: M2V
You have an .m2v — a raw MPEG-2 video elementary stream, the video-only intermediate that DVD and broadcast authoring tools produce — and you need it as an .flv for a legacy player, an old CMS, or a media server that only ingests Flash Video. This walk-through covers the codec choice that decides whether the result is watchable, the rate-control setting that behaves differently here than on most pages, and the handful of errors that DVD-era MPEG-2 reliably produces.
.m2v onto the page or click "Add Files". Several files can be queued and converted in one batch.The Flash Video container was defined by Adobe's Flash Video File Format Specification (version 10.1, 2010, the last public revision), which assigns a numeric codec ID to each payload it can carry. The four this page exposes are four different entries in that table, and two of them are not general-purpose video codecs at all.
| Option in the dropdown | What it really is | Use it when |
|---|---|---|
| FLV (preselected) | Sorenson Spark, an H.263-derived codec — FLV codec ID 2 | The player is old Flash Player 6/7-era software that predates VP6 and H.264 |
| H.264 | AVC inside FLV — codec ID 7, added in Flash Player 9 | Anything from roughly 2008 onward; far better quality per bit than Sorenson Spark |
| Flash Video | Screen video — codec ID 3, a screencast codec | Recorded desktop/screen content with flat colour and sharp text |
| Flash Video (v2) | Screen video version 2 — codec ID 6 | Same screencast use case, newer revision |
The two screen-video entries are lossless-oriented codecs designed for screen recordings, so the encoder skips bitrate controls entirely for them. Point them at DVD footage and you get an enormous file that looks no better than the source. For real-world video, the choice is between Sorenson Spark for maximum compatibility with ancient players and H.264 for anything else.
Quality Preset works one of two ways depending on the codec. For codecs with constant-quality encoding, it computes a quality target from your preset and the output resolution. Sorenson Spark has no such mode, so the preset falls back to scaling the source video's bitrate — and this is where .m2v is unusual:
ffprobe reports bit_rate=N/A at both stream and file level, because there is no container header to hold it.If output size or quality matters, do not rely on the preset. Set Constant Bitrate and name a figure in Kbps or Mbps, or use Constant Quality — for Sorenson Spark, the value in that field acts as a quantiser on a 1-31 scale where lower means better quality, and it starts at 5. If you have a hard size ceiling, Specific file size divides your target by the running time instead.
.m2v has no audio track to begin with; the format cannot store one. The Audio Codec dropdown still renders (AAC, ADPCM, MP3), but there is nothing for it to encode, so every output from this page is silent.FLV survives in two places: old archives, and RTMP — live encoders such as OBS still push RTMP, which wraps its payload in FLV-style tags before the platform repackages it as HLS. Neither of those is a file you upload. If your actual goal is a video that plays on a phone, a browser, or a modern TV, FLV is a dead end and you should target a current container instead.
| If your goal is… | Convert to | Why |
|---|---|---|
| Playing it anywhere today | M2V to MP4 | H.264 in MP4 is the universal baseline |
| Embedding in a web page | M2V to WebM | Native <video> playback, no plugin, open codecs |
| Keeping quality for archive | M2V to MKV | Flexible container, no re-wrap needed later |
| Just making the file smaller | Compress M2V | Stays MPEG-2, so DVD tooling still accepts it |
Because the source has no audio. .m2v is an MPEG-2 elementary video stream: a video-only bitstream with no container, which is exactly why DVD authoring uses it — the picture is encoded separately from the .ac3 or .mp2 soundtrack and the two are muxed at the end. Nothing is dropped in conversion; there was never an audio track in the file. To end up with sound, mux the audio back in with a desktop tool, or convert to a container-based target such as M2V to MP4 and add the audio there.
Desktop players, not browsers. VLC, mpv and MPC-HC all decode FLV directly, and ffmpeg-based tools handle it fine. What no longer exists is in-browser playback: Flash Player was discontinued at the end of 2020 and every mainstream browser removed the plugin, so an .flv embedded in a web page will not play for anyone. If you need something a visitor can watch, convert FLV to MP4 afterwards or target MP4 in the first place.
H.264 unless you have a specific reason not to. Both are legal FLV payloads — the format's codec table lists Sorenson H.263 as ID 2 and AVC as ID 7 — but H.264 arrived with Flash Player 9 and delivers dramatically better quality per bit. Sorenson Spark only makes sense when the software that has to read the file predates that: some Flash Player 6/7-era projectors, old Flash Media Server workflows and certain embedded devices really do only understand codec ID 2.
Usually yes, sometimes dramatically, because MPEG-2 is extremely inefficient by modern standards and DVD video runs up to 9.8 Mbit/s. But the amount depends entirely on the rate control you choose, and the default preset is unreliable on this page because .m2v carries no source bitrate for it to scale. Set Constant Bitrate or Specific file size if size is the point of the exercise.
Yes. The "Trim" section starts on Unchanged; switch it to Time Range and give a start time plus a duration, and only that section is encoded. This is worth doing on DVD rips, which often carry the whole title in one long stream. For frame-accurate work on the source instead, video cutter handles trimming as its own operation.
Because you are usually comparing two different codec generations. With FLV selected, the payload is Sorenson Spark, an H.263 derivative from the early 2000s with no B-frames worth speaking of, coarse motion compensation and a small toolset. H.264 in an MP4 has two decades of coding tools on top. Selecting H.264 inside the FLV container closes most of that gap — the container is not what limits quality, the codec is.
No. There is no deinterlacing step anywhere in this pipeline, so 480i or 576i fields are encoded as-is and the combing stays in the picture. That is deliberate: automatic deinterlacing damages progressive footage that was wrongly flagged, and a bad guess is worse than none. Run a proper deinterlace filter locally first if your source is interlaced and the output is for viewing rather than archiving.
Yes. The .m2v is uploaded over an encrypted connection, converted on our servers, and deleted automatically after a few hours, along with the FLV output. There is no sign-up, no watermark, and files are never shared or made public. We checked the claims on this page against a real MPEG-2 elementary stream: ffprobe identifies it as mpegvideo with no duration and no bitrate, which is exactly why the rate-control advice above matters more here than on a container-based source.