M2V to BMP Converter

Pulling a Still Out of a Bare MPEG-2 Stream

An .m2v is not a video file in the everyday sense. It is a naked MPEG-2 video elementary stream — the picture data defined by ITU-T H.262 / ISO/IEC 13818-2 with no container wrapped around it, no audio, no index and, as this page explains, no presentation timestamps. DVD authoring suites, broadcast encoders and MPEG-2 hardware cards all write them as an intermediate. This page decodes one and hands you a Windows bitmap.

Two things about the combination are worth knowing before you start. The moment you type into Frame Selection is an estimate on this format rather than a lookup, and the Quality Preset on a BMP target is not a compression slider — it chooses between three structurally different bitmap encodings.

How to Convert M2V to BMP

1. Upload Your M2V File: Drag the .m2v onto the page or click "Add Files" to browse. Several streams can be queued and they all use the same frame and quality settings.
2. Choose Frame Selection: "Specific Frame" grabs one still and exposes a Time (seconds) box that starts at 0. "Multiple Screenshots" instead walks the whole stream at a Capture Rate — anything from ten frames a second to one frame every ten seconds, opening at one second per frame — and writes one BMP per sample.
3. Pick a Quality Preset and a Resolution: Under Image Compression the list holds High, Medium and Low; those are the only three entries a BMP target shows, and the table below says what each one actually writes. Image resolution opens on Keep original, with Resolution Percentage, Preset Resolutions, Width, Height and Width x Height alongside it.
4. Convert and Download: Click Convert and take the bitmap. Files are uploaded over an encrypted connection, processed on our servers and deleted automatically after a few hours — no sign-up, no watermark, never shared or made public.

Walk-through: Why the Time You Type Is an Approximation

Every frame in a muxed file — an MPEG program stream, an MP4, a transport stream — carries a presentation timestamp, so a player can jump to a moment by looking it up. A raw elementary stream carries none. Run ffprobe over an .m2v and every frame reports pts_time=N/A, and the file as a whole reports duration=N/A and bit_rate=N/A. The decoder synthesises timing from the sequence header's frame-rate code, and a seek becomes an estimate against byte position.

The consequence is measurable. Encoding a 25 fps test clip to .m2v and asking for the frame at 0 seconds returned the twelfth frame, roughly half a second in; asking for 1 second returned that same twelfth frame; asking for 2 seconds returned frame 36 rather than frame 50. The same encode written as a muxed .mpg landed exactly where it was asked to.

So on this format:

• If the exact frame matters, use "Multiple Screenshots" instead. That mode samples the decoded stream at a fixed interval with no seeking involved, so the spacing between captures is exact even though the absolute offset is still approximate. Set a Capture Rate of one frame every second or two, take the sheet of bitmaps, and pick the one you wanted.
• If you only need a representative still, leave Specific Frame alone and type a time a second or two before the moment you actually want.
• If you need frame accuracy repeatedly, give the stream a container first. M2V to MP4 writes real timestamps and an index, after which timecode seeking behaves normally everywhere.

The decode itself is exact either way: the frame comes out of the stream as a lossless PNG first, and only then is it encoded into the bitmap you asked for. Every question about how the still looks is a question about the preset.

What Each Quality Preset Writes

BMP stores pixels literally, so there is no quality-versus-size dial in the JPEG sense. What the preset selects is the bitmap encoding itself:

High is not 24-bit truecolour. It is the 16-bit RGB 555 layout — five bits each for red, green and blue — which Microsoft's BITMAPINFOHEADER reference documents as the fixed meaning of 16 bits per pixel under BI_RGB. The arithmetic is exact rather than approximate: a 704 x 480 frame taken at High measured 675,894 bytes, which is precisely two bytes for each of the 337,920 pixels plus the 54-byte header.

One practical warning: the preset that is preselected on other image targets is one of the entries a BMP target hides, so the box does not already hold a usable value. Open the list and choose High, Medium or Low deliberately.

What an Elementary Stream Does and Doesn't Carry

Common Errors and How to Fix Them

• "The still is not the frame I asked for." Timestamp seeking on a raw elementary stream is approximate. Use Multiple Screenshots, or convert to a real container first.
• "Everything moving has comb lines through it." The source is interlaced, which most DVD-era MPEG-2 is, and there is no deinterlacing stage here. Choose a frame from a static shot, or deinterlace in an editor and re-export.
• "People look narrow or stretched." Anamorphic DVD essence. Set Image resolution to Width x Height and type the display dimensions you want — 640 x 480 or 854 x 480 for NTSC, 1024 x 576 for PAL widescreen.
• "The bitmap is huge." High writes two uncompressed bytes per pixel. A 720 x 576 frame is about 830 KB before you have done anything wrong. Choose Medium or Low, or reduce the resolution.
• "Skies and skin tones band." RGB 555 gives you 32,768 colours instead of 16.7 million, and the palette presets are harsher still. For true-colour lossless take the frame as M2V to PNG.
• "I got hundreds of files." Multiple Screenshots at one frame per second on a long stream. Slow the Capture Rate down.
• "The conversion failed outright." Partial or truncated elementary streams are common, because nothing in the format marks where the file should end. If a player also refuses it, the file is damaged rather than unsupported.

Frequently Asked Questions

Why does my M2V have no duration or timeline in any player?

Because there is no container to store one. An .m2v is only the coded picture data; duration, bitrate, timestamps, chapter marks and stream indexes all live in a container layer that a bare elementary stream does not have. ffprobe reports duration=N/A and bit_rate=N/A for the file and pts_time=N/A for every individual frame. Players cope by counting frames and assuming the sequence header's frame rate, which is why the scrub bar behaves oddly and why seeking is approximate.

Which frame do I get if I leave Time (seconds) at zero?

Roughly the beginning, but not reliably the first frame. In a measured test on a 25 fps stream, a request for 0 seconds returned the twelfth decoded frame — about half a second in — and a request for 1 second returned the same one. If you specifically need frame zero, take the whole stream through Multiple Screenshots at a fast Capture Rate and use the first bitmap it writes, or wrap the stream in a container first.

Does the High preset give me a 24-bit bitmap?

No. High writes an uncompressed 16-bits-per-pixel bitmap in the RGB 555 layout — five bits for each channel, 32,768 possible colours. It is the highest-fidelity option this page offers and it genuinely is uncompressed, but it is not the 24-bit truecolour BMP some software specifically requires. If a downstream tool insists on 24-bit, take M2V to PNG and set the depth in an image editor.

What is the difference between Medium and Low?

Both write an RLE-compressed palette bitmap with Floyd-Steinberg dithering, which spreads quantisation error into neighbouring pixels so banding reads as fine noise rather than hard steps. Medium keeps 256 colours, writes the classic BMP3 header form and applies a light selective blur before quantising, which helps flat graphics quantise cleanly. Low drops to 86 colours and tags the file at 150 pixels per centimetre. On photographic footage the difference is easy to see.

Why is my DVD frame 720 pixels wide but looks like 16:9 in a player?

DVD MPEG-2 stores a 720 x 480 or 720 x 576 raster and flags the display aspect separately, so the player stretches it at showtime. A BMP has no field a receiving application will honour for that, and the pipeline writes the stored raster rather than the display shape. Set Image resolution to Width x Height and enter the dimensions you actually want the picture to be.

Can I get every frame of the stream as a bitmap?

You can sample it densely, but not at full frame rate. Multiple Screenshots tops out at one capture every tenth of a second, which on 25 or 29.97 fps material is roughly one frame in three. At uncompressed sizes that is usually a mercy — a minute of 720 x 576 at High would be well over a gigabyte of bitmaps. Pick the slowest Capture Rate that still shows you what you need.

Will the still be as sharp as the frame inside the M2V?

Yes, as far as the decode goes. The frame is extracted as a lossless PNG before anything else touches it, so no detail or sharpness is lost in reading the stream. What you give up happens at the encoding step and is purely colour resolution: 32,768 colours on High, 256 on Medium, 86 on Low. Geometry and edge definition are untouched unless you also resize.

Is BMP a sensible choice at all, or should I use something else?

It is the right choice when something specific demands it — machine-vision software, an embedded display driver, a printer utility, a forensic workflow or Windows software old enough to trust nothing else. For anything else the trade is poor: BMP is uncompressed or crudely palettised, carries no metadata beyond dimensions and pixel density, and is far larger than the alternatives. M2V to PNG gives you true-colour lossless at a fraction of the size, and M2V to JPG is the sensible option for sharing.

How are my files handled during conversion?

Your .m2v is uploaded over an encrypted connection, decoded on our servers, and both the upload and every bitmap it produced are deleted automatically after a few hours. Nothing is shared or made public, there is no sign-up and no watermark. Elementary streams from DVD authoring are often large for their running time, so if you only want a still from one section it is quicker to cut it down first with video cutter.