AVCHD to PPM Converter

Convert AVCHD to PPM

This is frame extraction, not video conversion. An AVCHD file is H.264 video inside an MPEG-2 transport stream — what consumer HD camcorders have recorded since 2006 — and a PPM is a single uncompressed RGB bitmap with a fifteen-byte header and no compression of any kind. What comes back is a still, or a set of stills, taken out of the footage.

People reach for PPM rather than PNG or JPEG for one reason: the file needs no decoder. A PPM is three integers and then raw pixel bytes, which makes it the format of choice when the next step is a script, an image-analysis pipeline or code that would rather read bytes than link a codec library. If a human is going to look at the picture, AVCHD to JPG or AVCHD to PNG produce files a few percent of the size.

How to Convert AVCHD to PPM

1. Upload Your AVCHD File: Drag the .avchd file onto the page or click "Add Files". Uploads travel over an encrypted connection and are processed on our servers. Camcorder files are often several gigabytes, so this step usually takes longer than the conversion itself.
2. Choose Frame Selection: Specific Frame grabs one frame at the Time (seconds) you enter, starting at 0. Multiple Screenshots walks the whole clip at a Capture Rate — one frame per second by default, from ten frames a second down to one every ten seconds — and writes a PPM for each. Unlike raw video streams, AVCHD carries proper timestamps and regular keyframes, so seeking to an arbitrary time genuinely works here.
3. Set Image Resolution — Do Not Skip This on Anamorphic Footage: Image resolution opens on Keep original. For 1440×1080 camcorder footage, set Width x Height to 1920x1080 instead; the reason is below and it matters more than it sounds. Resolution Percentage, Preset Resolutions, Width and Height are the other modes.
4. Set Bit Depth and Convert: Bit Depth offers 8-bit (Recommended), 16-bit (High Precision) and 1-bit (Black & White). Click Convert and download. No sign-up, no watermark; uploads and outputs are deleted automatically after a few hours.

Anamorphic AVCHD and a Format With No Aspect Ratio

A very large share of AVCHD footage is stored anamorphically: the picture occupies a 1440×1080 pixel grid but carries a 4:3 sample aspect ratio, so a player stretches it out to a 16:9 display shape. That flag is what makes the footage look right in a video player.

PPM has nowhere to put that flag. Its entire header is a magic number, a width, a height and a maximum sample value — no aspect ratio, no colour profile, no metadata of any kind. We confirmed the loss end to end: the frame extracted from a 1440×1080 anamorphic clip still reported a 4:3 sample aspect ratio as a PNG, and the same frame written to PPM reported no aspect ratio at all. Anything that opens that PPM sees a 1440×1080 square-pixel image, which means the picture looks horizontally squashed.

The general rule for this pipeline is that leaving resolution alone preserves the source's pixel aspect, while any resize flattens it to square pixels. For most targets that is a hazard; for PPM it is the fix, because the format cannot carry the flag anyway. Entering the display dimensions yourself bakes the correct geometry into the pixels.

Interlacing: What You Get From 1080i Footage

Plenty of AVCHD camcorders record 1080i, and a still frame from interlaced video contains two fields captured at slightly different moments. There is no deinterlacing anywhere in this pipeline — no filter is applied at any stage — so what lands in the PPM is exactly the interleaved frame the decoder produced. On a static shot you will not notice. On anything with motion you get the classic comb-tooth pattern along moving edges, and it is permanent once written to an uncompressed bitmap.

There is no setting here that fixes it. If a clean still from interlaced footage matters, deinterlace the video first in an editor or with a dedicated tool, then extract the frame. Choosing a moment with little movement in it is the cheap workaround.

Bit Depth and File Size, Measured

PPM stores raw samples, so its size is arithmetic rather than a compression outcome. We wrote the same 1440×1080 AVCHD frame at all three depths:

Two of those three settings therefore change the picture without changing the size. 16-bit is worth choosing only when the receiving tool insists on 16-bit samples — consumer AVCHD is 8-bit H.264, so the extra bits carry no information from the camera. 1-bit posterises hard and saves nothing at all in this container. Scaling the image down is the only control here that meaningfully shrinks the output.

Frequently Asked Questions

Why is a single frame nearly 5 MB?

Because there is no compression involved. A 1440×1080 frame is 1,555,200 pixels, three bytes each, plus a short header — the size is fixed by the pixel count and nothing else. The same frame as JPEG might be 300 KB and as PNG perhaps 1.5 MB. That is not PPM doing something wrong; it is the trade the format exists to make, giving up size in exchange for needing no decoder at all.

Can I extract a frame from the middle of a long clip?

Yes. AVCHD is a transport stream with regular keyframes and real timestamps, so Specific Frame with a time value works properly here — we confirmed extraction at several offsets through a transport-stream source. That is worth calling out because raw video streams behave very differently: on a bare elementary stream there is nothing to seek with and a non-zero time returns nothing at all. AVCHD has no such problem.

How do I get a whole sequence of frames rather than one?

Use Multiple Screenshots and set the Capture Rate. One frame per second is the default; the fastest setting is ten frames per second and the slowest is one frame every ten seconds. Each captured frame becomes its own PPM. Do the arithmetic first — at 1440×1080 and ten frames a second, one minute of footage is roughly 2.8 GB of uncompressed bitmaps.

Is the extracted frame identical to what the camcorder recorded?

Identical to what the H.264 decoder produces from it, yes. The frame is pulled out as lossless PNG and then written to PPM with no further compression, so no pixel values change along the way. What you cannot undo is the compression the camcorder already applied when recording — H.264 artifacts in the original are decoded faithfully and preserved exactly.

Does the PPM keep the shooting date, camera model or GPS data?

No. AVCHD records a good deal of that, but PPM has no metadata section — the header stops after the maximum sample value. Timestamps, camera settings, colour profiles and location are all dropped, and metadata is stripped during the write in any case. If you need that information preserved alongside the image, extract to a format that can hold it, such as PNG or TIFF, and keep the original clip.

Why does my extracted still look squashed compared with the video?

Because the footage is anamorphic and PPM cannot store the flag that told your player to stretch it. The pixels are correct; the instruction to display them wider is gone. Set Width x Height to the display dimensions — 1920x1080 for typical 1440×1080 AVCHD — and the conversion bakes the right geometry into the pixel grid so it looks correct everywhere.

What is the difference between P6 and P3 PPM files?

They are the binary and ASCII forms of the same format. P3 writes every sample as decimal text, which a human can read but which inflates the file three to four times; P6 writes raw bytes after the header. This converter produces P6, which is what virtually all software expects. Open an output in a text editor and the first two characters confirm it.

How are my files handled while they are on your servers?

The .avchd file is uploaded over an encrypted connection, processed on our servers, and both the upload and the extracted frames are deleted automatically after a few hours. There is no account to create, no watermark, and files are never shared or made public. Because camcorder files are large, the real constraint is upload time — if you only need a still from one moment, cutting a short section first with the Video Cutter, which accepts .avchd directly, is much faster than uploading an hour of footage.