Initializing... drag & drop files here
Supports: RW2
RW2 is what a Lumix sensor recorded before the camera made any decisions — an undeveloped mosaic with the white-balance multipliers stored alongside it. SWF is Shockwave Flash: a 1990s multimedia container that Adobe stopped supporting on 31 December 2020 and began actively blocking content in on 12 January 2021.
Those two facts sit oddly together, and they should. Nobody converts a modern raw photograph to Flash for viewing. The reason this route exists is that some legacy authoring tools, kiosk systems and archival ingest pipelines accept exactly one video container and it is .swf. If that is you, this page explains what you will get, where the pixels go, and which single setting decides whether the result looks like a photograph.
| Property | Value on this route |
|---|---|
| Container | SWF, signature FWS (uncompressed Flash movie) |
| Video codec written | FLV / Sorenson Spark by default; MJPEG is the other entry in the Video Codec list |
| Codec lineage | Sorenson Spark is an H.263 derivative, standardised in the late 1990s |
| Audio | None on this route — a still image has no audio, and the audio controls are suppressed |
| Frame rate | 1 frame per second, with B-frames disabled |
| Clip length | Set by Image Duration, from 1/60s to 10 seconds per frame; opens at 5 seconds |
| Maximum frame size | 4096 pixels on the long edge — every still is normalised to that ceiling before encoding |
| Runtime status | Flash Player support ended 31 December 2020; content blocked from 12 January 2021 |
| What still opens it | Desktop players built on ffmpeg, such as VLC and mpv. Not browsers. |
This is the part specific to raw sources, and it is not a setting you can turn off. Before any video encoder sees your picture, the pipeline renders it to a clean intermediate image and clamps the long edge to 4096 pixels, preserving aspect ratio and rounding to even dimensions.
Measured on a Lumix DC-FZ80 file: the raw decoder emitted 4912 x 3688, and the frame that reached the encoder was 4094 x 3074. That is a 17% linear reduction, about 30% of the pixels. On a 24-megapixel or 40-megapixel body the cut is proportionally larger.
Nothing about Video resolution changes that ceiling. Keep original means "the clamped source size", not "the sensor size". If print-resolution pixels are the requirement, a video container is the wrong destination and RW2 to TIFF or RW2 to JPG is what you want.
.rw2 on the page or click "+ Add Files". Queue several frames if you want them in one movie.FLV is a quantizer-scale codec with no quality-target mode. Quality Preset works by scaling the source clip's bitrate, and a photograph has no bitrate to scale — the calculation resolves to zero and gets clamped to the pipeline's 48 kbps floor. The encoder then spends the entire job in buffer underflow, trying to fit a four-megapixel frame into a budget meant for a postage stamp.
Measured on the same clamped 4094 x 3074 Lumix frame, held for the default 5 seconds:
| File Compression setting | Output size | Fidelity to the developed frame |
|---|---|---|
| Quality Preset — the shipped default | 413,678 bytes | 28.04 dB PSNR |
| Constant Quality, CRF slider at its opening value of 5 | 1,263,761 bytes | 32.33 dB PSNR |
| Constant Quality, CRF slider at 2 | 3,465,847 bytes | — |
Three times the bytes and about four decibels of fidelity, from one dropdown. There is no size ceiling on SWF forcing you to accept the default, so on a still there is very little reason not to move the slider down toward 2 or 3.
Because the source is a photograph. Image-to-video routes suppress the audio controls entirely and write a single video track — there is nothing in a raw sensor capture that could become an audio stream. If you need a soundtrack over the held frame, you will have to add it in an editor afterwards, working from an SWF or from a more tractable container such as RW2 to MP4.
The one your Lumix recorded at capture. A raw file stores white balance as multipliers rather than baked pixels, and the decoder applies the camera's own values, then demosaics to 8-bit sRGB. There is no neutral or auto interpretation layered on top, and no control for it on this page. If the body guessed wrong under mixed lighting, develop the .rw2 in a raw editor first and convert the exported image instead.
FLV for almost any still. Sorenson Spark predicts each frame from the previous one, so five identical frames cost roughly one frame plus a rounding error. MJPEG stores every frame as an independent JPEG, which for a held still means paying full price five times over. Choose MJPEG only when something downstream specifically requires independently decodable frames — and note that it re-ranges the quality controls, because MJPEG runs on its own 2-to-31 quantizer scale.
A still has no frame rate of its own, so the encoder runs at 1 fps and holds the picture for the duration you chose. Five seconds means five identical frames, not 125, which is why the file stays small on FLV. B-frames are switched off for the same reason: they exist to predict motion between differing frames, and there is none here. If you want the still to behave like one frame inside a 30 fps timeline, pick 1/30s (single frame at 30fps) from the Image Duration list.
Yes. Upload them together and leave Merge strategy on Merge images; each rendered frame is held for the Image Duration you set, in upload order, in a single movie. Video per image produces one .swf per photo. Neither mode adds transitions, motion or audio, so the result is a run of held stills — which is exactly what a kiosk loop or a slate sequence usually wants.
The Flash runtime does not — Adobe ended support on 31 December 2020 and has blocked Flash content since 12 January 2021, and every major browser removed the plugin before that. What does still work is playback through desktop software built on ffmpeg: VLC and mpv will open the file and decode the video stream inside it. So the bytes remain readable even though the platform they were made for is gone.
No. Camera model, ISO, shutter speed, aperture, focal length and capture date do not survive. Video containers have no per-frame Exif equivalent, and the raw decoding stage drops those fields before the encoder ever sees the picture — a measured re-encode of a Lumix raw through the same image stage came out with only an orientation flag and a resolution field. Keep the .rw2 if the shot data matters.
They are the same family but not interchangeable filenames. RW2 has been Panasonic's raw extension since the mid-2000s; the container is a little-endian TIFF-like header carrying Panasonic's own magic number in place of TIFF's, wrapping a Panasonic-specific compressed payload. Older Lumix bodies used .raw, and some Leica models built on Panasonic hardware write .rwl with the same internals. If a very new body's file fails to decode, suspect a new compression variant rather than the extension.
There is no per-file setting that caps it, and an 18-to-24 megapixel Lumix raw of 20 to 40 MB is routine here. The practical constraint is upload size and time over your connection, not the converter. Bear in mind that whatever you send is clamped to 4096 pixels on the long edge before encoding, so a bigger sensor buys you nothing on this particular route.
Your .rw2 is uploaded over an encrypted connection, decoded and encoded on our servers, and both the upload and the .swf it produced are deleted automatically after a few hours. Files are never shared or made public, no account is needed, and nothing is watermarked.