ORF to MPEG-2 Converter

ORF to MPEG-2: What Goes Wrong and How to Avoid It

Most ORF-to-MPEG-2 conversions succeed on the first try and then disappoint on the second step: the file is much bigger than expected, or it refuses to play on the device it was made for, or the picture sits between two white bars. None of those are failures — they are all predictable consequences of feeding a 20-megapixel Olympus raw to a video codec designed in 1995 for standard-definition television. This page walks through each of them, and through the settings that actually change the result.

How to Convert ORF to MPEG-2

1. Upload Your ORF File: Drag the .orf onto the page or click "+ Add Files". Only .orf is accepted. 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.
2. Set the Video resolution first: This is the decision that determines everything else. "Keep original" encodes at the decoded raw dimensions; the height presets scale while keeping the Four Thirds 4:3 shape; "Width x Height" takes exact numbers when a disc or ingest specification demands them.
3. Set Image Duration and Background Color: Image Duration is how long the picture is held, from a single frame at 60 fps up to ten seconds per image, defaulting to five. Background Color fills any bars added when the output frame is a different shape from the raw and starts on White.
4. Convert and Download: Click Convert. With several raws queued, Merge strategy decides between one concatenated clip and one file per raw.

Common Problems and What Causes Them

• "The file is far bigger than I expected." At "Keep original" a modern Olympus raw is encoded at four thousand pixels wide. Our OM-D E-M1 Mark II sample decodes to 5240 × 3912, is normalised down to 4092 × 3056, and produced a 3,446,784-byte file for a five-second clip. The same frame at 770 × 576 came back at 55,296 bytes. How large that gap is depends on how much detail the picture holds, but resolution is always the fix.
• "It plays on my computer but not on the player I made it for." MPEG-2's Main Profile at High Level stops at 1920 × 1152. A four-thousand-pixel-wide MPEG-2 stream is a valid bitstream but no hardware decoder implements a level that large. Software players such as VLC will open it; DVD players, DVB and ATSC decoders and older signage boxes will not.
• "There are white bars beside my picture." The Background Color option starts on White on image sources and fills the pillarbox area created when a 4:3 raw is placed inside a 16:9 output frame. Pick Black from the dropdown, or leave Video resolution on "Keep original" or a height preset — both follow the raw's own aspect ratio and add no bars at all.
• "Changing the Quality Preset does nothing." Correct, and expected. MPEG-2 has no constant-quality mode, so the preset falls back to scaling the source video's bitrate and a photograph has none. Constant Quality and Constraint Quality are both rewritten as plain bitrate requests and then replaced by a floor derived from the output frame size, so the "Max bitrate" figure you type under Constraint Quality is discarded too.
• "The video has no sound and I can't add any." Image-to-video conversions are silent by design; the Audio Codec group is not rendered on this page and the output carries a single video stream. Add audio later in an editor.
• "I fed the result into an audio converter and it failed." It has no audio stream to convert, so the job errors rather than returning a silent file. That is the correct behaviour, not a bug.

Where the Pixels Go

ORF Reference

Frequently Asked Questions

Why is my ORF scaled down before it is even encoded?

Because the image normalisation step clamps the long edge to 4096 pixels. Recent Olympus and OM System sensors exceed that, so a 5240 × 3912 decode becomes 4092 × 3056. The reduction is proportional, so nothing is cropped and the aspect ratio is unchanged. Older Four Thirds bodies with smaller sensors are under the ceiling and pass through untouched.

Which resolution should I pick for a DVD?

720 × 480 for NTSC regions or 720 × 576 for PAL, entered through "Width x Height". Those are the only two frame sizes the DVD-Video specification allows, and MPEG-2 is the codec it requires. The disc structure itself — the VIDEO_TS folder, menus, IFO and BUP files — is produced by your authoring application afterwards; this converter writes the video stream.

Will a portrait shot come out sideways?

It should not. The step that prepares every image for encoding applies the orientation flag by default rather than ignoring it — we checked that behaviour directly by handing the same step a file tagged with orientation 6, which came back rotated upright rather than on its side. One side effect is worth expecting: a portrait frame is fitted inside a box computed from the unrotated dimensions, so it arrives with fewer pixels than a landscape frame from the same body.

How long is the video, and can I change the frame rate?

Length is the number of raws multiplied by the Image Duration. There is no frame-rate control on this page — the encode is fixed at one frame per second with B-frames disabled, because identical frames gain nothing from bidirectional prediction and B-frames introduce a start-time offset at low frame rates. Use the Image Duration dropdown to change the pace.

Does the Video Codec dropdown let me choose something other than MPEG-2?

Not on this page. The group renders, but MPEG-2 is the only selectable entry — every other codec is hidden because the output extension is .mpeg2. If you want a different codec, use the converter for that target instead, for example ORF to MP4.

Is the highlight headroom in the raw preserved?

No. The conversion demosaics the sensor data, drops to 8-bit 4:2:0 and applies lossy compression. Nothing survives that you could later use for exposure recovery. Develop the raw first if that matters — ORF to TIFF keeps considerably more of the original tonal range.

Why does an .mpeg2 file not open in my browser?

Because browsers do not implement MPEG-2 decoding, and an MPEG program stream is not a format any of them accept for <video>. The extension is also non-standard — the underlying container is an MPEG program stream, and renaming the file to .mpg improves its odds with desktop players. For anything web-facing, convert to MP4 instead.

What does the Merge strategy setting do with a single file?

Nothing visible. It only matters once more than one raw is queued, at which point "Merge images" concatenates them into a single MPEG-2 slideshow in queue order and "Video per image" returns a separate clip for each. With one file the result is the same either way.