RAF to 3GP Converter

RAF to 3GP: a Fujifilm Raw Inside a Feature-Phone Container

.raf is Fujifilm's raw format — the undeveloped capture from an X-series or GFX body. .3gp is the 3GPP file format from 3GPP TS 26.244, an ISO base media file format profile written for the days when video had to squeeze through a 3G data channel onto a small handset screen. Converting between them is a still-to-video job: the raw is developed into a picture, held on screen for a duration you choose, and encoded as a short silent clip. It is a niche conversion, and worth doing only when something specifically needs .3gp. What follows is the arithmetic of what happens to the photograph, and how to drive a batch of raws into one clip.

How to Convert RAF to 3GP

1. Upload Your RAF Files: Drag and drop onto the page or click "+ Add Files". "Merge strategy" is preselected on "Merge images", so a whole queue becomes one clip in the order it was added; switch to "Video per image" for a separate .3gp per raw.
2. Set the Image Duration: The dropdown runs from a single frame at 60 fps through 0.1, 0.2, 0.3 and 0.5 seconds up to 10 seconds per image, and starts at 5 seconds. This is the whole timing model — sub-second values give you a timelapse, multi-second ones a slideshow.
3. Pick the Video Codec and Resolution: H.264 is preselected, with H.263, MPEG-4 and Xvid as alternatives. "Video resolution" starts on "Keep original" and also offers Fixed Resolutions, Preset Resolutions, Width, Height and Width x Height; "Background Color" is white by default and fills any bars a fixed frame size leaves around your picture.
4. Convert and Download: Click "Convert" and take the clip. Photographs carry no sound, so there is no audio group here and no audio track in the output. No sign-up, no watermark.

The Downscale, in Numbers

Every image is normalised before encoding, with the long edge capped at 4096 pixels and both dimensions rounded to even numbers. Fujifilm sensors are all comfortably above that, so the cap is doing real work on every conversion:

Pick H.263 and the reduction goes much further, because that codec accepts only five picture sizes — 128 x 96, 176 x 144, 352 x 288, 704 x 576 and 1408 x 1152. The pipeline snaps to the largest one that fits, ignoring whatever you set under Video resolution, so a Fujifilm frame lands on 1408 x 1152. That is an 11:9 shape and your photograph is 3:2, so the picture is squeezed horizontally as well as shrunk. H.264 — the preselected codec, and perfectly legal inside a 3GP file — keeps the full 4096-pixel frame and its proper proportions.

Turning Several RAFs into One Clip

With "Merge strategy" on "Merge images", output length is simply the number of images multiplied by the Image Duration. Stills are encoded at one frame per second, so the frame count matches the seconds:

There are no transitions between images — each still replaces the one before it. Order follows the order the files were added, so name your files so they sort the way you want before uploading. And note that a 3GP file is a poor slideshow container in 2026: if the clip is for anything other than legacy hardware, RAF to MP4 produces the same slideshow at the full 4096-pixel frame in a container that phones, browsers and televisions actually favour.

What the Render Bakes In

Fujifilm raws are prized for their colour rendering and their headroom, and the video path spends both:

• The development is fixed at conversion time. The raw is demosaiced with the white balance the camera recorded and written as 8-bit sRGB. Film-simulation choices, highlight recovery and any change of colour temperature belong in the original file — set them in your raw editor and convert the exported image instead if the look matters.
• Extra bit depth does not travel. Fujifilm sensors read out well above 8 bits per channel; the developer used here writes 8, and the truncation happens as the file is read rather than as the video is encoded. There is no setting that changes it.
• Metadata is left behind. Camera body, lens, exposure, film simulation and GPS live in the raw's EXIF block. A 3GP video track has nowhere to record them, so they are gone from the clip.

If the goal is the photograph rather than a video, stop here: RAF to JPG gives you a viewable full-resolution picture, and RAF to MKV is the better home if you want the still as video with a wider choice of codecs.

Frequently Asked Questions

Why does my 3GP clip have no sound?

Because there was none to begin with. A raw photograph carries no audio, and this route does not invent a silent track to fill the gap — the audio controls are not shown at all and the output contains only a video track. If the clip needs music or narration, add it in a video editor afterwards.

How big is the frame if I leave the resolution on "Keep original"?

4096 pixels on the long edge, not your sensor's full width. Every image is normalised to that cap before encoding, so a 7728 x 5152 X-T5 frame arrives as 4096 x 2730 and a 11648 x 8736 GFX frame as 4096 x 3072. "Keep original" means "keep the aspect ratio and do not shrink further", not "keep every pixel". Full sensor resolution is only available on still targets such as RAF to JPG.

Does X-Trans raw convert differently from a Bayer raw?

Not in a way you have to think about here. Most Fujifilm X-series bodies use the X-Trans colour filter array rather than a conventional Bayer pattern, and the GFX medium-format line uses Bayer; the raw developer handles both and produces an ordinary RGB image either way. What you should expect is a neutral development rather than a match for a particular film simulation, since the simulations are applied by Fujifilm's own processing rather than stored as finished pixels in the raw.

Can I make a timelapse from a bracketed or interval-shot sequence?

Yes. Upload the frames together, leave "Merge strategy" on "Merge images", and set "Image Duration" to one of the sub-second values — 1/60, 1/30, 1/24, 0.1, 0.2, 0.3 or 0.5 seconds. At 1/24 second per image, 240 frames become a ten-second clip. Bear in mind that 3GP is a poor delivery format for anything with motion in it; use RAF to MP4 for a timelapse you actually intend to watch.

Which codec should I choose for an old handset?

H.263 if the device genuinely refuses anything else, H.264 otherwise. H.263 is what the earliest 3GP-capable phones were built around, and the cost is severe: the frame is forced to at most 1408 x 1152 with the aspect ratio not preserved, and the "Quality Preset" control has nothing to scale on a still image, so you would need "Constraint Quality" and an explicit maximum bitrate to control the result. H.264 is preselected, is standard inside 3GP, and keeps the frame intact.

What does the Background Color option actually change?

It fills the area your photograph does not cover. With "Keep original" or a height preset the video frame matches the picture's own 3:2 shape and you will never see the colour. Choose a fixed frame size — a 4:3 or 16:9 target, or your own Width x Height — and the leftover bars are painted in this colour. It is white by default, which is bright on a dark screen; black usually looks better.

Why is my file huge for such a short clip?

Because it is a 4096-pixel-wide frame. Held stills compress extremely well from the second frame onward, but the opening frame is a full-detail photograph and has to be encoded at that size. Dropping "Video resolution" to a preset height — 480p or 720p — shrinks the file dramatically with no visible loss on a phone-sized screen, and is the setting to reach for before touching quality.

Can several RAFs go into one file, or do I get one clip each?

Either. "Merge strategy" starts on "Merge images", which concatenates every uploaded raw into a single clip with each photograph held for the same duration. Switching to "Video per image" produces one .3gp per input file instead, which is what you want when the raws are unrelated or when a downstream device expects individual clips.

How are my files handled, and how long do you keep them?

Your raw is uploaded over an encrypted connection, developed and encoded on our servers, and deleted automatically a few hours after conversion — no sign-up, no watermark, never shared or made public. Fujifilm raws are large — a 40 MP X-T5 file runs to tens of megabytes and a GFX capture considerably more — so on a batch the upload takes far longer than the encode.