Recover Sony XAVC video from deleted or formatted memory cards.

There are many free disk-imaging tools and step-by-step guides available for Windows and macOS. Search Google for how to create a full-sector image of an SD card or CFexpress card and use a tool you are comfortable with. The important part for Sony XAVC Recovery is that you image the entire physical card, not just a partition or mounted volume.

Because recovery is much less exciting when the source stops changing underneath you. A fixed copy is the safer target. The program reads the image without writing to it, and you can analyze it again without repeatedly touching the original card. The card gets one job: give us the best copy of its sectors it can.

Because the software grew out of real Sony recovery cases and the reverse-engineering work needed to solve them. Publishing the tool also gives unusual cards a chance to expose layouts we have not seen yet. iCube Development does paid data recovery work, but the software itself is free. No licence key, subscription or paid unlock.

No. Analysis and recovery run on your computer. The recovery program does not upload your card image or recovered video to iCube Development.

No. Sony XAVC Recovery is built specifically around Sony XAVC recording structures. For GoPro, Nikon, Canon and other camera formats, search Google for recovery software for that camera or recording format. There are plenty of free recovery tools for other camera formats. Use one built for that job rather than asking a Sony-specific tool to pretend.

The tool does not use a camera-model whitelist. It looks for supported Sony XAVC recording structures in the card image. That means a body does not have to appear on a compatibility list to work, but the recording still needs to use a supported Sony XAVC MP4 layout.

Yes, provided you can make a readable full-sector image and the card contains a supported Sony XAVC MP4 recording layout. The recovery logic cares about the surviving recording structures, not the logo printed on the memory card.

Sometimes. A format can remove filesystem bookkeeping while leaving recording data behind, but an in-camera erase may also cause the card controller to unmap or erase blocks. The menu option alone does not tell you which happened. See Deleted or formatted?

That is normal. Card manufacturers advertise capacity in decimal gigabytes, while many operating systems and recovery tools display binary gibibytes. A nominal 256 GB card is roughly 238 GiB. Nothing went missing in the conversion.

It means the program could prove and preserve a usable portion of the recording, but a later video or audio extent could not be established safely. Rather than invent the missing part or throw away everything, Version 2.2 can keep the proven portion as .partial.mp4.

A diagnostic bundle is the best thing to create when the program finds the wrong number of recordings, finds nothing useful, marks unexpected items for review, produces a file that does not play correctly, or encounters a Sony layout it does not yet understand.

How to create one: select the source image as you normally would and reproduce the problem first. Then choose File > Create Diag. Bundle.... If you are testing an unformatted reference image, add that reference before creating the bundle so its characterization is recorded too.

The bundle records the program version, source identity, analysis results, Sony structures found, mapping and reconstruction decisions, recovery traces and other technical details needed to understand what the engine did. If a reference image is attached, its characterization and identity are recorded as well. The full card image and recovered videos are not placed in the diagnostic bundle.

The point is not simply to collect a log. The bundle lets iCube Development compare what the program expected with what was actually present on that card, which is how unsupported layouts and incorrect reconstruction assumptions can be isolated without guessing.

Extended Translation Recovery is an optional recovery path for cards whose reported sector addresses may not match where the useful data is really coming from. This can happen with damaged, counterfeit, fake-capacity or failing flash media. A card can appear readable while returning repeated, wrapped, shifted or otherwise abnormal data for parts of its address range.

Enable it with Advanced > Enable Extended Translation Recovery. It is normally left off because a healthy card image should not need alternate address translations. When enabled, the program tests additional translation possibilities and still requires the resulting Sony recording structure to satisfy the normal recovery checks. It does not simply accept more matches because they happen to contain video.

Use it when the normal analysis looks internally inconsistent, the sector map suggests repeated or strangely displaced data, the card reported a suspicious capacity, or the media is known to be counterfeit or failing. It can take longer because the program has more candidate mappings to test. It does not repair erased flash and cannot recover sectors that the card never returned in the image.

This is optional and is mainly for a Sony recording layout the program does not yet understand well enough from the damaged image alone. Choose File > Add Unformatted Image... and select a full-sector image of a healthy, unformatted Sony card that represents the same recording setup as closely as possible.

The best reference is from the same camera model and recording mode, including the same codec family, resolution, frame rate and related recording settings. It does not need to contain the missing footage. Its job is to provide a known-good example of how that Sony setup writes its recording structures and transitions to the card.

The reference is opened read-only and is used for comparison. Its video is not mixed into the damaged card's recovery and it cannot replace missing source bytes. When a diagnostic bundle is created, the tool can record the reference image's characterization and identity without placing the reference image itself in the bundle.

You do not need a reference image for a normal supported recovery. Add one when analysis suggests an unknown or unfamiliar layout, or when iCube Development asks for one while investigating a diagnostic bundle.

This option is for AVC/H.264 recordings that change from one physical region of the card to another and cannot be completed by the normal direct mapping path. A recording may be logically continuous even though its next group of frames is physically far away on the card.

When enabled, Sony XAVC Recovery follows the surviving Sony and MP4 structure across the break, calculates what the next AVC samples must look like, searches for a physical location that satisfies that structure, and accepts the transition only when there is a unique physical match. It then checks the rebuilt transition by decoding it in context with the surrounding video.

This is not blind fragment scanning and it does not accept a candidate merely because it contains valid H.264 data or produces a playable picture. If more than one physical location fits, the transition is not treated as proven. The option is therefore useful for difficult fragmented AVC cards, but it should not be enabled simply to make the program return more footage.

It applies to AVC/H.264. HEVC/H.265 uses its own reconstruction paths.

Disk Drill and Sony XAVC Recovery can both use camera-aware information to recover video that ordinary file carving misses, but they are built around different goals.

Disk Drill's Advanced Camera Recovery is a broader recovery product. Its Sony-aware recovery can locate camera metadata and matching video fragments across a card and reconstruct footage that a normal contiguous scan may miss. That broader approach can return more candidate footage on some cards.

Sony XAVC Recovery is narrower and deliberately conservative. It is built specifically around supported Sony XAVC MP4 structures. A fragment is not accepted merely because it looks like the right codec, is near the expected location, fits one boundary, or produces video that plays. The engine checks the surviving Sony metadata, the recording's expected sample sizes and order, the physical layout, and whether exactly one physical location satisfies those constraints. The reconstructed result is then decoder-checked in context.

The practical difference is the acceptance standard. If two physical locations can satisfy the available evidence, Sony XAVC Recovery does not silently choose one. The mapping remains unresolved or is marked for review. This can mean another recovery program returns more footage while Sony XAVC Recovery returns less, because this tool prefers to leave a fragment unresolved rather than guess which copy belongs to the recording.

When Sony XAVC Recovery reports a recording as fully recovered, it is not creating replacement video. It does not transcode the recording, interpolate missing frames, synthesize audio, conceal bad sections or make a damaged stream look playable. The accepted video and audio samples are copied from the source image using the original bytes, exact sample sizes and proven playback order.

That distinction matters. A successfully reconstructed recording contains the original Sony picture and sound data, not an approximation of it. The recovery process does not intentionally drop frames, substitute audio or re-encode the media. If the engine cannot prove a required piece of the recording, it does not manufacture one. The recording is instead left unresolved, marked for review, or, where possible, the proven portion is preserved as a .partial.mp4.

The finished MP4 container may contain structures rebuilt by Sony XAVC Recovery, so the recovered MP4 file itself should not be described as a byte-for-byte copy of the deleted MP4 file. The important part is that the recovered media payload is source-backed original data rather than regenerated video.

A file merely playing successfully is therefore not treated as proof. Sony XAVC Recovery writes its physical mapping decisions to the recovery report so you can see where the accepted pieces came from and why they were accepted.

And one other difference is unusually simple: Sony XAVC Recovery is free. No licence, subscription, upgrade screen, or “great news, we found your wedding video, now enter your credit card” moment.

There is no reason you cannot run both programs against the same read-only card image and compare the results. Disk Drill is useful when broad camera-aware recovery is the priority. Sony XAVC Recovery is useful when you want Sony-specific reconstruction where every accepted piece must trace back to the original card data and satisfy a strict uniqueness requirement. See the recovery-method comparison and five proof gates.

Finding Sony video bytes is only part of the problem. A recording can be split across the card, and the pieces still need the correct sizes, order, timing and association. Sony XAVC Recovery uses surviving Sony recording structures to prove how those pieces fit back together. The boxes-and-packing-list explanation shows why.

No. It is designed to reconstruct supported Sony XAVC recordings from a full-sector card image. If you already have an ordinary damaged MP4 file and no card image, this is probably the wrong tool.

It can mean the recording data was overwritten, the controller is returning an empty or unmapped view, the flash was actually erased, or the surviving Sony layout is not supported yet. Those are very different problems. The read-only Hex Viewer and diagnostic bundle can help separate them. See the SD erase explanation.

Yes. The tool is free to use on your own recovery work or client recovery work.

This tool needs a readable full-sector image. If the card is hot, unstable, showing the wrong capacity or producing repeated I/O errors, stop and address the hardware problem first. Software cannot reconstruct sectors the computer cannot read.