Osmo Pocket 3
This reference records a physical Osmo Pocket 3 survey on 11 September 2026, using DJI Mimo 2.11.9 (298019) on iOS. The camera’s About screen reported V01.06.1004. These findings describe that camera and app combination; OpenPocketCine support is documented separately in the iOS and Android pages.
The camera firmware corresponds to DJI’s v01.06.10.04 release notation. Firmware additions can change the menu and protocol behavior, so retain this version when comparing results. DJI release history.
Find a command or capability
Section titled “Find a command or capability”| Reference | Contents |
|---|---|
| Command comparison | Model-specific command families, differences and links to the captured evidence |
| Shooting modes and formats | Video, Low-Light, Slow Motion, native portrait, Photo, Panorama, Timelapse, Motionlapse and Hyperlapse |
| Exposure, focus and audio | Mode-specific controls, focus sequencing, white balance, exposure, Glamour Effects and the full audio DSP blob |
| Zoom and gimbal controls | Resolution-dependent zoom, candidate Med-Tele mapping, follow modes, speed, rotate and recenter |
| Original media | Color and recording commands, measured originals, HTTP transfers, RAW sets and audio companions |
| Mimo album and exports | Device/Local browsing, downloads, favorites, effects processing and inspected editor derivatives |
| Livestream | RTMP setup, BLE configuration and lifecycle, received output and capture completeness |
| USB webcam | Advertised USB formats, delivered host buffers, preserved audio/video and D-Log M follow-up |
| Connection and reconnect | Wi-Fi band writes and readbacks, plus the separate OpenPocketCine recording/warm-reconnect check |
| Coverage and implementation | General menus, remaining qualification, startup investigation and implementation gaps |
Shared framing and reusable commands remain in the command catalog. The pages here preserve the Pocket 3 values, restrictions and evidence needed to apply them.
How to read the evidence
Section titled “How to read the evidence”| Evidence | What it establishes |
|---|---|
| UI | Mimo displayed an option, selection, dialog or visible effect. A selected setting is not an original-file measurement. |
| Accepted | A CRC-valid DUML request matched a successful camera reply. This does not alone prove persistence or the requested physical effect. |
| Status | A separate camera report corroborated the resulting state. Reports can lag a menu change. |
| File | A preserved file was inspected independently for dimensions, codec, timing or sidecars. Full camera/phone SHA-256 matches are called out separately; receiver remuxes are not SD originals. |
| Unverified | A candidate mapping, published specification or untested branch. It is not a replay contract. |
The survey recorded phone IP traffic and Bluetooth controller traffic while Mimo was operated. The first network segment starts in an existing session. Bluetooth controller-log timestamps used the phone’s local wall clock; the writer source and measured phone offset support a two-hour normalization, with raw records preserved separately. This is not a subsecond cross-transport calibration. This is a bounded observation set, not a claim that every packet, setting combination or camera-body feature was captured. Private footage, credentials, identifiers and raw captures are excluded from this reference.