# facebook/transform360

Transform360 is an equirectangular to cubemap transform for 360 video.

Repository: https://github.com/facebook/transform360
Canonical: https://ross.abutalabs.com/products/transform360
Language: C++
License: NOASSERTION
License Family: other
Last push: 2026-05-08T22:16:23+00:00

## Health v2 (maintenance only)
Score: 69/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 81, release rhythm 35, longevity 100
- inputs: {"age_days": 3893, "days_push": 117, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1034, forks 241 (observed 2026-08-28T04:03:18.581422+00:00)

## What it is
Transform360 is a C++ video/image filter that converts 360-degree video between projections, typically from equirectangular to cubemap format. It is designed to be integrated as an FFmpeg video filter and offers improved memory usage, visual quality, and per-region quality control.

## Use cases
- convert equirectangular 360 video to cubemap projection
- reproject 360 footage for VR playback
- process 360 video with ffmpeg filters
- optimize memory usage when transcoding 360 video
- control quality of specific regions in a 360 video frame

## When to choose
- you need equirectangular-to-cubemap conversion in an ffmpeg pipeline
- you want fine-grained control over output quality regions of 360 video
- you need better memory/performance than naive projection transforms

## When to avoid
- you need a maintained turnkey tool with simple installation
- you are not comfortable rebuilding ffmpeg from source
- you work with non-360 video or non-projection transforms

## Facets
- artifact type: library
- maturity: maintenance
- function: video-processing, image-processing
- domain: media, graphics
- platform: cpp, cross-platform
- tags: 360-video, equirectangular, cubemap, ffmpeg-filter, vr-video, video, linux

## Member repositories
- facebook/transform360 (main) score 69

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:18.581422+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T07:06:33.001697+00:00, confidence not recorded.
  - readme: https://github.com/facebook/transform360 (fetched 2026-08-28T04:03:18.581422+00:00, sha 819e0d2197a2)
- Data as of 2026-08-30T08:39:29.467469+00:00.