Choose a codec and encoder

Start with the playback, editing, delivery, and licensing constraints. A codec with better compression is not useful when the target device cannot decode it or the required encoder is unavailable.

Decision order

1. List every required player, device, editor, and delivery platform.
2. Identify the codecs and profiles all of them accept.
3. Decide whether the job needs lossless, visually lossless, or lossy output.
4. Decide whether size, encode time, decode cost, or broad compatibility has priority.
5. Choose an encoder implementation and test a representative short segment.
6. Choose a container that accepts every video, audio, and subtitle stream you need.

Typical goal	Starting point	Check before committing
Broad consumer playback	AVC/H.264 with AAC in MP4	Profile, level, pixel format, and device limits
Efficient modern delivery	AV1, VP9, or HEVC	Hardware decode coverage and platform ingest rules
Editing/intermediate	ProRes, DNxHR, or lossless formats	Editor support, storage, chroma, and bit depth
Archival preservation	Lossless video and audio in a documented container	Validation, checksums, metadata, and recovery tooling

List the encoders in your FFmpeg build instead of assuming a package includes one:

ffmpeg -hide_banner -encoders
ffmpeg -hide_banner -h encoder=libsvtav1

Validation

Encode a short segment containing motion, texture, gradients, dark regions, and audio transitions. Inspect it on a real target device, then use ffprobe to confirm the output streams and metadata:

ffprobe -v error -show_format -show_streams -of json output.mkv

Do not compare CRF or preset numbers across different encoders as if they represented the same quality or speed. Compare the resulting media under the same source, scaling, pixel-format, and playback conditions.

Primary references

• FFmpeg codec documentation
• FFmpeg stream selection
• ffprobe documentation