Transcoding Video

What is Transcoding?

Transcoding is the process of decoding a video from its current format and re-encoding it into a different format. You might transcode to shrink a file for the web, make a video compatible with a specific device, or switch to a more efficient codec to save storage costs.

This is different from remuxing (which we covered in Lesson 1 using -c copy). Remuxing just changes the container without touching the encoded data. Transcoding re-encodes the data itself.

Containers vs. Codecs

Before diving into transcoding, it helps to understand the difference between containers and codecs — the two concepts people confuse most often:

• Container (MP4, MKV, WebM, MOV) — the "box" that holds video and audio streams, subtitles, and metadata
• Codec (H.264, H.265, VP9, AV1) — the algorithm used to compress and decompress the actual video data

An MP4 file could contain H.264 video, H.265 video, or even AV1 video. The container and codec are independent choices. When you transcode, you are changing the codec. When you remux, you are changing the container.

Choosing a Video Codec

Controlling Quality with CRF

CRF (Constant Rate Factor) is the most common way to control quality in FFmpeg. It tells the encoder: "I care about quality, not file size — pick whatever bitrate is needed to maintain this quality level."

Lower CRF = higher quality = larger file. The scale depends on the codec:

Target Bitrate Encoding

When you need predictable file sizes — for example, streaming at a fixed bandwidth — use target bitrate instead of CRF:

Common bitrate targets for H.264:

• 720p: 2-4 Mbps
• 1080p: 4-8 Mbps
• 4K: 15-30 Mbps

Encoding Speed with Presets

Presets control the trade-off between encoding speed and compression efficiency. A slower preset produces a smaller file at the same quality, but takes longer to encode.

The available presets, from fastest to smallest output:

medium is the default and a solid choice for most use cases. Use fast or veryfast when encoding time matters more than file size (e.g. live processing). Use slow for final renders where you want every byte optimized.

Two-Pass Encoding

When you need both a target bitrate and the best possible quality, use two-pass encoding. The first pass analyzes the video complexity. The second pass uses that analysis to distribute bits optimally.

Two-pass is especially useful for long videos with varying complexity — like a tutorial that alternates between a talking head (simple) and a screen recording (complex). The encoder allocates more bits to the complex scenes.

Practical Recipes

Web-optimized MP4 (fast start)

The -movflags +faststart flag moves the metadata to the beginning of the file, so browsers can start playing before the full download completes. Always use this for web video.

Batch convert a folder

Convert to H.265 for storage savings