Learn FFmpeg - Basic Commands & Transcoding

#Introduction

After understanding FFmpeg's architecture in episode 2 — the demux-decode-filter-encode-mux pipeline and the stream concept — it's now time to type real commands. This episode is your "first day on the job": you'll learn the basic syntax, do format conversions, and tell apart the two strategies that most determine the result — stream copy and re-encode.

Why does this matter? Because most beginner FFmpeg mistakes don't come from complicated options, but from misunderstanding two things: the input-output order, and whether data is really re-processed or merely copied.

#Basic Syntax

All of FFmpeg starts from one pattern: input first, output last.

Pola dasar

ffmpeg [opsi-input] -i INPUT [opsi-output] OUTPUT

The simplest example — copy (without re-encoding) one file to another format:

Konversi container paling sederhana

ffmpeg -i sample.mp4 sample.mkv

This command reads sample.mp4, then writes sample.mkv. Because there are no codec options, FFmpeg picks defaults: libx264 for video and aac for audio, so re-encoding happens. For quick tasks, this is often undesirable — we cover stream copy below.

Important

The order cannot be reversed: ffmpeg sample.mp4 -i sample.mkv will confuse FFmpeg and fail. Always write -i along with its input before the output. This is rule number one to internalize.

#Format Conversion: MP4, MKV, and WebM

Now let's break down the term we only mentioned briefly in episode 0: codec vs container. Format conversion can mean two very different things:

Changing only the container — the stream content stays, only the "packaging" changes.
Changing the codec at the same time — the data is truly decoded and re-encoded.

For pure container-swapping tasks, use stream copy:

MP4 ke MKV tanpa re-encode

ffmpeg -i sample.mp4 -c copy sample.mkv

MP4 ke WebM dengan re-encode

ffmpeg -i sample.mp4 -c:v libvpx-vp9 -c:a libopus sample.webm

The first completes in seconds because data is only copied. The second takes time because H.264 video must be decoded then re-encoded into VP9 and Opus — the standard codec pair for WebM.

#Stream Copy vs Re-encode

This is one of the decisions you'll make most often. Understand this table early:

Aspect	`-c copy`	Re-encode
Process	Data copied raw	Decode + re-encode
Speed	Very fast	Slow, depending on codec
Quality	Unchanged	Potentially reduced
File size	Generally unchanged	Can be reduced
Use case	Change container, join streams	Change codec, resolution, bitrate

-c copy is like photocopying an original document; re-encoding is like retyping the document on a new computer. Photocopying is instant but can't change content; retyping is slow but you can edit anything. Choose based on need, not habit.

Extending the options: -c:v selects the video codec, -c:a selects the audio codec. Example of re-encoding video only, keeping audio copied:

Video re-encode, audio copy

ffmpeg -i sample.mp4 -c:v libx264 -crf 23 -c:a copy sample-h264.mp4

#Reading Media with ffprobe

Before typing any command, a good technician inspects the raw material first. That's the job of ffprobe — the media inspector we met in episode 2. Start with the most concise form:

Lihat format dan durasi

ffprobe -hide_banner sample.mp4

Contoh output

Input #0, mov,mp4,m4a,3gp,3g2,mj2, from 'sample.mp4':
  Metadata:
    major_brand     : isom
  Duration: 00:00:10.00, start: 0.000000, bitrate: 123 kb/s
  Stream #0:0[0x1](und): Video: h264 (High) (avc1 / 0x31637661), yuv420p, 640x360
  Stream #0:1[0x1](und): Audio: aac (LC) (mp4a / 0x6134706d), 44100 Hz, stereo, fltp

From this output you can directly read duration, bitrate, video resolution, and audio codec — exactly the information needed before deciding on a transcode strategy.

For a more structured form that's easy to use in scripts, combine -show_format with -show_streams:

Format terstruktur untuk script

ffprobe -hide_banner -show_format -show_streams sample.mp4

#Controlling FFmpeg Output

These two options you'll use every day to keep the terminal clean and informative:

Option	Function
`-hide_banner`	Remove the version banner and build config
`-loglevel`	Set detail level: quiet, error, warning, info, debug

The example most used in production — showing only errors:

Mode senyap, error saja

ffmpeg -hide_banner -loglevel error -i sample.mp4 -c copy sample.mkv

With -loglevel error, the terminal only shouts when there's a problem — perfect for scripts and CI/CD that must stay clean.

Tip

Always test on a small file or a short clip before processing a batch: ffmpeg -ss 00:00:00 -t 10 -i sample.mp4 -c copy test.mp4 grabs the first 10 seconds in an instant — the best companion for trying new options without waiting long.

#Conclusion

In episode 3, you've written your first real FFmpeg commands: the basic syntax, MP4-MKV-WebM container conversion, the difference between -c copy and re-encoding, reading metadata via ffprobe, and controlling output with -hide_banner and -loglevel.

Key takeaways:

The basic pattern: ffmpeg [opsi-input] -i INPUT [opsi-output] OUTPUT.
Pure container conversion: -c copy; change codec: re-encode with -c:v and -c:a.
-c copy is a photocopy, re-encode is retyping — choose according to need.
ffprobe is a good habit before every transcode: read duration, bitrate, and streams first.
-hide_banner and -loglevel error keep output clean.

In the next episode 4, we enter the heart of video quality: video encoding & codecs — understanding H.264, H.265/HEVC, VP9, and AV1, the role of -crf and -preset, two-pass strategy, and setting resolution, fps, and GOP. See you in episode 4!