Raw, uncompressed video is enormous — far too large to stream practically. A single minute of uncompressed 1080p video can run into gigabytes. Video encoding, the broader process that includes compression as one of its central steps, is what makes streaming possible at all — turning that impractically large raw data into something your internet connection can actually deliver in real time.
The basic idea behind compression
Video compression removes redundant information. Two consecutive frames in a video are often very similar, so instead of storing each frame in full, compression stores only what changed between them. This is called temporal compression, and it's one of the biggest reasons video compresses so much more effectively than a single still image — most of a video's frames share the vast majority of their content with the frames immediately before and after them.
The full encoding pipeline, step by step
- Capture. Raw video is captured from a camera or source feed, typically at a very high, uncompressed data rate.
- Preprocessing. The raw signal may be scaled, color-corrected, or otherwise prepared before compression begins.
- Encoding. A codec compresses the video using both temporal (frame-to-frame) and spatial (within-frame) redundancy reduction techniques.
- Packaging. The compressed video is packaged into segments and a manifest, as covered in our HLS deep dive.
- Delivery. The packaged content is distributed via a CDN to viewers' devices.
- Decoding. Your device's decoder reverses the compression, reconstructing viewable video frames in real time.
Lossy vs. lossless compression
Streaming video uses lossy compression, discarding some detail the human eye is less likely to notice, in exchange for dramatically smaller file sizes. This is different from lossless compression (like a ZIP file), which preserves every detail exactly but achieves far smaller size reductions as a result. The entire streaming industry is built on the premise that a carefully, intelligently lossy approach delivers a far better practical outcome — smooth, accessible streaming — than insisting on lossless perfection that would make real-time delivery over typical home connections essentially impossible.
Spatial vs. temporal compression
| Type | What it exploits | Analogy |
|---|---|---|
| Spatial compression | Redundancy within a single frame (similar colors, patterns) | Similar to how a JPEG image compresses |
| Temporal compression | Redundancy between consecutive frames over time | Only storing what changed since the last frame |
Modern codecs use both techniques together, which is part of why video compresses so much more efficiently than a sequence of individually-compressed still images would — temporal compression alone can eliminate the vast majority of a video's raw data before spatial compression even begins its work on what remains.
Codecs do the heavy lifting
The specific compression method is handled by a codec — see our guides to H.264 vs H.265 and HEVC vs AV1 for the most common ones used in streaming today. The codec defines the specific algorithms and rules for both spatial and temporal compression; the encoder is the actual software or hardware implementation that applies those rules to a specific piece of content.
Tip
If you're ever comparing streaming quality between providers, remember that codec choice, encoder quality, and chosen bitrate all affect the final result independently — two providers using the identical codec can still produce noticeably different quality.
Pros and cons of aggressive compression
| Pros | Cons |
|---|---|
| Dramatically reduces bandwidth needs, enabling streaming at all | Very aggressive compression can introduce visible artifacts |
| Allows higher resolutions to fit within realistic bandwidth budgets | Requires more processing power to encode and decode |
| Modern codecs minimize visible quality loss at reasonable settings | Fast motion and complex scenes compress less efficiently |
Best practices for understanding compression's practical impact
- Understand that quality and file size are always a trade-off, not independent variables a provider can maximize simultaneously.
- Recognize that codec efficiency affects your data usage — a more efficient codec delivers comparable quality using less bandwidth.
- Don't judge compression quality from a single frozen frame — motion handling matters as much as static image quality.
- Check your device's hardware decoding support for whichever codec your content uses, covered in our codec comparison guides.
Common mistakes about video compression
- Assuming all compression is equally aggressive, when bitrate and codec choice both significantly affect the final quality-to-size trade-off.
- Confusing compression with resolution — a high-resolution video can still be heavily compressed and look worse than a lower-resolution, less compressed alternative.
- Blaming "bad compression" for issues actually caused by network delivery, like buffering, which is unrelated to how the source content was encoded.
- Assuming lossless compression would be better for streaming, when it would make real-time delivery over typical connections completely impractical.
Troubleshooting compression-related quality issues
Video looks blocky or blurry, especially during motion
This typically indicates the bitrate is too low for the content's complexity — fast motion is harder to compress efficiently, revealing artifacts more visibly than in static scenes.
Quality varies noticeably between different channels
Different channels or providers may use different encoding settings, codecs, or bitrates — some genuine quality variation between sources is normal and expected.
4K content doesn't look dramatically better than HD
If bitrate wasn't increased proportionally with resolution, the extra pixels may be poorly compressed — check whether your provider allocates sufficient bitrate specifically for their 4K tier.
Common mistake
Assuming a higher resolution number automatically means better quality. A poorly-compressed 4K stream can look worse than a well-compressed 1080p one — resolution and compression quality are related but genuinely separate factors.
How keyframes and inter-frames work together
Diving one level deeper into temporal compression: video codecs organize frames into a repeating pattern of keyframes (also called I-frames, which store a complete image independently) and inter-frames (which store only the differences from nearby frames). A keyframe acts as a reference point the decoder can start from at any time — necessary for seeking, for a player app joining a live stream mid-broadcast, or for recovering after a dropped connection. Inter-frames between keyframes are much smaller, since they only encode what changed, but they depend on that preceding keyframe to be decoded correctly. This is part of why seeking to a very precise point in a video sometimes lands slightly off — the player often jumps to the nearest keyframe rather than decoding every inter-frame in between to reach an exact frame.
How encoder settings affect real-world results beyond just bitrate
Bitrate gets the most attention, but encoder configuration involves many other settings that affect final quality at a given bitrate. Encoding speed presets trade processing time for compression efficiency — a "slower" preset spends more computational effort finding the most efficient way to represent each frame, typically producing better quality at the same bitrate than a "faster" preset, at the cost of taking longer to encode. Keyframe interval settings affect the balance between seeking precision and compression efficiency, since more frequent keyframes improve seek accuracy but increase file size. For live streaming specifically, encoding must happen in real time, which constrains how much computational effort can be spent on the "slower, better" end of these trade-offs compared to pre-encoded on-demand content that can be processed for as long as needed before publication.
A real-world example: comparing two providers' 4K streams
Consider two IPTV providers both advertising 4K channels, yet one consistently looks noticeably sharper and cleaner than the other during fast-moving sports content. Understanding compression explains why this happens despite both claiming the same resolution:
- Provider A allocates a generous bitrate specifically tuned for fast-motion sports content, using a slower, more efficient encoding preset given their infrastructure investment.
- Provider B uses a lower bitrate and a faster encoding preset to reduce their own infrastructure costs, prioritizing broad channel availability over per-channel encoding quality.
- Both streams are technically "4K," but Provider A's stream contains meaningfully more actual visual information per frame, especially during the fast motion that sports content demands.
- A viewer comparing the two side by side would likely notice Provider A's stream holds up better during fast camera pans and rapid player movement, exactly where compression is put under the most strain.
This example illustrates why resolution alone is an incomplete way to evaluate streaming quality — the encoding choices behind that resolution number often matter more to your actual viewing experience than the number itself.
Semantic terms worth knowing
A few related terms are useful here: keyframe (I-frame) (a complete, independently decodable frame used as a reference point), inter-frame (a frame that stores only differences from nearby frames), encoding preset (a setting balancing encoding speed against compression efficiency), and rate control (the method an encoder uses to manage bitrate across a piece of content, whether constant or variable).
Key takeaways
- Video encoding is a multi-step pipeline; compression, using both spatial and temporal techniques, is its most impactful stage.
- Streaming relies on lossy compression, trading some imperceptible detail for dramatically smaller file sizes.
- Codec choice, encoder quality, and bitrate together determine final streaming quality — no single factor alone.
- Fast motion and complex scenes are inherently harder to compress well than static content.
- Resolution and compression quality are related but separate — a higher resolution doesn't guarantee better final quality.
Conclusion
Video compression is the quiet engineering achievement that makes modern streaming possible at all — without it, delivering video over typical home internet connections in real time simply wouldn't work. Understanding the basic trade-offs involved, and where compression fits within the broader encoding pipeline, makes it much easier to understand why streaming quality varies the way it does, and what's actually within your control versus what's determined upstream by your provider's encoding choices.