Reference Glossary
Compression ratio
The relationship between a file's original size and its compressed size, expressed as a ratio (e.g., 10:1) or a percentage reduction.
Why it matters in a DAM
DAM storage and bandwidth budgets are shaped directly by the compression ratios applied across rendition tiers — a highly compressed thumbnail versus a near-lossless master multiplies cost differently when repeated across a library of millions of assets, which is why most DAM transcoding rules define different target ratios per tier rather than one blanket setting. Because achievable ratio depends heavily on image content (a flat-colour graphic compresses far more than a detailed photograph at the same quality setting), the same setting applied library-wide produces very different real-world file sizes.
A worked example
Common mistake
Applying one fixed compression setting to an entire library regardless of content type, which over-compresses detailed photography into visible artifacts while under-compressing simple graphics into needlessly large files.
Compression ratio just compares before and after: a 10 MB file compressed to 1 MB is a 10:1 ratio. What determines the ratio a DAM can safely apply isn’t a single universal number — it’s a function of both the compression method and the image content itself. Busy, high-detail photography resists compression more than flat-colour graphics or illustrations at the same visual-quality target, so two assets processed with identical settings can end up with very different actual ratios.
For that reason, most DAM transcoding pipelines don’t chase a single target ratio — they target a quality level per rendition tier (thumbnail, web preview, print master) and let the resulting ratio vary by asset. Trying to enforce a specific ratio uniformly across a diverse library tends to either degrade detailed photography visibly or leave simple graphics far larger than necessary.
Frequently asked
What is compression ratio?
The relationship between a file's original size and its compressed size, expressed as a ratio (e.g., 10:1) or a percentage reduction — a 10 MB file compressed to 1 MB is a 10:1 ratio.
What determines the compression ratio a DAM can safely apply?
Two factors combine: the image content itself, and how the resulting file will be used. Busy, high-detail photography resists compression more than flat-colour graphics at an identical quality setting. But the deciding factor is the derivative's purpose — a thumbnail meant for fast browsing tolerates aggressive compression because detail loss goes unnoticed at that size, while a print master needs minimal or zero compression because every detail matters. The safe ratio is set by usage context, not one universal number.
Why don't DAM transcoding pipelines chase a single target ratio?
Because two assets processed with identical settings can end up with very different actual ratios; pipelines instead target a quality level per rendition tier and let the resulting ratio vary by asset.
What compression ratio is typical for a thumbnail versus a print master?
A thumbnail typically uses a high ratio, around 20:1 or more, prioritizing speed and storage, while a print master uses a low ratio or stays lossless, prioritizing fidelity above file size.
What's the mistake of applying one fixed compression setting library-wide?
A single fixed setting ignores that different content types and different use cases need different compression levels. Thumbnails and print masters serve opposite goals, and a simple logo behaves nothing like a detailed photograph at the same ratio — flat graphics tolerate heavy compression while photography shows artifacts fast. Applying one setting library-wide either over-compresses assets that need quality preserved, or under-compresses assets that should stay lightweight, wasting storage and bandwidth.
How do compression ratios affect DAM storage budgets at scale?
A highly compressed thumbnail versus a near-lossless master multiplies cost very differently when repeated across a library of millions of assets, which is why different rendition tiers get different target ratios.