Every JPEG export tool asks for a quality setting, usually a slider or a percentage, and rarely explains what the number actually controls. Here's JPEG quality settings explained without the jargon.
It's not "80% of the original quality"
The percentage doesn't mean the image retains 80 percent of its original visual information — that's a common assumption, and it's not quite right. It's a setting that controls how aggressively the JPEG format simplifies the image data during compression. Higher numbers preserve more fine detail (and produce larger files); lower numbers discard more detail in exchange for a smaller file. There's no universal, precise mapping from percentage to "how good it looks" — that depends heavily on the specific image, since a simple, low-detail photo compresses more cleanly than a busy, high-detail one at the same setting.
Where quality loss actually shows up
JPEG compression works by simplifying detail in ways the human eye is less likely to notice — small color variations, fine texture. Pushed far enough, that simplification becomes visible as blockiness around sharp edges, smearing in fine textures (hair, fabric, foliage), and banding in smooth gradients like skies. That's the visible signature of quality being set too low for that particular image.
What setting to actually use
There's no single "correct" number — it depends on what you're optimizing for:
- 90 to 100 percent: Effectively visually lossless for most photos, but produces the largest files. Reasonable for a master copy you might edit again later.
- 75 to 90 percent: The sweet spot for most everyday sharing — noticeably smaller than the top range, with quality loss that's difficult to spot without zooming in.
- 50 to 75 percent: Meaningfully smaller files, with mild softness that becomes visible on close inspection, especially in detailed areas.
- Below 50 percent: File size drops fast, but so does visible quality — usually only worth it when you're under real pressure to hit a strict size limit and quality is a secondary concern.
These ranges are general guidance based on typical encoder behavior, not a precise measurement — the right number for a specific photo still depends on its content.
Why targeting a file size beats guessing a percentage
Picking a quality percentage requires knowing in advance how that number will behave on your specific photo — which is hard to predict without testing it. Targeting a file size instead flips the problem: you say how small you need the file to be, and the compression works backward from there, only reducing quality as far as it actually needs to.
- Drop your photo into the compressor and set a target size instead of a quality percentage.
- It searches for the highest quality setting that still hits your target — rather than you guessing a percentage and hoping the result is small enough.
- Preview before downloading, so you see exactly how much (if any) quality was traded away to hit the number you set.
The short version
"80 percent quality" isn't a precise, universal measurement — it's a dial that behaves differently depending on the photo. If what you actually care about is hitting a specific file size, working backward from that number is more reliable than guessing at a percentage.