Does Saving a JPEG at Higher Quality Fix Artifacts?

Recompressing an existing JPEG image at a higher quality setting will not remove, reduce, or repair existing compression artifacts. Because JPEG is a lossy format, original image data discarded during previous saves cannot be recovered. Instead, saving a degraded JPEG at a higher quality setting permanently preserves those existing flaws, potentially introduces subtle new artifacts through generational loss, and unnecessarily inflates the file size.

The Permanence of Discarded Data

When an image is first compressed as a JPEG, a mathematical process called the Discrete Cosine Transform (DCT) converts the pixel data into frequency components. The algorithm then applies quantization, which permanently discards high-frequency data (fine textures, sharp transitions, and subtle color gradations) that the human eye is less likely to notice.

Once this data is eliminated, it is gone forever. Opening the image in an editor does not restore this lost information; it merely reconstructs an approximation based on the remaining data.

How the Encoder Treats Existing Artifacts

A JPEG encoder does not possess contextual awareness. It cannot distinguish between genuine visual details (like the texture of fabric or hair) and compression defects (such as 8x8 pixel blockiness, mosquito noise around sharp edges, or color banding).

When you re-export an artifact-heavy JPEG at a high quality setting (such as 90 to 100), the encoder treats every visible artifact as an intentional, high-priority detail. It dedicates substantial data to accurately preserving the jagged edges, blurry blocks, and noise created by the earlier compression passes.

Increased File Size Without Quality Improvement

Choosing a higher quality slider value lowers the quantization compression rate, instructing the encoder to preserve as much visual data as possible. Because the encoder is now meticulously preserving both the underlying picture and the existing artifacts, the resulting file size increases significantly. The user ends up with a much heavier file that looks visually identical to—or slightly worse than—the lower-quality source.

The Impact of Generational Loss

Recompressing a JPEG rarely yields a completely neutral result, even at the highest quality settings. Each save cycle re-evaluates the image through color space conversion, chroma subsampling, and DCT calculations:

• Rounding Errors: Floating-point math during the transform stages introduces small rounding errors each time an image is decoded and re-encoded.
• Grid Misalignment: If the image was cropped or resized by an amount not divisible by 8 or 16 pixels, the 8x8 block grid shifts. The encoder will then generate a new layer of boundary artifacts across areas that were previously smooth.
• Color Subsampling: If the re-save uses 4:2:0 chroma subsampling, color information is degraded yet again, softening saturated edges further.

Saving a JPEG at a higher quality setting acts as a high-fidelity snapshot of an already degraded image. To avoid baking in artifacts and accumulating generational loss, original images should always be stored in a lossless format like PNG, TIFF, or RAW, and only exported to JPEG as the final step in the workflow.