Why Re-saving Low Quality JPEG at 100 Increases Size

Re-saving a low-quality JPEG file at maximum quality (100) significantly increases its file size without restoring any lost visual detail. This counterintuitive result happens because JPEG is a lossy compression format; once original pixel information is discarded during the initial compression, it is gone permanently. When re-saved at maximum quality, the encoder treats existing visual degradation—such as compression artifacts and blockiness—as complex, intentional details that require massive amounts of data to preserve.

The Irreversible Nature of Lossy Compression

JPEG compression relies on mathematical algorithms, primarily the Discrete Cosine Transform (DCT), to reduce file size. During this process, the image is broken into 8x8-pixel blocks, and fine details—especially subtle color transitions and high-frequency textures that human eyes struggle to see—are rounded off and permanently deleted through a process called quantization.

Once this data is removed, no software can guess or recreate the original information. Setting the quality slider to 100 does not reconstruct lost pixels; it simply dictates how aggressively the encoder should compress the current state of the image.

Artifacts Are Encoded as Complex Detail

When an image is heavily compressed, it develops visible flaws known as artifacts. These include:

• Blocking: Distinct, visible boundaries around 8x8 pixel grids.
• Banding: Harsh, step-like transitions where smooth gradients used to be.
• Mosquito noise: High-contrast ringing or haziness around sharp edges.

To a JPEG encoder, a smooth gradient is easy to compress because it contains low-frequency data that can be represented with very little code. However, the sharp edges of compression blocks and ringing artifacts represent high-frequency data. When you instruct the encoder to save at quality 100, it applies minimal quantization, forcing the file to meticulously preserve these harsh, artificial edges. The file size expands because the encoder is expending large amounts of storage to maintain the visual flaws created by the first compression pass.

Generation Loss

Far from improving the image, re-saving a JPEG at any quality setting actually introduces a phenomenon known as generation loss. Each time a lossy file is decompressed into raw pixels and re-encoded back into a JPEG, new rounding errors occur during the DCT calculation.

Saving a low-quality JPEG at quality 100 results in the worst of both worlds: the image accumulates subtle additional distortion from being re-encoded, while the file size balloons to store the artifacts of previous compression cycles.