Lossless jpeg compressor

What a lossless JPEG compressor does

A lossless JPEG compressor reduces file size while keeping the picture’s pixel values the same. That is a narrower promise than “the photo still looks good.” Everyday JPEG is a lossy format: when a camera or editor first saves the image, it runs a discrete cosine transform (DCT) and then quantization, which discards detail to save space. A later tool can still be lossless if it avoids throwing any more of that data away.

The phrase is ambiguous in a way that changes what you should use. It can mean rewriting an existing JPEG so it decodes to identical pixels, encoding uncompressed pixels with a true lossless JPEG mode, or shrinking a photo until it only looks the same on screen. Only the first two keep the samples unchanged. This guide does not test or rank named apps, sites, or prices; treat any specific product claim as unverified here.

The three meanings people usually want

• Lossless optimization of a JPEG you already have: the decoded pixels stay identical; the tool only packs the same coefficients more tightly and may drop extra tags.
• True lossless encoding: every pixel is stored exactly, typically from an uncompressed capture or another lossless original, using JPEG’s lossless mode or a related lossless still-image format.
• High-quality lossy compression: a smaller file that still looks acceptable. Data is discarded; the result can look sharp in a preview and still differ from the original samples.

If your goal is identical pixels, you need optimization or a true lossless encode. If your goal is a smaller attachment that still looks fine on a call, a form, or a phone screen, you are choosing high-quality lossy JPEG and should judge it by inspection, not by the word lossless.

How an existing JPEG can get smaller without changing the picture

Once a photo is already a JPEG, a lossless compressor works on the compressed coefficients and on extra bytes wrapped around them. Rebuilding Huffman tables (the entropy coding that stores those coefficients) can cut size while leaving the decoded image unchanged. Stripping metadata you do not need—duplicate thumbnails, editing history, or location tags—does the same. JPEG also allows some lossless operations in place, such as certain 90-degree rotations and flips, so the file is rewritten without a full decode-and-resave.

How much you save depends on how the file was encoded and how much extra data it carries. There is no single typical percentage in the evidence for this article. A file that is already tightly encoded and stripped may barely shrink. Detail the camera discarded on the first save cannot be restored; a lossless tool only prevents further loss.

True lossless encoding

A true lossless JPEG mode skips the usual quantize-and-discard step and stores pixels exactly, using prediction rather than the everyday DCT pipeline. That is useful when you start from RAW, TIFF, PNG, an uncompressed webcam still, or another lossless master and the recipient specifically needs JPEG. The lossless file is often larger than a normal camera JPEG of the same scene, because those camera files already threw data away. If the recipient can take another lossless format, that can be a better fit than forcing JPEG.

Actions that still change quality

• Opening a JPEG and exporting it again, even at the maximum quality setting, re-encodes the image and can change pixels.
• A quality or compression slider on an editor or upload page is a lossy control, including when the preview still looks clean.
• Resizing, cropping that decodes the image first, sharpening, filters, and background replacement all change samples before any compressor runs.
• Round-tripping through another format and back to ordinary JPEG reintroduces lossy encoding on the last save unless that last step is a documented lossless encode.

Online compressors that claim no quality loss

An online compressor is lossless only if it preserves decoded pixels or uses a true lossless encode, and if the file you download matches that claim. Copy that says “without losing quality” often means the result still looks acceptable. That is a visual judgment, and this article has no test results for any particular online tool.

Uploading a camera photo, a call snapshot, or a document sends that file to a remote service. Use a local optimizer when the picture is private. To verify lossless output, decode the original and the result and compare the raw samples (a difference image or a checksum of the pixels). A smaller file size or a sharp preview does not prove the pixels are unchanged.

What to do next

Decide whether you need identical pixels or a smaller file that still looks good. For identical pixels on an existing JPEG, use lossless optimization and optional metadata stripping, then compare samples if identity matters. For identical pixels from a lossless original, use a lossless encoder. For a lean file for a message, a form, or a video-call still, use high-quality lossy JPEG and inspect the result. Compression does not fix focus, lighting, or a messy physical background; it only changes how the capture is stored.