You have probably seen a quality slider labeled something like 0 to 100 every time you save or compress a JPG, but few people know what those numbers actually change. They do not measure how pretty the image looks in any absolute sense. Instead, they tell the encoder how aggressively to throw away visual detail in exchange for a smaller file. Understanding that trade-off is the single most useful skill for anyone who works with photos online, and it is exactly what our convert to JPG tool relies on under the hood.

In this guide we will unpack what JPG compression really does, why a quality of 80 is usually indistinguishable from 100, and how to choose settings for web pages, email, and archiving. By the end you will be able to look at a slider and predict what will happen to both your file size and your image. For the friendly free version of all of this, jpgoo handles the math for you.

What JPG Compression Actually Removes

JPG is a lossy format, which means it permanently discards some of the original data to make files smaller. The clever part is that it discards the information your eyes are least likely to miss. The encoder breaks the image into small 8x8 pixel blocks, converts each block into frequency information using a math operation called the discrete cosine transform, and then rounds off the high-frequency details that human vision is poor at noticing.

Color is treated even more loosely than brightness. Our eyes are far more sensitive to changes in light and dark than to changes in hue, so JPG often stores color at half resolution through a process called chroma subsampling. This is why a JPG of a detailed landscape can look crisp while a JPG of bright red text on blue can look smeared at the edges.

What the Quality Slider Really Controls

The quality number controls a quantization table, which is essentially a set of rounding rules. A high quality value uses gentle rounding and keeps most of the frequency detail. A low quality value rounds aggressively, zeroing out large amounts of data and producing a much smaller file.

  • Quality 90 to 100: Very little is discarded. Files are large and the difference from the original is invisible to almost everyone.
  • Quality 75 to 90: The sweet spot for photographs. Files shrink dramatically while staying visually identical for most viewers.
  • Quality 50 to 75: Noticeable softening begins, useful when small size matters more than perfection.
  • Below 50: Visible blocky artifacts and color banding appear, best reserved for thumbnails.

Crucially, the scale is not linear. Dropping from 100 to 90 might cut a file in half with no visible loss, while dropping from 30 to 20 might barely shrink it at all but ruin the image. Most of the savings live in that first big step.

Why 80 Usually Beats 100

A quality of 100 does not mean perfect; it means the encoder applies almost no rounding, so files balloon in size for detail no human can perceive. For nearly all web and email use, a quality around 80 delivers a file that is two to four times smaller than the 100 version while looking the same on a normal screen. That is why our JPG compress tool targets this range by default.

The exception is when an image will be edited and re-saved many times. Every save discards a little more data, a problem called generation loss. If you know a file will pass through several editing rounds, keep quality high until the very last export.

Spotting Compression Artifacts

Once you know what to look for, low-quality JPGs become easy to identify. The telltale signs are:

  • Blocking: faint 8x8 squares visible in smooth areas like skies.
  • Ringing: ghostly halos or fuzz around sharp edges and text.
  • Banding: visible steps in gradients where there should be smooth transitions.
  • Color bleed: red and blue edges that look smeared due to chroma subsampling.

Images with lots of flat color, sharp text, or hard edges suffer most. Photographs with natural texture hide artifacts well, which is precisely why JPG was designed for photos in the first place. If your image is mostly text or line art, a lossless format may serve you better, a topic we cover in our comparison of JPG vs PNG vs WebP.

Quality Versus File Size: A Practical Comparison

Imagine a 4000 by 3000 pixel photo straight from a phone. Here is roughly how the same image behaves at different settings:

  • Quality 100: around 6 to 9 MB, indistinguishable from the source but wasteful.
  • Quality 85: around 1.5 to 2.5 MB, the version most people would choose for sharing.
  • Quality 70: around 800 KB to 1.2 MB, fine for web pages where speed matters.
  • Quality 40: around 300 to 500 KB, visibly degraded but acceptable for tiny previews.

The lesson is that the move from 100 to 85 buys you most of the savings for almost no visible cost, while pushing below 70 trades real quality for diminishing returns. When file size is your hard constraint, such as an email attachment limit, our walkthrough on how to compress JPG for email shows how to hit a target size precisely.

Resolution and Quality Work Together

Compression is only half the story. A 12-megapixel photo squeezed to quality 60 may still be larger than the same photo resized to 1200 pixels wide at quality 85, and the resized version will often look better on screen. If your image is far bigger than the space it will be displayed in, resizing first is the most effective way to cut size without ugly artifacts. Our JPG resize tool pairs naturally with compression for this reason, and our guide to resizing JPG dimensions explains how to pick the right pixel width.

A Simple Decision Process

When you are unsure what quality to choose, follow this short routine:

  1. Decide where the image will be seen and resize it to roughly that display size first.
  2. Start at quality 85 and view the result at full size.
  3. If the file is still too big, step down to 75, then 70, checking each time.
  4. Stop the moment you can see artifacts, then back up one step.

This approach lets the image itself tell you when you have gone too far, instead of guessing at a magic number.

Troubleshooting Common Quality Problems

Even with sensible settings, JPGs sometimes disappoint. Here are the issues people hit most often and how to fix each one:

  • The file is still huge at quality 80. The image is almost certainly too large in pixel dimensions. A 24-megapixel photo will be big at any quality. Resize it to the width you actually need before compressing, and the size will plummet.
  • Skies and gradients show ugly bands. This is quantization banding. Raise the quality a little, or if the source allows, work from a higher-bit-depth original and apply a subtle dither before exporting.
  • Text looks fuzzy. JPG handles sharp edges poorly. If the image is mostly text or a screenshot, switch to PNG or WebP rather than fighting the quality slider.
  • The image got worse after several edits. That is generation loss from repeated lossy saves. Re-export from your original source instead of re-saving the same JPG again and again.
  • Saturated colors look dull. Aggressive chroma subsampling can mute vivid reds and blues. Higher quality settings use lighter subsampling, so nudge the quality up when color fidelity matters.

A Real-World Example

Suppose you are listing a product on a marketplace that caps images at 2 MB and displays them about 1000 pixels wide. Your camera produced a 5500-pixel, 8 MB JPG. The naive fix is to drop quality until the file squeezes under 2 MB, but that forces a very low setting and introduces visible blocking across the smooth product surface.

The better workflow is to resize the photo to roughly 1500 pixels wide first, then apply quality 85. The result lands comfortably under 1 MB and looks crisp at the size shoppers actually see, with plenty of headroom under the limit. This is the recurring lesson of working with JPG: shape the pixel dimensions to the destination, then let a moderate quality setting do the rest. Trying to solve everything with the quality slider alone almost always produces a worse image than combining a sensible resize with a sensible quality.

How Quality Settings Differ Across Tools

One subtle trap is that a quality value of 80 is not a universal standard. Different encoders use different quantization tables, so quality 80 in one program may produce a slightly larger or smaller file than quality 80 in another, and the visual result can vary too. The numbers are a guide, not a guarantee. This is why the most reliable approach is to judge the image with your own eyes at full size rather than trusting the number alone. When you compress with a consistent tool, however, the numbers become a dependable shorthand you can reuse across projects, which is part of why sticking to one converter pays off over time.

Bringing It All Together

JPG quality is not a measure of beauty but a dial that controls how much imperceptible detail the encoder is allowed to discard. Most of the file-size savings happen in the first reduction from maximum quality, the scale is non-linear, and the best setting depends on whether your image is a photo, text, or graphic, and on where it will be viewed. Resize before you compress, keep quality high while editing, and only push the slider low when size truly matters.

Ready to put this into practice? Use our convert to JPG tool to turn any image into a clean, well-compressed JPG, and let jpgoo handle the rounding tables so you can focus on the picture.