ilovejpg was built around a simple idea: most image jobs need only a handful of well-chosen controls, and those controls should run in the browser without uploading anything. People who search for "I love JPG editor" usually want to know exactly what the editor can do and how to get a good result on the first attempt. This guide answers both questions, one control at a time.
Each section below explains what a control does technically, when it is the right tool, and which settings tend to work. The final sections bring the controls together into practical recipes for email attachments, web pages, online forms, social posts and printing, and explain how the live size estimate and the target-size option remove most of the guesswork from compression.
Opening an image and what happens behind the scenes
The editor accepts JPG and JPEG files, together with PNG, WebP, GIF and BMP. When you open a file, the browser decodes it into raw pixels and draws it onto an HTML5 canvas. Every subsequent change, from resizing to adding text, is applied to that canvas in your browser's memory. Nothing is sent to a server, which is why the editor works on slow connections and why sensitive documents stay on your device.
Two consequences are worth understanding from the start. The output is always a newly encoded file, not a modified copy of the original bytes. That is what allows format conversion and compression, and it is also why EXIF metadata, including camera details and any GPS coordinates, is stripped automatically. In addition, modern browsers apply the EXIF orientation flag when they decode a photo, so a picture that appears upright in the preview will be saved upright, with the rotation baked into the pixels.
| Input format | What to expect | Practical note |
|---|---|---|
| JPG / JPEG | Opens normally; metadata removed on export | Keep the original as a master copy |
| PNG | Transparency preserved if you export as PNG or WebP | Exporting to JPG fills transparent areas with a solid colour |
| WebP | Opens in browsers that support it, which is nearly all current ones | Useful for converting WebP downloads back to JPG |
| GIF | A single still frame is edited | Animation is not kept |
| BMP | Opens as an uncompressed bitmap | Converting to JPG or PNG shrinks it dramatically |
If an image has transparency that matters, such as a logo, export it as PNG or WebP. JPG has no alpha channel, so transparency cannot survive the conversion.
Resizing by pixels or percentage
Resizing changes the number of pixels in the image. It is the single most effective way to reduce file size, because a JPG's size depends heavily on how many pixels it must describe. Halving the width and height leaves a quarter of the pixels, and the file typically shrinks by a similar proportion at the same quality.
Pixels or percentage
Enter exact pixel dimensions when a destination states them, for example a form that asks for 600 pixels wide or a website column that displays images at 800 pixels. Use percentage when you simply want a smaller version, such as 50 percent of a 4000-pixel phone photo, and the precise number is unimportant.
The aspect-ratio lock
With the lock on, changing the width updates the height automatically so that proportions are preserved. Leave it on in almost every case. Turning it off lets you set width and height independently, which squashes or stretches the picture; that is rarely desirable for photographs. If a destination demands a different shape, such as a square, the honest answer is that the image needs cropping, and the editor does not include a crop tool, so do the crop in your system's Photos or Preview app before opening the result here.
- Downscaling is safe and usually improves apparent sharpness at viewing size.
- Upscaling cannot create detail that was never captured; it makes pixels larger and the image softer.
- Resize before adjusting quality, because size and quality interact when you aim for a file-size limit.
- For screens with high pixel density, an image about twice the displayed width looks crisp.
Rotate and flip
Rotate left and rotate right turn the image by 90 degrees in either direction. Use them for phone photos that a particular application displays sideways, for scanned pages fed into the scanner on their side, and for screenshots taken in landscape mode that need to be portrait. Rotation by multiples of 90 degrees is exact; no pixels are interpolated.
Flip horizontal mirrors the image left to right, which corrects the reversed text many front-facing phone cameras produce in selfies, or reverses the direction a subject faces to suit a page layout. Flip vertical mirrors top to bottom and is mostly useful for correcting scans or for deliberate reflection effects.
Because the output is re-encoded, rotating here is not the lossless rotation some desktop tools perform on JPG data. At sensible quality settings the difference is invisible, but avoid rotating and saving the same file over and over.
Brightness, contrast, saturation and grayscale
These controls adjust the tonal and colour character of the whole image. They are global adjustments, applied evenly to every pixel, so they suit corrections rather than local retouching.
Brightness
Brightness raises or lowers all tones. It helps rescue photographs taken indoors or against a bright window. Small moves are best; pushing brightness far upward turns the lightest areas pure white and loses detail that cannot be recovered from a JPG.
Contrast
Contrast spreads tones apart or draws them together. A modest increase gives flat, hazy images more depth and makes scanned documents easier to read. Too much contrast clips shadows to black and highlights to white, and it exaggerates compression artefacts.
Saturation
Saturation controls colour intensity. A small boost can make food, landscapes and product photos look livelier; reducing it gives a muted, understated look. Skin tones are the first thing to look unnatural when saturation is pushed too far.
Grayscale
The grayscale toggle removes colour entirely. It is ideal for scanned text, signatures and forms that request black-and-white images. Grayscale can also reduce file size slightly, because the encoder no longer needs to describe colour variation.
| Problem | Control to try | Sensible starting point |
|---|---|---|
| Photo looks dark and murky | Brightness, then a touch of contrast | Small increases of around 10 to 15 percent |
| Image looks flat or hazy | Contrast | A gentle increase, checking the shadows |
| Colours look dull | Saturation | A small boost; watch skin tones |
| Scanned page is grey and hard to read | Grayscale with extra contrast | Increase contrast until the paper looks white |
| Colours clash with a page design | Saturation reduced, or grayscale | Lower saturation for a muted style |
Adding a text overlay
The text tool places a single line of your own words over the image. You type the text, choose a font size, pick a colour, optionally make it bold, and then drag it into position on the preview. The text becomes part of the pixels when you export, so it cannot be edited afterwards in the saved file.
- Choose a size relative to the image width; text that looks large at full size can become illegible after the image is resized down.
- Use a colour that contrasts with the area behind it, such as white on a dark sky or black on a pale wall.
- Bold improves legibility for short labels, dates and names, especially on busy backgrounds.
- Keep text away from the edges, where social platforms and messaging apps sometimes crop thumbnails.
- Add text after resizing so that you judge its size at the final dimensions.
Typical uses include labelling evidence photos with a date, adding a name to a document image, marking ownership on a picture you are sharing, or captioning a quick product shot. The editor adds new text; it cannot read or change text that is already part of a photograph, which would require OCR or retouching software.
Output format, quality and the live size estimate
The format selector offers JPG, PNG and WebP. The choice matters more than any single slider, because each format suits a different kind of picture.
| Format | Compression | Best for | Quality slider |
|---|---|---|---|
| JPG | Lossy | Photographs and anything with gradual tonal change | Active |
| PNG | Lossless | Screenshots, diagrams, text, logos, transparency | Not used; PNG is always lossless |
| WebP | Lossy in this editor | Web pages, where it is usually smaller than JPG at similar quality | Active |
The quality slider for JPG and WebP controls how aggressively the encoder discards detail. Higher values keep more detail and produce larger files; lower values produce smaller files with more visible artefacts such as blockiness and halos around edges. The scale is not linear in its effect. Between roughly 75 and 90 most photographs look excellent; above 90 the file grows quickly for little visible gain; below about 60 artefacts start to show in skies and skin.
The live estimate of output file size updates as you change dimensions, format and quality. Treat it as your feedback loop. Instead of exporting, checking the file and trying again, watch the number while you move the slider and stop when it falls within the size you need. Because the estimate reflects the actual encoding settings, it shows clearly how much resizing contributes compared with quality reduction.
If moving quality from 85 to 70 barely changes the estimate, the image is probably simple already. If it still exceeds your limit, reduce the dimensions instead of pushing quality lower.
Target size in KB
Many upload forms state a maximum file size, often 100 KB, 200 KB, 500 KB or 1 MB. The optional target size in KB is designed for exactly this. When you enter a limit, the editor searches for the highest quality setting whose output fits beneath it. You get the best image that satisfies the requirement, without trial and error.
Target size works through the quality setting, so it applies to JPG and WebP. It cannot reduce the number of pixels on its own. If a very large photo cannot fit under a small limit even at low quality, or fits only with obvious artefacts, the correct remedy is to reduce the dimensions first and then let the target size find the best quality for the smaller image.
- Set dimensions to what the destination actually displays or requires.
- Enter the size limit slightly below the stated maximum, for example 95 KB for a 100 KB limit, to allow for how different systems count kilobytes.
- Check the preview for blockiness in smooth areas such as sky or walls.
- If quality looks poor, lower the dimensions and let the target size search again.
Recommended settings for common jobs
The settings below are reasonable starting points rather than rules. Always check the destination's own requirements, which take precedence.
| Job | Dimensions | Format and quality | Other advice |
|---|---|---|---|
| Email attachment | About 1600 to 2000 pixels on the long edge | JPG at 75 to 85 | Several photos together should stay well under typical attachment limits |
| Web page image | About twice the displayed width, such as 1600 pixels for an 800-pixel column | WebP or JPG at 75 to 82 | Smaller files load faster on mobile connections |
| Online form or portal | As stated by the form; otherwise around 1000 to 1200 pixels | JPG with target size set just under the limit | Use grayscale for scanned documents if colour is not required |
| Social post | Around 1080 pixels wide is widely used | JPG at 85 to 90 | Platforms recompress uploads, so starting with good quality limits double loss |
| Print size in inches multiplied by 300, such as 1800 × 1200 pixels for a 6 × 4 inch print | JPG at 90 to 95 | Avoid upscaling; for several pages use the JPG to PDF tool |
Why print needs more pixels
Printing quality depends on how many pixels cover each inch of paper. Around 300 pixels per inch is the usual standard for close viewing; posters viewed from a distance can use fewer. The DPI number stored in a file is only a label, and since the editor strips metadata, set the physical size in your print dialogue instead. What matters is that the pixel dimensions are large enough for the intended print size.
Frequently Asked Questions
Does the ilovejpg editor upload my photos?
No. Every step, from opening the file to encoding the output, runs inside your browser using the HTML5 Canvas and JavaScript. The image stays on your device throughout. This also means the editor keeps working on poor connections once the page has loaded, and it is suitable for personal documents that you would prefer not to send to an online service.
Why is my exported image smaller than the original even at high quality?
Phone cameras often save photos with generous quality settings and large metadata blocks, including thumbnails and location data. When the editor re-encodes the image, metadata is removed and the encoder uses your chosen settings, which together usually produce a smaller file. If you also reduced the dimensions, the saving will be far larger, because file size depends heavily on pixel count.
Can I crop an image in the editor?
No, the editor has no crop tool. It can resize, rotate, flip, adjust tone and colour, add text and convert formats. To crop, use the built-in Photos app on Windows, Preview on a Mac, or the photo editor on your phone, then open the cropped file in the editor to resize, compress or convert it. Changing dimensions with the aspect lock off distorts rather than crops.
What happens if the target size cannot be reached?
Target size works by lowering quality until the file fits, so it cannot remove pixels by itself. If a large image still exceeds a small limit, or only fits with visible blockiness, reduce the width and height first. A smaller image at decent quality nearly always looks better than a large image compressed severely, and the target size option can then find the best setting.
Should I choose JPG, PNG or WebP?
Choose JPG for photographs that must work everywhere, including older systems and upload forms. Choose PNG for screenshots, diagrams, text and anything needing transparency, accepting larger files. Choose WebP for web pages, where it usually produces smaller files than JPG at similar visual quality and is supported by all current major browsers. When unsure about a form's requirements, JPG is the safest choice.
Is location data removed from my photos?
Yes. Re-encoding the image through the canvas writes only pixel data, so EXIF metadata such as camera model, capture time and GPS coordinates is not carried into the exported file. This happens automatically with no setting to remember. If you need to keep metadata for archiving, store the untouched original separately and share only the edited copy.
Conclusion
The editor is intentionally focused. Resizing with the aspect lock, rotate and flip, four tonal and colour controls, a text overlay, three output formats, a quality slider, a live size estimate and a target size in KB cover the large majority of everyday image tasks. Understanding what each control does, and in which order to use them, turns those tools into a reliable workflow: resize first, adjust gently, add text at the final size, then choose format and quality while watching the estimate.
For anyone searching for an I love JPG editor that works without installation or uploads, the practical advice is to start from the destination's requirements and work backwards. Know the dimensions and file-size limit you need, keep the original safe, export once, and let the target-size option find the best quality that fits. For cropping, layers or RAW files, pair the editor with your system's built-in apps or a desktop program, and use each for what it does best.