How to compress images for websites, email and messaging
Reduce image size without guessing by choosing the right dimensions, format and quality for the destination.
Start with dimensions, not only quality
A 4000-pixel camera photo is usually unnecessary in an email or a 900-pixel website column. Reducing dimensions removes far more data than repeatedly lowering quality, and it often produces a cleaner result.
Choose dimensions based on the largest place the image will actually appear. Keep the original separately if you may need a print or a larger crop later.
Choose a format that matches the content
JPG is dependable for photographs. WebP often makes modern website images smaller at similar visual quality. PNG is valuable for transparency, screenshots and graphics with sharp edges, but photographic PNG files can be very large.
Conversion does not restore missing detail. Moving a heavily compressed JPG into PNG prevents another lossy save, but it does not repair artifacts already present in the source.
Use quality as a finishing control
Begin with a moderately high setting and inspect faces, text, gradients and detailed edges. Lower the setting gradually until the size is practical. The number on a quality slider is not comparable between every format or encoder, so judge the downloaded result.
For a batch, test one representative image before processing everything. A setting that works for a simple portrait may be too aggressive for a detailed product photo or screenshot.
- Zoom to 100 percent for inspection
- Check text and logos for halos
- Compare the actual output size
- Keep the original source
Match the delivery channel
Email and messaging prioritize compact files and broad compatibility, so JPG is often suitable. Websites benefit from WebP and correctly sized responsive images. Graphics that require transparent edges should remain PNG or WebP.
Do not enlarge a small source merely to reach a preset dimension. Upscaling increases pixel count but cannot create genuine detail.
Protect privacy in photographs
Camera images can contain location, device and capture metadata. Remove metadata before public posting when those details are unnecessary, then verify the cleaned file rather than deleting the original immediately.
MyFileNest performs supported compression in the browser, but you should still review the final image for visible private information before sharing it.
Target sizes by destination
These are practical starting points rather than rules. Always check the specific site's or platform's own limit.
| Destination | Width | Format | Quality |
|---|---|---|---|
| Email attachment to share photos | 1,600 px | JPG | 65–75% |
| Full-width website hero | 1,920 px | WebP | 70–80% |
| Blog body image | 1,200 px | WebP or JPG | 70–80% |
| Online form with 1 MB limit | 1,200–1,600 px | JPG | 65–75% |
| Messaging app | 1,600 px | JPG | 65–75% |
| Thumbnail or avatar | 200–400 px | WebP or JPG | 75–85% |
A repeatable workflow
- Decide the largest size at which the image will be shown.
- Choose a matching Quick goal in Image Compressor, or set width, format and quality yourself.
- Process one representative image and inspect it at 100% zoom: faces, text, fine textures and skies.
- If it looks good, lower quality by 5 points and compare again. Stop when you can see a difference, then step back up.
- Apply the chosen settings to the whole batch and check the reported sizes.
- Keep the originals in a separate folder.
Why page speed depends on images
On most websites, images make up a large share of the bytes a visitor downloads. A single unoptimised 5 MB hero photo can take several seconds to appear on a mobile connection, during which the page looks broken or empty. Search engines measure aspects of loading experience, such as how quickly the largest element on screen appears, and slow images are one of the most common reasons for poor scores.
Compressing images is therefore one of the cheapest performance improvements available. It needs no code changes, and the savings are often 80% or more for photos taken on modern phones.
Worked example: a monthly newsletter
A community group sends an email newsletter with six event photos. Originally they attached each photo straight from a phone, about 4 MB each, which pushed the message over some recipients' size limits and made it slow to open. Now they compress the batch with the Email goal, which resizes to 1,600 pixels wide at 68% JPG quality. Each photo drops to a few hundred kilobytes, the message sends without bouncing, and the pictures still look sharp on phones and laptops.
Things compression cannot fix
- Blurry or out-of-focus photos stay blurry, however carefully they are compressed.
- An image already saved at very low quality cannot be cleaned up by saving it again at high quality.
- Screenshots with lots of text are better kept as PNG or WebP and reduced in dimensions, rather than compressed hard as JPG.
- Animated GIFs are not handled by the image compressor; use GIF Compressor for those.
How email clients and platforms handle large images
Email services limit the total size of a message, commonly to around 20–25 MB including all attachments, and the limit is set by both the sender's and the recipient's provider. Large images also take longer to send and download, and some mobile email apps offer to shrink attached photos automatically, which can produce unpredictable quality.
Messaging apps and social networks almost always recompress images when you upload them. Sending an already well-compressed image of a sensible size gives you more control over the result, because the platform has less to change. Sending a huge original often produces a softer result, because the platform's own compression is more aggressive.
Images in web pages: beyond file size
On websites, how images are delivered matters as much as their size. Responsive images, using the srcset attribute, let the browser download a smaller version on phones and a larger one on desktops. Lazy loading, with loading=lazy on images below the fold, delays downloading images until the visitor scrolls near them. Setting width and height attributes reserves space so the page does not jump as images load, which improves perceived quality and layout stability.
Many content management systems and site builders handle some of this automatically, generating several sizes from each upload. In that case, upload a reasonably sized, high-quality master, for example 2,000 pixels wide, rather than a huge camera original, and let the platform create the smaller versions.
Batch processing without losing track
When compressing many images, keep the originals in one folder and the compressed versions in another, never in the same place with similar names. Add a suffix such as -web or -email to compressed files so they are easy to distinguish. Process one test image first, decide on settings, then process the rest with the same settings so results are consistent.
If a few images in a batch look worse than the others, they are probably more detailed, such as photos of foliage, textured fabric or crowds. Reprocess those individually at a slightly higher quality rather than raising quality for the whole batch.
Quick answers
What is a good size for an email photo? Around 1,600 pixels wide at moderate JPG quality usually gives a file of a few hundred kilobytes that looks sharp on any screen.
Should I use PNG for website photos? Generally no. Photos are much smaller as JPG or WebP with no visible difference.
Are my images uploaded when I compress them? No. MyFileNest compresses images in your browser.
Understanding the numbers on the quality slider
The quality value is a setting for the encoder, not a percentage of the original image that is kept. Quality 80 does not mean 20% of the detail is thrown away; it means the encoder uses a particular level of rounding that typically looks very close to the original. The relationship between the number and the file size is not linear either: dropping from 95 to 85 often halves the file, while dropping from 60 to 50 saves much less and introduces more visible artefacts.
Different encoders and formats also interpret the number differently. WebP at 75 and JPG at 75 do not produce identical results, and two applications exporting JPG at the same number may give different file sizes. That is why the practical advice is always to look at the result and read the actual file size, rather than relying on a number from another tool or guide.
Privacy when compressing photos
Photos straight from a phone often contain GPS coordinates, the capture time and the device model. When MyFileNest re-encodes an image during compression, that embedded camera metadata is not carried into the new file, which is usually helpful when sharing pictures publicly or with people you do not know well. If you need to keep the details, for example for your own photo library, keep the original file. Compression happens in your browser, so the photos themselves are not uploaded to MyFileNest. Before posting, still look at the picture itself for visible details such as house numbers, number plates or documents on a desk, which no amount of metadata cleaning can remove.
Summary
Resize first to the largest size the image will be displayed at, choose the format that suits the content and destination, then lower quality gradually while watching faces, text and gradients. Keep originals separate, process batches with consistent settings, and check the real output sizes. These steps typically cut camera photos by 80% or more without any visible loss.