How to reduce a photo below a target KB limit
A target size asks the tool to find a file at or below a limit, not to produce an exact byte count. Check format and dimension requirements too, so a smaller file is still usable.
List the format, dimensions and size limit first
If a destination requires “JPG, 800×800px, no more than 500KB”, all three conditions apply. Converting to JPG, setting dimensions and reducing encoded bytes are separate choices. First check whether you are allowed to make the dimensions smaller.
PhotoLight calculates 1KB as 1000 bytes. Entering 500 requests a file at or below 500,000 bytes; it does not pad a smaller output up to exactly that size. This is an input example, not a universal upload requirement.
Enable the target and inspect the actual result
- Choose your files in Compress photos.
- Select a format accepted by the destination. Set dimensions and background first if those are specified too.
- Enable “Target size” and enter the KB limit.
- Process the photos, then check target status, actual bytes, dimensions and visible quality.
- Save results that meet your needs.
For JPG and WebP, the tool tries encodings from your selected quality toward lower quality settings and compares the actual byte counts. When it finds a result below the target, it uses a fitting result. The quality number is not a file-size reduction percentage or a percentage of original detail retained.
Download any existing output you want to keep before changing settings. A change clears the previous results, and the next run uses the current settings.
Allow additional downscaling only when dimensions may change
If quality changes are not enough, “Also reduce dimensions to reach the target” can make the width and height smaller before encoding again. The result shows its actual dimensions and whether additional downscaling was used.
Reducing dimensions further would break a requirement for an exact width and height. Additional downscaling is therefore not used in exact-dimensions mode. Compare allowed formats and quality settings while keeping those dimensions; if no suitable result fits, recheck the requirements.
PNG does not use the lossy quality adjustment. For a large PNG that needs transparency, consider smaller dimensions or compare WebP if the destination allows it. JPG is another option when transparency is unnecessary, but transparent areas will receive a background color.
A missed target is not a successful size match
“Target size not met” means the output is larger than the requested limit. It can be smaller than the original and still fail that condition. Uploading a flagged result may therefore be rejected.
| Condition to preserve | Settings to consider |
|---|---|
| Exact width and height | Keep dimensions; compare allowed formats and quality |
| Only a maximum edge is specified | Consider a smaller longest edge or additional downscaling |
| Transparency is required | Compare PNG and WebP and inspect the background |
| Only one format is accepted | Adjust within that format and recheck the output conditions |
Even additional downscaling cannot guarantee every very small target. Warnings use the encoded file’s actual bytes. A file that loses necessary detail may be unsuitable for the task even if its byte count passes.
Check omitted ZIP entries and the file you upload
Over-target outputs are excluded from “Save all as ZIP”. If only some results qualify, only those are included. If none qualify, ZIP saving is unavailable. “Save over-target result” downloads an individual file without changing its missed-target status.
The ZIP has its own size; it is not the size of any one photo inside it. Extract image files before uploading when the destination expects photos. File-manager labels can use different units and rounding, so check the destination’s actual acceptance too.
You can download an original and a measured output from the file-size and dimensions example. That example’s reduction percentage does not predict the result for another photo. Judge the result produced from your own file.