The job that is easy once and painful four hundred times
Converting one image is trivial. Converting an entire camera card, a product catalogue or a decade of phone photos is a different task, and the things that make it go wrong are not about image formats at all — they are about ordering, naming, memory and the fact that you only find out something was wrong after you deleted the originals. This guide covers how to plan a batch, what to check before committing, and the specific traps in browser-based bulk conversion.
Decide the target once, properly
Every file in the batch gets the same treatment, so a wrong decision is multiplied by the file count. Three questions settle it:
- Where are these going? A website wants WebP. An upload form or a client wants JPG. An editor wants PNG. Archives should keep the original.
- Is there transparency anywhere in the set? One logo among three hundred photos is enough to make a blanket JPG conversion wrong — that file comes out with a solid background and nobody notices until later.
- Does anything downstream cap file size or dimensions? Resize during the batch rather than discovering the cap afterwards and running everything twice.
Run a single representative file through first, look at it at full size, and only then commit the batch. Five minutes of checking is cheaper than four hundred files at the wrong quality.
What actually decides the output size
| Lever | Effect on size | Effect on how it looks |
|---|---|---|
| Halve the width and height | About 75% smaller | Invisible if the display size is smaller than the new dimensions |
| JPEG quality 0.9 → 0.75 | About 40% smaller | Essentially invisible on photographs |
| JPEG quality 0.75 → 0.5 | About 35% smaller again | Visible blocking in skies and flat areas |
| JPG → WebP at equal quality | 25–35% smaller | No visible difference |
| Strip EXIF metadata | A few KB per file | None, but loses date, camera and GPS |
| Convert photos to PNG | 3–10× larger | None — PNG is lossless, and wasteful here |
The order matters: resize first, compress second. Dimensions are the cheapest saving by a wide margin, because quality loss from downscaling is invisible as long as the result is still larger than where it will be displayed, whereas quality loss from heavy compression is visible at any size.
Names, order and the things that break later
- Sort order is not upload order. A tool typically processes files in the order the browser hands them over, which usually follows the file picker's sort. If the sequence matters — scanned pages, a photo story — check the ordering in the tool before converting, not after.
- Numbers sort as text.
img2.jpgcomes afterimg10.jpgin a plain sort. Pad with zeros,img002.jpg, before you start, and the problem disappears permanently. - Duplicate base names collide. Converting
photo.jpgandphoto.pngfrom different folders into the same format produces two files wanting the same name. Flatten and rename first. - Non-Latin filenames can be mangled by the ZIP step if any tool in the chain is old. Check one extracted file before trusting the whole archive.
- Keep the originals until the batch is verified. Spot-check the first, middle and last file at full size. Conversion is irreversible and a wrong quality setting is only obvious in the details.
What EXIF carries, and whether to keep it
Photographs carry metadata: camera and lens, exposure settings, the date and time, and frequently GPS coordinates accurate to a few metres. Batch conversion is the moment to make a deliberate decision about it, because whatever you choose applies to everything.
- Keep it for a personal archive. Losing capture dates means photo software falls back to file timestamps, and years of chronology collapse into the conversion date.
- Strip it before publishing anywhere public. Home addresses have been revealed through photo GPS more times than is comfortable. Image Compressor drops metadata during re-encoding; JPG Compressor (Keep EXIF) is the opposite choice when you want it preserved.
- Check what is there with the EXIF Data Reader on one representative file before deciding.
Practical limits of converting in a browser
Browser conversion has a real advantage for large batches — nothing is uploaded, so there is no waiting on a network and no copy of your photos on someone's server — and a real constraint: everything happens in your machine's memory, in one tab.
- Batches of a few hundred typical photos are comfortable. Thousands of high-resolution files, or 50-megapixel images, can exhaust the tab's memory.
- If a batch stalls or the tab reloads, split it into chunks of one or two hundred. The result is identical and each chunk finishes.
- Keep the tab in the foreground. Background tabs get throttled, which can stretch a two-minute job into twenty.
- Closing the tab loses the work. There is no server-side job to come back to.
- Output arrives as a single ZIP, which avoids several hundred separate download prompts.
The Bulk Image Converter handles the batch and the ZIP. For a set that needs to land under a specific size limit rather than a quality setting, Compress to Target Size searches for the right quality per file instead of applying one guess to everything.