JPG to PNG Converter — Lossless Output, Honest About Limits
Stops further compression loss. Won't recover detail JPEG discarded.
Converting JPG to PNG is a one-way ticket out of the lossy compression cycle: every subsequent save in PNG preserves every pixel exactly. What it does not do — despite claims on many converter sites — is restore quality already lost when the JPEG was first encoded. The DCT artifacts, ringing, and chroma subsampling baked into the JPEG remain in the PNG. Drop JPEGs below, click Convert, get a lossless PNG copy ready for editing, compositing, or background removal. All processing runs in your browser via the Canvas API.
Real-World Benchmark — Our Own Tests
Below are file sizes (rounded to the nearest readable KB; byte-exact values in src/assets/benchmarks/jpg-to-png/results.json) from running this exact converter (tooldev.app v1.5) on five representative JPG inputs synthesised from our reference PNG set. No theoretical estimates — these numbers reflect the actual Canvas-API output produced by the same engine you are using on this page.
| Sample | Source JPG | PNG output | Size change |
|---|---|---|---|
| icon (256×256) | 9.9 KB | 59 KB | 6.0× larger |
| photo (1920×1280) | 137 KB | 710 KB | 5.2× larger |
| screenshot (1440×900) | 50 KB | 160 KB | 3.2× larger |
| gradient (1280×720) | 20 KB | 196 KB | 10.0× larger |
| composite (1280×853) | 36 KB | 244 KB | 6.9× larger |
Methodology & honest caveats
Test images are all self-generated by tooldev.app (see src/assets/benchmarks/samples/SOURCES.md). Reference PNGs were first re-encoded to JPG at quality 85 via the same engine to produce realistic JPG inputs, then converted to PNG. Conversions executed via tooldev.app v1.5 on 2026-05-20 in Chromium 148 on macOS. PNG output is lossless by design — there is no quality slider, so the table shows one output column. Reported file sizes are rounded to the nearest KB from the actual Canvas-API toBlob output. In our tests JPG → PNG did not reduce file size, and for normal photos and screenshots it usually grows. PNG is a lossless container, so it faithfully preserves the decoded pixels of the (already lossy) JPG source — in this benchmark that means 3–10× the original size. The JPEG compression artifacts have been baked into those pixels, and PNG's DEFLATE compression handles random-looking artifact noise poorly, which is why the size ratio is largest for the gradient sample (10.0× larger). Use this conversion when a downstream workflow explicitly requires a PNG container — for example before adding transparency in another tool, or for a system that does not accept JPG input. Do not use it expecting storage savings. Sample files: icon.png, photo.png, screenshot.png, gradient.png, composite.png — all in our repository under src/assets/benchmarks/samples/.
Your Files Never Leave Your Browser
Conversions run locally in your browser using the Canvas API, with Web Workers used when supported. Your image data stays on your device and is never uploaded to any server.
We use Google Analytics and Google AdSense for aggregated traffic stats and contextual ads when consent allows it. Theme and language preferences stay in your browser. We never see, read, or store the images you convert.
Browser-based conversion has trade-offs: large files (>50 MB) may hit memory limits; animated GIF output flattens to a single frame; EXIF metadata is stripped; ICC color profiles may differ across browsers.
> jpg_to_png_technical_truth
JPG to PNG is one of the most misunderstood conversions on the web. Many tools advertise it as a way to "fix" or "improve" a JPEG — that is not how lossy compression works. Here is what actually happens during the conversion and what you can realistically expect from the output.
Why PNG cannot recover what JPEG threw away
JPEG compression discards information permanently. The encoder splits the image into 8×8 pixel blocks, applies a discrete cosine transform (DCT) to each block, then quantizes the resulting frequency coefficients — the higher-frequency coefficients (fine detail) get aggressively rounded toward zero. That detail is mathematically gone. It is not stored in the file in any reduced form. There is no algorithm — including PNG, AI upscalers, or "JPG repair" tools — that can recover the original bits from quantized coefficients alone.
What you see when you decode a JPEG is the inverse DCT of those quantized coefficients. The decoded image contains approximations of the original — close enough that human vision typically does not notice the difference, but mathematically not the same image. When that decoded raster is encoded into PNG, PNG preserves it exactly as decoded — quantization noise, ringing artifacts, blocking artifacts, and all. The PNG is a perfect copy of an imperfect source.
How PNG compression actually works (DEFLATE + filters)
PNG uses DEFLATE — the same compression algorithm as gzip and ZIP — combined with one of five line-by-line filters applied before compression. The filters (None, Sub, Up, Average, Paeth) transform pixel values into delta sequences that compress better with DEFLATE's LZ77+Huffman pipeline. The encoder picks the filter that yields the smallest result for each scanline.
This is fundamentally different from JPEG's frequency-domain approach. PNG works in the spatial domain — every pixel is preserved, but redundancy between neighboring pixels is exploited through prediction. DEFLATE excels at compressing structured data with repetition (logos, screenshots, flat colors), and underperforms on the noisy decoded output of a JPEG. That is why a PNG of a decoded JPEG is typically 3–5× larger than the source JPEG: PNG is forced to preserve all of JPEG's quantization noise, which does not compress efficiently.
When the conversion is genuinely useful
Despite the disclaimers above, JPG to PNG conversion has clear, defensible use cases — they are just specific. The conversion makes sense when you want to:
- Stop further generation loss before a long editing session. Once in PNG, every save is lossless.
- Enable transparency editing by removing the background later in a PNG-aware editor.
- Comply with platform requirements that demand PNG specifically (some logo upload portals, slide deck templates, digital signage software).
- Composite onto another image without baking JPEG block boundaries into the output.
> how_to_convert
- Upload your JPG fileDrag and drop your JPEG image into the converter, or click "browse files". Multiple files are supported for batch conversion.
- Select PNG output formatClick the PNG button in the format selector. The quality slider will hide — PNG is always lossless, no quality setting needed.
- Download your PNGClick "Convert All", then download your lossless PNG files individually or as a ZIP archive.
> jpg_vs_png
| Feature | JPG | PNG |
|---|---|---|
| Compression | Lossy | Lossless |
| File size | Smaller | Larger |
| Transparency | ✗ No | ✓ Yes |
| Re-save quality | Degrades each save | No degradation |
| Best use case | Web photos | Graphics, editing, logos |
| Browser support | All browsers | All browsers |
> when_not_to_convert_jpg_to_png
JPG to PNG is the wrong choice more often than the right one. Five scenarios where you should keep the JPEG (or convert to a different format instead):
1. You are deploying images to a website
For web delivery, PNG is almost always the wrong format for a JPEG photograph. The PNG will be 3–5× larger with no visible quality benefit (since the source JPEG already discarded the detail PNG would otherwise preserve). For web use, convert to WebP at quality 85 instead — same visual result, file size 30–50% smaller than the original JPEG, far smaller than any PNG of the same image.
2. You believe PNG will repair JPEG artifacts
It will not. PNG preserves the decoded JPEG exactly as-is, including all blocking, ringing, and chroma subsampling artifacts. If your goal is to remove JPEG artifacts, you need an artifact-reduction tool (FFmpeg's smartblur filter, AI-based detail recovery models, or manual touch-up in Photoshop). The format conversion alone does nothing.
3. You need EXIF metadata preserved
PNG technically supports text metadata via tEXt, iTXt, and eXIf chunks (the latter added in 2017), but Canvas API conversion in the browser does not carry over EXIF from the source JPEG. GPS coordinates, camera model, shooting parameters — all stripped. If you need metadata in the output, use a desktop tool like ImageMagick with -strip off, or ExifTool to copy EXIF after conversion.
4. The image is a photograph that will not be edited
If you simply want a copy of a JPEG photograph for storage, sharing, or display — keep it as JPEG. Converting to PNG inflates the file size with zero visual benefit. The only legitimate reason to put a JPEG photograph into a PNG container is if downstream software (a CMS, a slide template, a print bureau form) explicitly requires PNG input. Otherwise, leave the JPEG alone.
5. You only need transparency for a JPEG that has none
Converting a JPEG to PNG does not magically add transparency — the converted PNG has the same solid background as the source. If your goal is a transparent version of a JPEG (a logo on a white background that you want isolated), you need a background-removal step first — strip the background in a dedicated tool, then export the result as PNG.
> when_jpg_to_png_is_the_right_choice
JPG compresses photographs efficiently, but every time you save a JPG, you lose a small amount of quality — and that loss compounds. If you plan to edit an image, overlay it onto other content, or run it through multiple tools, starting with a lossless PNG means no further degradation during the editing pipeline. PNG also supports transparency, which JPG fundamentally cannot provide.
Use cases where converting JPG to PNG makes practical sense:
- Background removal — tools like Photoshop's Remove Background, GIMP, or AI-based removers produce cleaner results from PNG source files. The lossless data gives edge detection algorithms more precise pixel information to work with.
- Images edited multiple times — if an image will be cropped, filtered, adjusted, and re-exported across multiple editing sessions, start in PNG to avoid compounding JPG compression artifacts at each save.
- Slide decks and documents — PowerPoint, Keynote, and Google Slides embed images more cleanly in PNG format. JPGs embedded in slides can look noticeably worse when the presentation is re-exported to PDF or projected at high resolution.
- Logo and watermark overlay work — placing a JPG element over another image produces cleaner edges in PNG, especially when precision matters around soft edges or gradients.
- Archiving before editing — a standard workflow: convert to PNG first, make all edits in PNG, then export the final result to JPG or WebP. This avoids generational quality loss across multiple export cycles.
Keep in mind: the PNG output will be noticeably larger than the source JPG — often 3–5x. That's the expected trade-off for lossless quality. Convert back to JPG or WebP when you're done editing.