Read the on-screen status, then check format → file bytes → decoded width and height → browser or network → vector complexity. Renaming an extension or placing the same file in a ZIP does not convert its contents and usually hides the useful clue.
The limits that are actually enforced
DrowPack checks the uploaded byte size separately from the decoded image dimensions. A 10MB PNG can pass the first check and still fail after decoding if it exceeds 64 megapixels. A modest-resolution PNG can fail earlier if the file itself is larger than 32MB.
| Input or operation | Current limit | What happens | Direct fix |
|---|---|---|---|
| PNG, JPG, or WebP | 32MB maximum per file | The app reports that the image file is too large | Export at smaller pixel dimensions or resave without unnecessary metadata |
| SVG upload | 8MB maximum per file | The app reports that the SVG file is too large | Remove unused paths, metadata, and hidden objects, or use PNG/WebP |
| PDF or AI upload | 40MB maximum per file | The app reports that the PDF/AI file is too large | Save the needed first page separately or export a raster preview |
| Decoded raster image | 16,384px maximum per side and 64MP total | The app reports that the image dimensions are too large | Reduce both the longest side and the total pixel count |
| PDF/AI first-page render | 8,192px maximum per side and 12MP total | The render scale is capped within these bounds | For finer detail, export the required artboard directly as PNG or JPG |
The 64MP rule is width multiplied by height, not the sum of the dimensions. A 12,000×5,000px image is 60MP and also stays under the 16,384px side limit. A 10,000×7,000px image is 70MP, so it fails even though neither side is individually too long. Check both values in the file properties or your editor's image-size panel.
DrowPack renders only the first page of a PDF/AI file to a white preview canvas. Subsequent image exports and SVG conversion start from that rendered canvas. Export the required page or artboard first if you need another page, native layers, or controlled rasterization.
A diagnostic order that avoids wasted work
- Record the exact status message. “File too large,” “dimensions too large,” and “no PDF-compatible data” point to different remedies.
- Confirm that the file is not empty. A zero-byte file, an incomplete cloud placeholder, or an interrupted download must be downloaded or exported again.
- Check the real format, not only the suffix. Supported uploads are PNG, JPG/JPEG, WebP, SVG, PDF, and PDF-compatible AI. Renaming another format to
.pngdoes not change its encoding. - Compare its byte size with the table. If it exceeds that limit, changing browsers will not make the validation pass.
- For raster input, inspect decoded dimensions. Test the 16,384px-per-side and 64MP limits independently. Highly compressed files can be small on disk but huge in memory.
- For PDF/AI, check compatibility and connectivity. AI must contain PDF-compatible data. If the PDF rendering library cannot load, restore network access and retry before re-exporting the artwork.
- If upload works but output fails, reduce working complexity. Lower the preview-canvas size first, then reduce colors, detail, tiny fragments, and path count as needed.
Fixes for each design-file format
AI: save with PDF compatibility
A file ending in .ai is not automatically readable by a browser. DrowPack can open it only when it contains a PDF-compatible representation. Resave it from Illustrator with PDF compatibility enabled, or export the target artboard as PNG or JPG. The “no PDF-compatible data” message means the browser-readable PDF representation is absent; it does not mean the filename itself is wrong.
PSD: export a flattened preview
DrowPack does not open an original PSD directly. Export the visible composite from Photoshop or another compatible editor as PNG, JPG, or WebP, then upload that file. Choose PNG or WebP when transparency matters. JPG is often smaller for opaque photography or painted artwork. Merely changing .psd to .jpg does not perform an export.
PDF: verify page one and the renderer
If the artwork is on page two or later, save that page as a new one-page PDF or export it as an image. When the message says the PDF rendering library could not be loaded, check offline mode, network blocking, and browser extensions before rebuilding the file. A general render failure can indicate damaged data or a PDF structure the browser cannot read; opening and resaving it as a new PDF is the quickest controlled test.
SVG: distinguish the 8MB upload limit from the preview safeguard
An SVG uploaded into DrowPack is limited to 8MB. A newly generated pure-vector SVG has a separate, lower safeguard for inline display: when it is over 1,800,000 bytes (about 1.8MB) or has more than 9,000 paths, DrowPack skips rendering it inside the page to prevent browser lag. The generated SVG remains available, and download still works. “Preview skipped” is therefore not a failed conversion.
White seams, white edge noise, and heavy vectors
These symptoms need different adjustments. Inspect the result at 100%: determine whether a consistent hairline lies between neighboring color areas, many small white fragments sit around an edge, or the entire browser slows under the path count.
| Visible symptom | First adjustment | Why it helps |
|---|---|---|
| Thin white seams between filled shapes | Enable color seam fixing and set its width between 0.8 and 1.0 | A thin stroke matching each fill covers rendering gaps between paths |
| Small white dots or fragments along edges | Lower color count or increase small-dot removal | Fewer antialiased intermediate colors become separate paths |
| Photo trace produces a very large SVG | Reduce color count and detail | Fewer regions and boundaries reduce both path count and file size |
| Logo or line art is unnecessarily complex | Use the dedicated logo or line-art mode instead of photo tracing | The tracing assumptions better match the input structure |
| Complex art remains too heavy as pure vector | Use original-quality SVG or PNG/WebP | Photographic texture no longer needs thousands of editable paths |
Original-quality SVG preserves the appearance by carrying a raster image inside the SVG container. It is not the same as a pure vector whose color regions can be edited as paths. Path count is not a quality score; choose according to whether you need editable geometry or reliable visual fidelity.
When SVGZ or ZIP download fails
SVGZ depends on CompressionStream
SVGZ is SVG text compressed with gzip. DrowPack uses the browser's CompressionStream feature, so the SVGZ action is unavailable or reports no support when that API is missing. A compression error does not erase the vector result. The supported fallback is to download the regular SVG; it contains the same vector content without gzip compression and can be compressed later with another tool if required.
For ZIP failure, reduce the current image dimensions first
The package operation generates several JPGs from the current preview canvas and adds the latest SVG when one exists, then assembles those files in browser memory. If DrowPack reports “ZIP creation failed,” reduce the current canvas dimensions and create the package again. Re-uploading the same large input or repeatedly refreshing does not reduce the memory needed. If an important SVG is already ready, download it separately before changing the canvas.
- Close other memory-heavy tabs and retry the package once
- Reduce the preview's long edge to the actual size needed for publishing
- For a huge pure vector, lower colors, detail, and tiny-path generation
- If only SVGZ fails, verify that regular SVG download works
- Save any finished result before refreshing or switching browsers
A 30-second check before retrying
- Classify the stage. Decide whether the message concerns bytes, decoded dimensions, format, rendering, or export.
- Check the numbers. Compare against 32MB raster, 8MB SVG, 40MB PDF/AI, and the 16,384px/64MP decoded-image limits.
- Resolve format-specific requirements. Use PDF-compatible AI, export PSD to PNG/JPG/WebP, and isolate the required PDF page.
- Reduce complexity only when input already works. Start with canvas dimensions, then adjust colors and detail.
- Use the fallback output. Choose SVG when SVGZ is unavailable, or original-quality SVG/PNG/WebP when pure vector is impractically heavy.
If one source still fails after these checks, preserve it and export a small test copy from the original editor. When the test copy opens, the original file's structure or scale is more likely to be the cause than the browser itself.