How to convert images
The flexible converter decodes supported sources and creates JPG, PNG or WebP output. Format choice determines transparency and compression behavior, while animated, vector, multipage and editable source properties require explicit limitations.
Steps
- 1Review source support
Check whether the browser can decode the source and read the specific notices for GIF, sanitized SVG, classic TIFF or compatible PSD composite input.
- 2Choose the output properties
Select JPG for an opaque photographic result, PNG for lossless pixels or transparency, or WebP for a modern raster output that can retain transparency.
- 3Compare the finished copy
Verify dimensions, transparency, format and visible content while retaining the source that contains properties the raster output cannot represent.
Key decisions
What happens to non-raster source features?
GIF contributes its decoded first frame, sanitized SVG is rasterized, supported TIFF pages become separate ordered raster inputs, and compatible PSD contributes its embedded flattened composite.
Which output should preserve transparency?
Choose PNG or WebP when transparent pixels matter. JPG is opaque and requires a solid background.
Before you download
Know the limits
- Browser-dependent AVIF and BMP input can be used only when the current decoder accepts it.
- TIFF support is limited to classic stripped, chunky eight-bit white-is-zero or black-is-zero grayscale, RGB, declared-alpha RGBA or palette pages using None, PackBits, LZW, Deflate or old Deflate compression; horizontal differencing is supported only with LZW or Deflate. The first supported page opens by default. Opening every supported page requires the whole chain to fit the same subset and creates separate rows in document order. BigTIFF, tiles, separate planes, non-top-left orientation, reversed fill order, JPEG and CCITT are excluded. Limits are 25 MB per TIFF; 8,192 pixels, 12 megapixels and 256 strips per page; 16 pages and 16 megapixels combined; plus 150 MB of TIFF sources per batch.
- PSD support is limited to a version-one RGB or grayscale embedded flattened composite at 8 or 16 bit with up to four channels, an exactly four-channel CMYK composite, an exactly three-channel Lab composite, or an 8-bit indexed composite with one palette-index channel and an exact 256-color RGB palette. Raw, RLE and ZIP work at both depths; ZIP with prediction is 8-bit only. Every source becomes an ordinary 8-bit flat sharing copy. CMYK uses a deterministic screen approximation; Lab uses a deterministic D50-to-D65 sRGB conversion. Neither is a color-critical proof. Limits are 25 MB, 8,192 pixels and 10 megapixels per PSD, plus 10 files, 100 MB and 30 decoded megapixels per PSD batch. Layers, plates, spot channels, palette identity and color profiles are not reproduced; PSB, 32-bit, duotone and multichannel files are unsupported.
Verify the copy
Common questions
Does converting a PSD preserve its layers or color reference?
No. A compatible PSD supplies only its embedded flattened composite. CMYK and Lab become bounded screen approximations, not color-critical proofs. Continue editing and preparing print output with the original PSD.
Does an animated GIF stay animated?
No. The converter uses its decoded first frame. Use the GIF maker or an animation-preserving editor when motion must remain.