SVG vs PNG
PNG preserves decoded pixels losslessly and can carry alpha. It is useful for transparent graphics and screenshots; it does not preserve editable document layers or camera sensor data, and photographic files can be large.
Image format
XML-based vector graphics with shapes, text and optional animation. Raster export chooses a specific pixel resolution.
Explore the format
Choose a target format for your SVG images.
Marked paths may need steps such as flattening, rasterization, tone mapping or animation loss.
Choose a source format to convert to SVG.
Scalable Vector Graphics
SVG is a W3C language for describing two-dimensional graphics in XML. Paths, shapes, text, gradients and transformations describe how a picture should be drawn rather than recording only a fixed pixel array. It addresses scalable graphics for diagrams, interface symbols and illustrations that need to render at different sizes.
Opacity, clipping and masking can create transparent compositions. SVG can also reference raster images, fonts, styles and external resources; those dependencies affect portability. Animation and interactive scripting are possible in suitable contexts, but an SVG loaded as a static image may run under a more restricted processing mode.
Use SVG for diagrams, logos and symbols whose shapes should scale cleanly. A photographic image embedded inside SVG remains raster data and does not acquire vector detail. To export PNG or JPEG, choose dimensions, fonts and background deliberately; rendering freezes vector, text and animation semantics into pixels.
SVG describes geometry rather than using a lossy/lossless pixel codec. XML can be compressed for transport, but rasterization is a separate operation that fixes resolution and can discard editability.
SVG develops through W3C specifications for structured web graphics, with a later revision aligning more closely with web platform features.
Different formats, different trade-offs. Choose for the image and the workflow.
PNG preserves decoded pixels losslessly and can carry alpha. It is useful for transparent graphics and screenshots; it does not preserve editable document layers or camera sensor data, and photographic files can be large.
PDF preserves a page description that can mix text, vector paths and raster images. It suits fixed-layout documents; extracting a picture or rasterizing a page requires choosing the page, resolution and colour handling.
WebP combines lossy or lossless 8-bit image coding with alpha and animation. It is a web delivery option when those features matter; high-precision samples and editable source structures need another working format.
ICO packages several icon sizes in one resource. It suits Windows application icons and compatible favicons; a general image format is simpler for a single full-size picture, while icon export requires deliberate size selection.
SVG is a W3C language for describing two-dimensional graphics in XML. Paths, shapes, text, gradients and transformations describe how a picture should be drawn rather than recording only a fixed pixel array. It addresses scalable graphics for diagrams, interface symbols and illustrations that need to render at different sizes.
SVG describes geometry rather than using a lossy/lossless pixel codec. XML can be compressed for transport, but rasterization is a separate operation that fixes resolution and can discard editability.
Embedding the photo only wraps its raster pixels. Reconstructing paths requires a separate tracing process and cannot reproduce arbitrary photographic detail exactly.
Check fonts, linked images, filters and the processing mode. A standalone interactive SVG and the same file used through an image element may expose different features.
Raster export chooses output dimensions and a static rendering state. It loses paths and editable text; JPEG also flattens transparency. Embedded raster images limit the detail available at larger sizes.
Use SVG for diagrams, logos and symbols whose shapes should scale cleanly. A photographic image embedded inside SVG remains raster data and does not acquire vector detail. To export PNG or JPEG, choose dimensions, fonts and background deliberately; rendering freezes vector, text and animation semantics into pixels.