Choose the right image format before you convert
A smaller file is not automatically a better file. Start with the image’s job—photo, transparent graphic, archive or web delivery—then choose the format that keeps the properties you actually need.
Use JPG for photographs that do not need transparency
JPG is broadly supported and often compact for camera photographs, gradients and other continuous-tone images. Its compression is lossy: lowering quality removes detail and can create halos or blocks around sharp edges.
Because JPG has no transparent pixels, a transparent source must be flattened onto a solid color. PixHandy uses a visible warning so a logo or cut-out does not unexpectedly gain a dark background.
Use PNG when exact pixels or transparency matter
PNG is lossless and dependable for interface graphics, screenshots, diagrams and logos. It preserves transparent pixels and crisp edges, but photographic PNGs can be much larger than JPG or WebP.
Choose PNG when editing headroom or exact edge quality matters more than transfer size. For a final web photo, compare a WebP or JPG result before publishing.
Use WebP for modern web delivery
WebP supports transparency and usually produces smaller web-ready files than PNG for the same photographic content. Modern browsers support it well, but a legacy publishing system may still expect JPG or PNG.
PixHandy lets you compare the output byte size before download. If a compression attempt makes a file larger, the compression tool keeps the smaller original instead of claiming a false saving.
Use still GIF only when a destination requires it
GIF can store no more than 256 colors in one palette. That can suit flat artwork with few colors, but photographs and smooth gradients may show visible banding and can be larger than a WebP version.
The JPG-focused converter creates one separate single-frame GIF for every JPG. It preserves dimensions within the safe GIF bound, then proportionally reduces both dimensions when either side exceeds 4,096 pixels. Selecting still GIF output for several JPGs keeps those files independent; choose Animate these JPGs when you want the same ordered selection carried directly into one timed animation.
Treat TIFF as an original or archive, not just a web format
TIFF is a container used by scanners, publishing, archives, scientific imaging and geospatial workflows. It can hold multiple pages, higher bit depths, specialist color data and rich metadata. A JPG, PNG or WebP sharing copy cannot preserve all of those properties, so keep the original TIFF.
PixHandy deliberately accepts a bounded subset: classic stripped, chunky 8-bit white-is-zero or black-is-zero grayscale, RGB, RGBA with declared alpha, or palette TIFF pages stored with None, PackBits, LZW, Deflate or old Deflate compression. Horizontal differencing is accepted only with LZW or Deflate. The first supported page opens by default; opening every supported page creates separate rows in document order and requires the complete chain to fit the same subset. Limits are 25 MB per TIFF; 8,192 pixels per side, 12 megapixels and 256 strips per page; 16 pages and 16 megapixels combined; with 150 MB of TIFF sources per batch. BigTIFF, tiles, separate planes, non-top-left orientation, reversed fill order, JPEG, CCITT and higher bit depths are rejected. Page structure and specialist metadata are not copied.
Treat PSD as an editable original, not a flat sharing format
A Photoshop PSD can contain layers, masks, editable text, effects, channels and document metadata. JPG, PNG and WebP are flattened pixel outputs, so converting a PSD cannot preserve the document for continued layer editing. Keep the original PSD even when the sharing copy looks correct.
PixHandy reads only the embedded flattened composite from a bounded subset: PSD version 1, 8- or 16-bit RGB or grayscale with up to four channels, exactly four-channel CMYK, exactly three-channel Lab, or 8-bit indexed color with one palette-index channel and an exact 256-color RGB palette. Raw, PackBits 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 file, plus 10 PSD files, 100 MB of PSD sources and 30 decoded megapixels per batch. Layers, plates, spot channels, palette identity and color profiles are not reproduced. PSB, 32-bit, duotone and multichannel files are rejected. A PSD saved without a compatible composite cannot be converted here.
What about SVG, AVIF and BMP?
SVG is vector markup rather than a pixel format. PixHandy sanitizes self-contained SVG input before raster conversion, removing scripts and external resources. That security boundary can change the appearance of SVGs that depend on linked assets or active content.
AVIF and BMP conversion depends on the browser’s decoder. Animated GIF input is deliberately converted from its first decoded frame; use the GIF maker when the output should remain animated.
- Keep an original source file before any lossy conversion.
- Inspect transparency and small text at full size.
- Judge quality and byte size together, not from the extension alone.