- JPEG remains the universal delivery format; quality 82-90 hits the sweet spot between file size and visual fidelity for most web and social media use.
- WebP and AVIF offer 30-50% smaller files than JPEG at equivalent quality, and browser support in 2026 is broad enough for production web use.
- Color space selection (sRGB vs. Adobe RGB vs. Display P3) must match the output destination -- embedding the wrong profile causes muted or oversaturated colors.
- DPI is a print instruction, not a quality metric. A 4000-pixel-wide image prints identically at 72 DPI and 300 DPI; only pixel dimensions and print size matter.
JPEG: The Universal Format
JPEG has been the default photographic delivery format since the mid-1990s, and despite the arrival of modern alternatives, it remains the safest choice when you need maximum compatibility. Every browser, email client, social platform, messaging app, and operating system handles JPEG without question.
JPEG uses lossy compression based on the Discrete Cosine Transform (DCT). It divides the image into 8x8 pixel blocks, transforms each block into frequency components, and discards high-frequency detail that the human visual system is less sensitive to. The quality slider in your export dialog controls how aggressively this discarding occurs.
At quality 100, almost no data is discarded, but the file is typically only 20-30% smaller than an uncompressed TIFF. At quality 60, files shrink dramatically but compression artifacts -- block boundaries, color banding in gradients, mosquito noise around sharp edges -- become visible, especially on close inspection.
The practical sweet spot lies between quality 82 and 90 (on a 0-100 scale). In this range, file sizes are 60-75% smaller than quality-100 output while artifacts remain invisible at normal viewing distances. For web delivery, quality 82-85 is typically optimal. For print files delivered to clients, quality 90-95 preserves maximum detail with reasonable file sizes.
Never re-save a JPEG from a JPEG. Each save-cycle recompresses the image, compounding artifacts. Always export from your master file (raw, PSD, or TIFF) to a fresh JPEG. If you need to make a minor edit to a delivered JPEG, use a lossless JPEG editor like JPEGCrop to avoid recompression.
PNG, TIFF, and Beyond
PNG uses lossless compression, meaning no image data is discarded during saving. This makes it ideal for graphics with sharp edges, text overlays, screenshots, and images requiring transparency (alpha channel). For photographs, however, PNG files are significantly larger than equivalent-quality JPEGs -- often 3-5 times larger -- because photographic detail does not compress efficiently with PNG's deflate algorithm.
Use PNG when you need transparency, when the image contains text or line art that JPEG would smear, or when you are saving intermediate editing stages that will be re-opened and modified (though PSD or TIFF are better choices for this).
TIFF (Tagged Image File Format) is the archival and professional interchange standard. It supports lossless compression (LZW or ZIP), 16-bit and 32-bit depth, multiple layers (in some implementations), and embedded ICC profiles. Print shops, publishers, and stock agencies frequently require TIFF submissions because it guarantees no compression artifacts and preserves full color depth.
The downside is file size. A 45-megapixel image saved as a 16-bit TIFF with LZW compression can easily exceed 100 MB. For delivery to clients or online use, TIFF is impractical; its role is archival masters and print production handoff.
Modern Formats: WebP, AVIF, and HEIC
WebP, developed by Google, offers both lossy and lossless compression with transparency support. In lossy mode, WebP produces files 25-35% smaller than JPEG at equivalent perceptual quality. Browser support reached effective universality in 2024, and by 2026, WebP is a production-safe web format. Major platforms including WordPress, Squarespace, and Shopify support automatic WebP conversion.
AVIF (AV1 Image File Format) pushes compression further, delivering 30-50% savings over JPEG and roughly 20% over WebP. It supports HDR content, wide color gamuts, and transparency. Encoding is computationally expensive -- significantly slower than JPEG or WebP -- but decoding is fast, making it well-suited for web delivery where the encode cost is paid once during export. Browser support is strong across Chrome, Firefox, Safari (since version 16), and Edge.
HEIC (High Efficiency Image Coding) is Apple's preferred format, used as the default capture format on iPhones since iOS 11. It offers compression similar to WebP with support for depth maps, live photos, and image sequences. Outside the Apple ecosystem, support is inconsistent -- Windows requires a codec extension, and web browser support is limited. For cross-platform delivery, HEIC is best avoided; for Apple-to-Apple workflows, it is space-efficient and feature-rich.
Choosing a file format is not a technical exercise in isolation -- it is a communication decision. The format must serve the recipient's ability to open, display, and use the image as intended.
Color Spaces Explained
A color space defines the range (gamut) of colors available in an image. Choosing the wrong color space for your output causes colors to appear muted, oversaturated, or shifted -- problems that no amount of editing can fix after export.
sRGB is the standard color space for web, social media, email, and most consumer displays. It represents approximately 35% of the visible spectrum and is the assumed color space when no profile is embedded. If you export for screens and do not embed a profile, the image will be interpreted as sRGB. For web delivery, always export in sRGB.
Adobe RGB encompasses a wider gamut than sRGB, particularly in the cyan-green range. It is used in print production where CMYK conversion benefits from the additional color information. If you deliver Adobe RGB files for web display without proper profile embedding, greens and blues will appear desaturated because the browser interprets the wider-gamut values as sRGB coordinates.
Display P3 is the wide-gamut standard used by modern Apple displays, many flagship Android phones, and increasingly by HDR monitors. It covers roughly 25% more of the visible spectrum than sRGB. For photographers targeting Apple devices or modern wide-gamut displays, Display P3 enables more vivid reds, greens, and oranges -- but only when the viewing device supports it and the file's ICC profile is correctly embedded.
Work in the widest color space your camera captures (usually Adobe RGB or ProPhoto RGB) throughout your editing pipeline, then convert to the target color space only at the final export step. This preserves maximum color information during editing and avoids gamut clipping until delivery.
DPI Myths Debunked
Few topics in digital imaging generate more confusion than DPI (dots per inch). The core misunderstanding: DPI is a print instruction, not an image quality metric. It tells a printer how densely to space the pixels when laying them on paper. It has no effect on screen display, file size, or image quality.
A 4000 x 3000 pixel image contains 12 million pixels regardless of whether its DPI metadata reads 72, 150, or 300. Set to 300 DPI, it prints at approximately 13.3 x 10 inches. Set to 150 DPI, the same image prints at 26.7 x 20 inches. The pixels are identical -- only the print size changes.
The persistent "72 DPI for web, 300 DPI for print" rule is a simplification that obscures the real question: do you have enough pixels for your output size? For a high-quality print, you need roughly 240-360 pixels per inch of printed output. For a website hero image displayed at 1200 pixels wide, DPI is irrelevant -- only the pixel width matters.
When clients or print shops request "300 DPI," they mean they need sufficient pixel dimensions for the intended print size at 300 pixels per inch. A request for a 10x8 inch print at 300 DPI translates to a 3000 x 2400 pixel file. Communicate in pixel dimensions to avoid confusion.
Batch Export Workflow Tips
Efficient export workflows become essential as your volume grows. Manually configuring export settings for each image is unsustainable when delivering 50 wedding photos, 200 product shots, or a week's worth of editorial content. Every major editor supports export presets -- build them once and apply them consistently.
Create output-specific presets for your common delivery targets. A typical set might include: Web (sRGB, JPEG quality 85, 2400px long edge, 150 DPI), Instagram (sRGB, JPEG quality 90, 1080px square crop), Client Proof (sRGB, JPEG quality 80, 1600px long edge, watermarked), Print Master (Adobe RGB, TIFF 16-bit, full resolution), and Portfolio (sRGB, WebP quality 85, 2000px long edge).
Automate renaming and organization during export. Lightroom's export dialog supports dynamic tokens for date, sequence number, original filename, and custom text. A naming template like 2026-09-15_ClientName_001.jpg keeps deliverables organized and traceable without manual intervention.
Sharpen at export, not before. Output sharpening must match the final pixel dimensions, which means it should be the last processing step. Lightroom and Capture One both apply output sharpening after resizing within the export pipeline -- use these built-in options rather than sharpening your master file and then resizing.
Verify before delivering. After a batch export, spot-check 5-10 images at full size. Confirm color accuracy (open on a different device if possible), check that sharpening looks correct at the output resolution, and verify that metadata (copyright, contact info, keywords) is embedded as expected. A two-minute check prevents embarrassing delivery mistakes.
Keep your master files in a non-destructive format (raw, PSD, or layered TIFF) and never overwrite them with an export. Treat exports as disposable outputs -- you can always regenerate them from the master. Store masters on redundant storage and exports on fast, temporary drives.