Photo TIFFs vs document TIFFs — different tradeoffs

"TIFF" covers two very different file types. The first is the multi-page bilevel CCITT scan from an office scanner — a 30-page contract is a single 1 MB TIFF. The second is the single 16-bit RGB Photoshop save — a single 300 MB TIFF of one studio photograph. TIFF2PDF handles both, but you should know which you have, because the right post-processing differs.

Document TIFFs

The dominant case in the wild: a workflow built around scanning paper into searchable archives.

• Source: office multifunction printer/scanner (Brother, Canon, HP, Xerox, Ricoh), document feeder scanner (Fujitsu ScanSnap, Epson DS), production-grade book scanner.
• Structure: multi-page TIFF, one IFD per scanned page, all pages sharing similar properties (DPI, color mode).
• Mode: bilevel (text-only documents), grayscale (handwriting, faded prints), or color (forms with stamps, marketing materials).
• Compression: G4 for bilevel (almost universal), LZW or deflate for grayscale/color.
• DPI: 200, 300, 400, 600 — scanner resolution settings.
• File size: 5–50 KB per page for bilevel, 200–800 KB per page for grayscale, 600 KB–3 MB per page for color.

The conversion produces one PDF page per IFD; the source pixels are decoded and re-embedded with the conversion library's default filter (typically Flate). Output is typically 2–3× larger than input for bitonal CCITT sources, similar size for grayscale and color sources. The natural next step is OCR (see the dedicated post) to make it searchable.

Photo TIFFs

The minority case: archival of a single high-resolution image.

• Source: Photoshop save with "Maximize compatibility" enabled (which forces TIFF over Photoshop's native PSD), drum scanner output for film digitization, professional camera RAW conversions, fine-art reproduction scans.
• Structure: single IFD (one image). Sometimes embeds a thumbnail in a second IFD.
• Mode: 8 or 16-bit RGB, sometimes 8-bit CMYK for pre-press, occasionally with alpha channel (8-bit RGBA).
• Compression: LZW (most common, lossless), deflate (modern alternative), JPEG-in-TIFF (lossy when size matters), uncompressed (rare; for time-critical workflows).
• DPI: 300 (offset print standard), 600 (high-quality), 1200 or higher (drum scanner film).
• File size: 5–500 MB. A 16-bit 6000×4000 RGB TIFF (medium-format film scan) is ~144 MB uncompressed (6000×4000×3×2 bytes); LZW-lossless compresses it somewhat less on photographic content.

The conversion produces a single PDF page sized to match the original physical dimensions (image width / DPI × 72 per axis). PDF carries pixel components at 1, 2, 4, 8, or 16 bits — 16-bit RGB content remains high-precision when the underlying conversion preserves it; some pipelines reduce 16-bit to 8-bit for compatibility, which is visually transparent on a screen but discards subtle gradient headroom.

Layered Photoshop TIFFs

Photoshop can save TIFFs with non-standard "private" tags storing layer information, masks, alpha channels, and editing history. The base image is a flattened version that any TIFF reader can show; the private tags carry the editable layered structure.

The conversion reads the base flattened image. Layers, masks, and Photoshop-specific private tags are not represented in the output PDF — the resulting PDF is a single image of the visible composition.

If you need layered editability preserved, save as the native format of your image editor (PSD or similar) and use a different tool to create PDFs that retain layers — most full-featured commercial PDF tools support a "PDF/X with layers" export.

HDR TIFFs

High dynamic range TIFFs use 16-bit or 32-bit floating-point per channel, capturing a much wider range of brightness than standard 8-bit imagery. Used in:

• HDR photography workflows (multiple exposures merged).
• Computer graphics (rendered output before tone mapping).
• Astronomy (long exposures with very dark and very bright regions).

PDF doesn't support floating-point image data. Floating-point TIFFs are converted to integer-valued pixels by the underlying conversion library before embedding, using whatever default reduction the library applies (typically a linear or clipped mapping rather than a perceptual tone-map). The result is a viewable PDF but loses HDR information; for control over the tone-mapping, run the float TIFF through a dedicated HDR-to-LDR tool first, then upload the resulting 8-bit TIFF.

Size economics

Document TIFFs are tiny per page; the resulting PDFs are still small but not as small. A 200-page bilevel CCITT scan archive of ~10 MB becomes a 25–35 MB PDF after re-encoding to Flate. For maximum size efficiency on bitonal archives, look for specialist archival software that preserves CCITT compression inside the PDF.

Photo TIFFs are huge; the resulting PDFs match. A 300 MB studio photo TIFF becomes a roughly 300 MB PDF. For sharing or web distribution, recompress the photo first: open the TIFF in any image editor, "Save As" a JPEG at quality 85 (likely 30× smaller), then upload to JPG2PDF for a small JPEG-based PDF.

When the source format should change

If you're producing PDFs for screen viewing or web distribution and your TIFFs are photo-grade, lossless preservation isn't useful — JPEG-based PDFs are 5–10× smaller with no visible quality loss. Convert to JPEG first, use JPG2PDF instead.

If your TIFFs are document-grade (bilevel scans), TIFF2PDF produces faithful output suitable for OCR and archival. For tighter size control on bitonal sources, specialist archival software is the right tool.