Convert TIFF to JPG

Choose the TIFF page before making a JPG

JPEG stores one rectangular image. TIFF can hold a sequence of pages. Those facts make page choice part of the conversion rather than an advanced option: when the source has more than one top-level image directory, this tool requires you to select the exact page. It does not export page one by habit and leave the remaining pages unmentioned.

A preview of the page arrives first. A multipage file lists every page with a thumbnail, its upright dimensions and color depth; a single-page file is selected for you. Stored dimensions, orientation, compression, color interpretation, sample depth, alpha, and resolution sit under Page details. The download name always includes the decision — scan-page-3.jpg, for example — even when the TIFF contains only one page.

TIFF to JPG is deliberately a lossy conversion

TIFF is a container and can store pixels without loss, with simple run-length compression, or with another codec. JPEG always introduces a new lossy generation. Its strength is a compact, widely accepted photographic file; its cost is discarded detail around edges and texture. The quality setting controls that trade, and the receipt describes the output as a new JPEG encode rather than “preserved quality.” No setting can restore information already absent from a compressed TIFF page. When the JPG still has to fit an upload limit, the image compressor works backward from the size you name instead of making you guess at quality numbers.

For diagrams, text scans, palette artwork, sharp screenshots, or anything that must retain the decoded 8-bit pixels exactly, TIFF to PNG is the safer sibling. If the source is a multi-page document rather than one picture, use TIFF to PDF to keep the selected pages together instead of converting them one at a time.

Transparency has to become a visible color

JPEG has no alpha channel. When a decoded page contains declared, genuinely transparent pixels, a solid background choice appears before conversion with white as its default. Associated TIFF alpha is unpremultiplied before compositing; unassociated alpha is used directly. Extra samples that are not declared as either alpha form are rejected, because treating an unknown channel as opacity can erase color data while producing a file that still looks valid.

The receipt says Transparency: Filled with #rrggbb, naming the color that was actually used. When the page is fully opaque, the background control is absent because it could not change a pixel. Controls without an effect make a converter look flexible while concealing which decisions actually matter.

Upright pixels replace an orientation note

A TIFF can say that its stored rows should be mirrored, rotated, or both. The inventory reports all eight orientation values, and the output applies them. Pixels are normalized and dimensions are swapped where necessary; the JPG does not need a new EXIF orientation tag to look correct. The built-in reopen checks JPEG magic and the oriented dimensions. Independent regression fixtures check asymmetric corners; the on-page receipt does not claim that broader oracle ran.

Supported unsigned sixteen-bit grayscale samples must also cross a visible boundary. JPEG's browser encoding path is eight-bit, so values are deterministically scaled and the receipt says Reduced from 16-bit to 8-bit. Signed and floating-point TIFF samples are outside this route and receive a typed unsupported message.

Color and metadata are not silently carried

The supported grayscale, RGB, and palette pages are decoded to an sRGB-like eight-bit working image. Embedded source ICC bytes are neither applied nor copied. Chromium's JPEG encoder currently writes a new sRGB ICC profile; other browser encoders may leave the file untagged or write another profile. The receipt reads the finished JPG's APP2 segments and reports what was actually embedded. GeoTIFF positioning, camera metadata, scanner-private tags, thumbnails, and other directories are not written into the JPG. This makes the file suitable as a general viewing or delivery copy, not as a color-managed master, georeferenced dataset, or archival substitute.

Decoder support is explicit as well. Verified classic TIFF combinations use strips, contiguous samples, and the named raw, PackBits, LZW, Deflate, and bilevel fax paths. BigTIFF, tiled pages, JPEG-compressed TIFF, planar color, Predictor 3, signed or float samples, CMYK, YCbCr, Lab, and unknown combinations stop with the reason. Renaming one of those files cannot force it through because the first bytes and page tags, not the suffix, decide the format.

A receipt that exposes every irreversible step

The finished panel shows the format, dimensions, size, and transparency of the new file. Under Technical details it lists source page count, chosen page number, compression and bit depth, orientation, precision, profile and metadata handling, JPEG quality, and chroma sampling. The quality range stops at 99 because the browser's 100 setting changes chroma sampling as well as compression, causing a sharp and otherwise hidden size jump. The download starts only after an independent reopen confirms JPEG magic and measured dimensions. Changing the page, quality, longest side, or background removes that download; download again from the retained TIFF.

Large files are not refused by a hard size limit. A very large file gets advice before it is read, and you continue locally; once page dimensions are inspected, the memory advice follows decoded pixels rather than compressed bytes, and a running task can be canceled. Nothing is uploaded or written into URL or storage state. You can verify that boundary with the network-checking guide.