Lossless Photo Formats Explained: Which to Use (2026)

Published November 7, 2026 · Daniel Tello · RevelRaw

You've just copied a card to your Mac, and the export menu offers JPEG, PNG, TIFF and HEIF, quality sliders, bit depth menus and color spaces. You don't want to throw away detail by accident, but you also don't want a folder of files too large to upload or back up sensibly.

Lossless isn't a label on a file extension. It's a property of the whole workflow. One file stays untouched as your camera original, another serves as the working master while you edit, and a third is the delivery copy made for a screen, a lab or a client. Once you sort those three jobs, the format choices get much simpler.

Which Photo Formats Are Lossless?

Some formats can store your pixels exactly, some can't, and several can do either depending on the setting you choose.

Common photo formats by compression behavior:

How to read this table

Read the middle columns together. The extension alone doesn't decide the outcome; the compression setting inside the file does. A TIFF saved with LZW keeps every pixel, while a TIFF saved with JPEG compression doesn't. A RAW file can be recorded with lossless compression on one camera setting and lossy compression on another; the comparison of Sony compressed, uncompressed and lossless RAW shows what that trade involves on one brand.

Rule of thumb: Keep originals and masters lossless, and match the delivery format to where the image will go.

Before every export, open the compression or quality setting, not only the file type menu. A quality slider usually means a lossy file. Options such as LZW, ZIP or no compression for TIFF mean you're in lossless territory. That ten-second check prevents most accidental quality loss.

What Does Lossless Mean for Your Pixels?

A lossless save makes one promise. You write a file, close it, reopen it, and the decoded pixels match the ones you stored, value for value.

A zip archive works the same way. You zip a folder of documents, send it to a friend, and they unzip it. Every document returns byte for byte, or the archive has failed. A photocopy behaves differently: copy a page, then copy that copy, and each generation drifts a little further from the original. Lossless formats sit on the zip side of that divide.

How lossless compression shrinks a file

Lossless compression doesn't throw detail away. It writes repeated patterns in a shorter form. A clear sky holds thousands of neighboring pixels with nearly the same value, and the encoder can describe that pattern far more compactly than by listing every value. When you reopen the file, the decoder rebuilds every pixel from that description.

Lossy compression goes one step further. It also discards detail your eye is unlikely to miss at the intended viewing size. That discarding is permanent. No decoder can bring back the exact values that were removed, which is why the format matters most at the moments a file gets saved.

The part most guides skip

Lossless describes the encoding, not everything you'd like a file to remember. RAW keeps largely unprocessed sensor data, while TIFF and PNG keep decoded pixels after that sensor data has been turned into a picture. Neither TIFF nor PNG brings back the original sensor reading, and neither keeps your RAW editor's adjustable settings.

The same word can mean different things across formats. As the JPEG XL reference notes, JPEG XL supports lossless compression and can losslessly recompress existing JPEG files, typically about 20% smaller, in a way that can be reversed to rebuild the original JPEG bit for bit. The JPEG's earlier loss stays, but the recompression adds no new loss. Browser support is recent: Safari has shown JPEG XL since version 17, and Chrome and Firefox have only just added it, so older browsers and many apps and upload forms still won't take it.

Three questions tell you which role a format can fill:

• Does it keep sensor data? That's the camera original, usually RAW.
• Does it keep decoded pixels exactly? That's a lossless TIFF, PNG or similar master.
• Does it keep editability? That depends on layers, sidecars and stored settings, not only on pixels.

To check a setting yourself, save a test file, reopen it, and compare it with the original using a difference blend, or compare checksums of the decoded pixel data. A zero difference confirms that round trip was lossless for that format and setting.

How Does Lossy Compression Cost You Detail?

Lossy isn't simply bad and lossless simply good. What matters is where in the workflow the loss happens.

What gets discarded first

Lossy encoders sacrifice the detail you're least likely to notice on first viewing: fine texture in foliage, subtle tonal steps in a sky, precise color in smooth gradients and crispness along high-contrast edges. At a sensible quality setting and final size, you may see no problem at all. The trouble begins when you edit that file hard, enlarge it, or save it repeatedly.

Open a JPEG, adjust exposure and color, save it as a JPEG again, and repeat the cycle five times. Each save recompresses a picture that already contains small errors from the last save, so the errors stack. This is generational loss, and it's the main reason experienced photographers won't edit from a JPEG if a better source exists. Our RAW versus JPEG comparison for Sony Alpha shooters shows how much more room the RAW version leaves you when you push an edit.

Symptoms you can spot yourself

Take a clear evening sky and push contrast hard after several JPEG saves. You'll often see bands where the tone should change smoothly. Make the same edit from a high-bit-depth lossless master and the sky usually stays clean, because no encoder has rounded away the in-between tones.

• Blocky patches: Flat areas break into visible squares or blotches.
• Halos and ringing: Light or dark fringes appear along sharp edges, such as branches against sky.
• Color steps: Smooth gradients turn into stripes after strong exposure or saturation lifts.
• Muddy texture: Grass, hair and fabric lose their fine structure and look smeared.

When lossy is the right choice

A single high-quality JPEG, exported once at the size it will be viewed, often looks identical to the master it came from. That's the format doing its intended job at the end of the chain. Lossy delivery makes sense when the file won't be edited again, the viewing size is known, and small size matters for uploading or sharing.

Never let a lossy file become the source for your next edit.

Keep that one rule and you can use JPEG and HEIF confidently for delivery. Break it, and every new version starts from a slightly worse photograph than the one before.

How Big Do Lossless Files Get?

Lossless files are larger than delivery JPEGs, and how much larger surprises many people the first time they export a master.

Measured file sizes for lossless formats:

The scan figures come from a 2026 UglyPear comparison of TIFF and PNG for scanning by a maker of file-compression software, where a 300 DPI A4 color scan (2,480 × 3,508 pixels, 8 bits per channel) measured 24.6 MB as uncompressed TIFF, 8.2 MB as TIFF LZW and 7.1 MB as PNG. The camera figures come from a comparative table in a 2026 academic article indexed by PubMed Central. The article is about archiving images in a small herbarium's digitization workflow, and its 80-megapixel files came from a Panasonic Lumix DC-G9, a camera with built-in pixel shift, with the derivative files made in Adobe Camera Raw.

Why a TIFF dwarfs the RAW file

A RAW file stores one value per photosite on the sensor. When software decodes it into a picture you can edit, every pixel needs three color values, one each for red, green and blue. The data expands because the software has filled in the full-color picture through demosaicing.

The arithmetic is plain. At 16 bits per channel, each pixel needs 6 bytes, so 80 million pixels need about 480 MB before compression. That sits close to the 461 MB measured for the TIFF, which suggests it was saved with little or no compression. The 125 MB RAW shows how compactly a camera format can pack sensor data before that expansion happens, and the 64.1 MB JPEG 2000 shows that a stronger lossless codec can shrink even the decoded pixels a long way.

The DNG in that table was made from the camera's RAW, yet it came out larger. The article doesn't break the difference down; a DNG can carry full-size previews and uses its own compression, so converting doesn't always save space. For how camera settings change these numbers, see our guide to how big a RAW file gets.

Planning storage for a shoot

Multiply your expected frame count by the size of the format you'll keep, then add a generous margin for masters and exports. A wedding photographer who masters only selected frames needs far less space than one who converts every frame to TIFF. Treat these figures as orders of magnitude rather than exact budgets, since subject detail changes how well any image compresses. Smooth skies pack down well; busy foliage and high-ISO noise don't.

Which Lossless Format Should You Edit In?

Assign a format to each of the three roles and most of the decision makes itself.

The original is the RAW file, and it stays untouched. With its settings stored alongside, it works like a negative with a notebook of instructions: the capture stays intact and the recipe stays editable. You can revisit white balance, exposure or color months later, because nothing has been baked into pixels yet.

The working master is where pixel work happens. Move a frame out of RAW only when the job demands it: layers, compositing, detailed skin retouching, or a handoff to another program or person. A 16-bit TIFF is the usual vehicle for that handoff because it carries high bit depth, an embedded color profile and metadata reliably between applications.

The delivery copy is the export, made once, at the end.

Four questions that decide the format

• Must it stay re-editable? Keep it as RAW, or as a raw DNG from the camera or Adobe DNG Converter, with its settings stored alongside or inside the file. Our explanation of XMP sidecar files covers how those small companion files preserve your edits without touching the original.
• Must it keep 16-bit precision? Stay in RAW for as long as you can, then use a 16-bit TIFF for pixel work and stay there until the final export.
• Do you need layers? Use a layered TIFF or PSD in your pixel editor, and keep that layered file as part of the master.
• Do you want the smallest lossless footprint? Stay in RAW as long as possible, then use compressed TIFF rather than uncompressed TIFF for the frames that need one.

Reserve TIFF masters for frames that earn them. Making TIFF the routine working format for every frame multiplies storage and slows backups and file copies.

One portrait, start to finish

You've shot a portrait in soft window light. In the RAW editor you set exposure, white balance and a gentle color grade, and you straighten the crop. The original hasn't changed, and every slider remains adjustable.

The skin needs careful retouching, so the picture opens in a pixel editor as a 16-bit TIFF. You retouch on separate layers, keeping the untouched pixels underneath in case you overdo it. When the retouching is done, you save the layered TIFF as the master and make the final color decisions there. Only then do you export, once, as an 8-bit sRGB JPEG sized for the client gallery.

Mistakes that quietly cost quality

• Letting a lossy file become the working file: Every save of a JPEG, and every tweak-and-re-export of a finished file, discards a little more detail even at high quality settings. Keep the working file lossless and generate the delivery copy once.
• Flattening and deleting layers: You lose the ability to soften a retouch or reopen a composite you later regret.
• Choosing the format by habit: The right file depends on the role it plays in the job, not on what you used last time.

Which Format Should You Export for Web, Print and Sharing?

Each destination asks for a different file. A web gallery wants small files that load quickly, a print lab wants the tonal data it asked for, and a social platform will recompress whatever it receives.

Export format by destination:

Where lossless WebP fits

PNG has long been the safe exact-pixel choice for the web, but it isn't the smallest. Google's WebP lossless bitstream specification describes a format that stores and restores pixel values exactly, transparency included, and reports compression about 25% denser than PNG. It gets there with transforms such as spatial prediction, color decorrelation and color indexing before entropy coding, plus LZ77-style back references and a color cache.

That gain is measured against PNG, so it matters for the images PNG is used for: logos, screenshots, diagrams and anything with transparency. For a photograph headed to a screen, keeping every pixel buys nothing you can see, and lossy WebP or JPEG makes a far smaller file. Chrome, Firefox, Edge and Safari all display WebP in their current versions, but a client's older system or a print portal may still refuse it.

The order of operations

A sensible default is to work from the master in one pass: convert the color space, resize, then compress last. Some workflows resize before converting; what matters more is that each step runs once and compression comes at the end. Repeating a step on an already exported file invites softness and color shifts.

When the gallery asks for a different size next month, go back to the master and export fresh rather than compressing a copy of a copy. The master works like a negative, and every file you send out is a print made from it. The guide on how to resize an image without losing quality covers that step in detail.

One finished image often needs two exports. For the online gallery, export a lossy WebP or JPEG sized for screens. For a framed print, export at full resolution in the format and color space your lab requests, often a 16-bit TIFF, and confirm the profile before sending. When a TIFF is too large for an email attachment, LinkShip, which sells link-based file-sharing tools, has a guide on how to deliver image files through browser links instead; before you rely on any such service, check that the recipient can download your original file and not only a compressed preview.

Social platforms recompress every upload, so a huge lossless file gains you nothing there. Send a clean 8-bit sRGB JPEG at a sensible size, and let their system do its worst to a copy rather than to your master.

How Should You Archive Originals and Masters?

A format policy can fit on an index card. Its job is to make sure you can always find the untouched original, the editable master and a clean delivery copy for any picture that matters.

File names and folders that keep each role clear

The policy only holds when the file name and folder tell you a file's role without opening it. A workable scheme uses one folder per shoot, with subfolders numbered in the order files move through them:

2026-03-14_lisbon/
  1-originals/
    IMG_4521.CR3
  2-masters/
    IMG_4521_v1_16bit.tif
  3-delivery/
    IMG_4521_web_2048px.jpg
    IMG_4521_print.tif

Set the originals folder to read-only after import, embed the color profile in each master, and export delivery files per destination. When an edit changes, bump the version tag on the master from v1 to v2 and re-export the delivery files, rather than renaming things in place. A glance at the path then tells you what is safe to recompress and what must survive.

Libraries publish their own baselines. The U.S. Library of Congress recommended formats statement, which sets out the formats the Library prefers for collecting and preserving creative works, asks for digital photographs at the highest resolution available without rescaling, the highest bit depth available (16 bits per channel if available), an embedded color profile or specified color space, and uncompressed, unlayered files. Its preferred formats are TIFF, JPEG 2000, PNG and JPEG; lossless compression, DNG and camera RAW formats appear on its acceptable list.

Choosing the master format

For the master tier, keep 16 bits per channel and an embedded profile; reserve 8-bit for delivery files headed to destinations that ask for it.

TIFF is generally the safer interchange choice for masters because it handles high bit depth, embedded profiles, layers and photographic metadata across professional tools. PNG is excellent for exact-pixel delivery and transparency, but it has no CMYK mode and apps handle camera metadata in PNG inconsistently. Our guide to whether PNG is a lossless format covers where it can still lose quality.

Practical housekeeping rules

Master only the frames you've selected. An 800-frame shoot doesn't need 800 TIFF masters. It needs the originals kept safely, and masters for the picks you'll print, sell or revisit. That one habit keeps archives affordable.

Keep EXIF, GPS and copyright fields in originals and masters, since they document when, where and how a picture was made and who owns it. Strip GPS from public web and social copies, because your home address or a client's location shouldn't travel with a shared JPEG.

Write exports to a new folder and never overwrite the source file. Some editors can also export a linear DNG: a full-resolution 16-bit file whose pixels are already demosaiced, with your edits baked in. Another raw editor can open it and keep adjusting, but it starts from that baked result and can't return to the sensor data the way your RAW and its settings can, so keep the RAW as well. The guide to what a DNG file is explains the format itself.

Back up on the 3-2-1 pattern: at least three copies, on two different kinds of media, with one copy off site. A drive failure, a stolen laptop or a house fire should each cost you time, never your archive.

Your five-step format policy

1. Keep camera RAW files untouched and backed up before formatting cards.
2. Edit non-destructively in a RAW editor so adjustments stay reversible.
3. Export a 16-bit TIFF master only for frames needing pixel work, compositing or print.
4. Deliver from the master: lossy JPEG or WebP for photos, PNG or lossless WebP for graphics, sized and profiled for each destination.
5. Test one frame end to end on a second device, checking color, sharpness and metadata before you trust the pipeline with a whole shoot.

Whatever editor you use, confirm that its TIFF export keeps the bit depth, profile and metadata you expect before relying on it for client work. A ten-minute test with one frame beats discovering a missing profile after a deadline export.

How Do You Run This Workflow in RevelRaw?

We make RevelRaw, a native RAW editor for Macs with macOS 26 or later on Apple M1 or later. It opens 24 RAW extensions across 15 brands and exports JPEG, PNG, 8- or 16-bit TIFF and HEIF in sRGB, Display P3 or Adobe RGB, so a 16-bit TIFF master and a delivery JPEG can come from the same RAW. It also exports a full-resolution 16-bit linear DNG with your edits baked in, which other raw editors can open and keep adjusting from that baked result. AI scene detection ranks 50+ looks by category as a starting point, and every slider stays yours to adjust.

Editing is free and the first full-quality export is included, with no account. Pro adds unlimited and batch export, as an auto-renewing subscription or a one-time lifetime purchase; RevelRaw Pro, Monthly ($3.99/month) is one of those plans, and App Store prices vary by region. Processing happens on your Mac; downloading the app, buying Pro and the one-time Apple selection-model download for click-to-select masks need internet. Get RevelRaw on the Mac App Store (requires macOS 26 or later).

Frequently Asked Questions

Which photo formats are lossless?

PNG, TIFF saved with LZW, ZIP or no compression, lossless WebP, and JPEG XL or JPEG 2000 in lossless mode all decode to exactly the pixels you saved. JPEG is always lossy and HEIF is usually lossy. RAW and DNG are lossless only when saved uncompressed or with lossless compression.

Is a RAW file always lossless?

No. Many cameras save RAW uncompressed or with lossless compression, but many also offer lossy options that discard some data to shrink the file, such as Sony's compressed ARW, Canon's C-RAW and Nikon's lossy compressed NEF. Lossy RAW still keeps white balance and tone open for editing. Check the RAW compression setting in your camera's menu.

Does saving as DNG keep my photo re-editable?

It depends on the DNG. A raw DNG from the camera or Adobe DNG Converter still holds sensor data, and edits live as settings you can change. A linear DNG exported from an editor holds demosaiced pixels with your edits baked in: another raw editor can open it and adjust further, but only from that baked result.

Should I use lossless WebP for a photo gallery?

Usually not. A lossless WebP or PNG of a photograph is several times larger than a lossy WebP or a JPEG at quality around 80–90, and on screen the difference isn't visible. Use lossless WebP or PNG for graphics, screenshots and images with transparency, and keep the lossless master in your archive.

Which lossless formats can RevelRaw export?

RevelRaw, our Mac app, exports lossless PNG, 8- or 16-bit TIFF and a 16-bit linear DNG with your edits baked in, plus JPEG and HEIF for delivery. The sRGB, Display P3 and Adobe RGB choice applies to every format except DNG. It is a free download with free editing and one free full-quality export, and it needs macOS 26 or later on an Apple M1 or later Mac.

Visit RevelRaw to open a shoot on your Mac, where editing is free. Use the first full-quality export, also free, to make a 16-bit TIFF master or a delivery copy and check it at its real destination before you set your own format policy.