General

Digital Preservation: File Formats That Stand the Test of Time

Published Mar 19, 2026 7 min read By ChangeThisFile Team

Quick Answer

For files that must remain accessible decades from now, choose formats with open specifications, wide tool support, and institutional adoption: PDF/A for documents, TIFF for images, WAV or FLAC for audio, MKV with FFV1 for video, and plain text for anything that doesn't need formatting. Proprietary formats die when the company does.

Every file you create today is a bet on a format's future. If you save a document as DOCX, you're betting that software capable of reading DOCX will exist when you need it in 2046. If you archive photos as PSD, you're betting that Adobe Photoshop or a compatible reader will be available. If you store video as WMV, you've already lost that bet.

Digital preservation is the discipline of choosing formats, storage media, and practices that maximize the likelihood of files remaining accessible over decades. Libraries, archives, museums, and government agencies have been grappling with this since the 1990s, and their recommendations are directly applicable to anyone with files they care about keeping.

This guide covers which formats have the best preservation characteristics, why open specifications matter more than technical features, and the practical strategies for ensuring your files survive.

What Makes a Format Archival

The Library of Congress (LoC), the National Archives (NARA), and the Digital Preservation Coalition evaluate formats on seven criteria:

Criterion	Why It Matters
Open specification	If the spec is public, anyone can build a reader. No dependency on one company.
Non-proprietary	Formats controlled by one company die when the company does (or pivots).
Wide adoption	Popular formats have more tools, more expertise, and more institutional investment in their survival.
Self-documenting	Formats that describe their own structure (headers, metadata) are easier to recover from partial corruption.
No DRM	DRM-encumbered files may become inaccessible when the license server disappears.
Lossless option	Lossy formats permanently discard information. For preservation, lossless is strongly preferred.
Backward compatibility	New versions of the format should be able to read old versions. PNG from 1996 works identically in 2026.

Documents: PDF/A

PDF/A (ISO 19005, first published 2005) is the only document format specifically designed for long-term archival. It's a constrained subset of PDF that removes features that could break future accessibility:

No JavaScript (scripts may not execute in future readers)
No external dependencies (all fonts must be embedded; no linked images or resources)
No encryption (encrypted files become inaccessible if the password is lost)
Required metadata in XMP format
Required embedded ICC color profiles

PDF/A has four conformance levels: PDF/A-1 (2005, based on PDF 1.4), PDF/A-2 (2011, based on PDF 1.7, adds JPEG 2000 and transparency), PDF/A-3 (2012, allows embedded file attachments), and PDF/A-4 (2020, based on PDF 2.0).

PDF/A is mandated for legal documents, government records, and court filings in many jurisdictions. The EU's eIDAS regulation references PDF/A. U.S. federal courts require PDF/A for electronic filing.

Convert DOCX to PDF for archival — though for true PDF/A conformance, use a tool that specifically targets PDF/A output (LibreOffice can export as PDF/A-1 and PDF/A-2). Standard PDF conversion may not meet all PDF/A requirements.

The Unkillable Format: Plain Text

Plain text (UTF-8 encoded) is the most durable file format in existence. ASCII text from 1963 is still perfectly readable. Text files have no dependencies — no fonts, no rendering engines, no codecs. They're readable by literally every computing device ever made.

For any content that doesn't require formatting — source code, configuration files, notes, data records, logs — plain text is the archival gold standard. Convert documents to TXT when the content matters more than the formatting. Markdown (DOCX to Markdown) adds minimal formatting while remaining human-readable text.

Images: TIFF and PNG

TIFF (Tag Image File Format, Adobe/Aldus, 1986) is the Library of Congress's recommended format for raster image preservation. It supports lossless compression (LZW, ZIP, or uncompressed), 16-bit per channel color, ICC color profiles, and multi-page documents. TIFF's tagged chunk structure is self-documenting — even partial or damaged files can often be partially recovered.

TIFF is universally supported by professional image software and has been stable for decades. The format hasn't had a major revision since TIFF 6.0 in 1992 — this stability is a feature, not a limitation. A TIFF created in 1992 opens identically in 2026.

PNG (W3C, 1996) is the web-friendly archival alternative. It's lossless, open, widely supported, and smaller than uncompressed TIFF (though larger than LZW-compressed TIFF for photographic content). PNG is a strong choice when TIFF is overkill or when files need to be web-accessible.

For archival: convert JPEG to TIFF (note: this won't restore quality lost to JPEG compression — it just preserves what remains in a lossless container). Convert BMP to PNG for lossless size reduction. Convert PSD to TIFF to escape Photoshop's proprietary format while preserving quality.

What About Camera RAW?

Camera RAW files (CR2, NEF, ARW, DNG) contain the sensor's unprocessed data and are the highest-quality source available. But CR2 (Canon), NEF (Nikon), and ARW (Sony) are proprietary formats that depend on each manufacturer's continued support.

DNG (Digital Negative, Adobe, 2004) is an open RAW format designed for archival. Adobe published the specification and committed to it as a preservation format. The Library of Congress lists DNG as a preferred format for digital photo archival. Convert proprietary RAW to DNG using Adobe's free DNG Converter, or convert to TIFF for maximum safety: CR2 to TIFF | NEF to TIFF | ARW to TIFF.

Audio: WAV and FLAC

WAV (RIFF WAVE, Microsoft/IBM, 1991) stores uncompressed PCM audio. It's the simplest, most universally supported audio format. CD-quality WAV (44.1 kHz, 16-bit stereo) is 10.1 MB per minute. High-resolution WAV (96 kHz, 24-bit) is 34.6 MB per minute. WAV's simplicity is its preservation strength — the format is trivial to parse and has no codec dependencies.

FLAC (Free Lossless Audio Codec, Josh Coalson, 2001) compresses audio to 50-60% of WAV size with bit-perfect reconstruction. FLAC is open-source, royalty-free, and widely supported. The Library of Congress accepts FLAC as a preservation format alongside WAV.

The choice between WAV and FLAC for archival: WAV is simpler and needs no decompression step (lower future risk), FLAC is 50% smaller with the same content (saves storage). Both are excellent choices. Neither MP3 nor AAC should be used for preservation — they're lossy and discard data permanently.

Convert MP3 to WAV for archival (preserves the already-degraded audio in a lossless container). Convert WAV to FLAC to save storage without quality loss.

Video: MKV with FFV1, or Uncompressed

Video preservation is the hardest problem because of the sheer data volume. One hour of uncompressed 1080p video is approximately 560 GB. Even institutions with large budgets can't store everything uncompressed.

FFV1 (FF Video Codec 1) is a lossless video codec developed by Michael Niedermayer for FFmpeg and standardized by IETF. The Library of Congress and the International Association of Sound and Audiovisual Archives (IASA) recommend FFV1 in MKV containers for video preservation. FFV1 achieves 2:1 to 3:1 compression losslessly — that 560 GB hour becomes 200-280 GB.

MKV (Matroska) is the recommended container because it's open, well-documented, and supports any codec plus multiple audio tracks, subtitle tracks, and chapters. The combination of MKV + FFV1 + FLAC audio is the current institutional standard for video preservation.

For practical archival (personal collections, small organizations), H.264 in MP4 at high bitrate (CRF 18 or lower) is a reasonable compromise — not lossless, but visually transparent and 10-20x smaller than lossless. Convert legacy AVI to MKV for a modern, preservation-friendly container.

Ebooks: EPUB and PDF

EPUB (International Digital Publishing Forum, 2007) is an open ebook format based on XHTML and CSS. It's reflowable (adapts to screen size), supports embedded fonts and images, and is widely supported by readers. EPUB is the recommended format for text-centric ebooks.

PDF is better for fixed-layout content (textbooks with complex layouts, illustrated books, technical manuals). For archival, use PDF/A to ensure long-term accessibility.

Avoid proprietary ebook formats for archival: Kindle's KFX, Apple's iBooks format, and any DRM-protected format. DRM-protected files become inaccessible when the license server disappears — and it will eventually disappear. Convert EPUB to PDF for fixed-layout preservation, or keep both formats for flexibility.

Formats to Avoid for Preservation

Format	Risk	Migrate To
PSD (Photoshop)	Proprietary, depends on Adobe	TIFF or PNG
DOC (Word 97-2003)	Binary, underdocumented, declining support	DOCX then PDF/A
WMA/WMV	Proprietary Microsoft, declining support	FLAC / MP4
HEIC	Patent-encumbered (HEVC), uncertain licensing future	PNG or TIFF
RAR	Proprietary (Alexander Roshal), closed source	ZIP or 7Z
MOBI/AZW3	Amazon proprietary, DRM possible	EPUB
Any DRM-protected format	Inaccessible when license server dies	Remove DRM (where legal) and save as open format

A Practical Preservation Strategy

Keep originals. Never delete the camera RAW, the master recording, or the source document. Store them alongside any converted versions.
Convert to archival formats. Create archival copies in open, lossless formats: TIFF/PNG for images, WAV/FLAC for audio, PDF/A for documents, EPUB for ebooks.
Use the 3-2-1 backup rule. Three copies, on two different types of storage media, with one copy offsite. Format durability means nothing if all copies are on one hard drive that fails.
Embed metadata. Add title, date, description, and creator to every file. Metadata is the only context that travels with the file itself.
Revisit every 5-10 years. Check that your files are still readable with current tools. Migrate to newer open formats if the current one is losing support. This is cheaper than emergency recovery after a format becomes truly obsolete.
Document your file organization. A plain text README in each archive directory explaining what's in it and how it's organized costs nothing and helps immensely.

Digital preservation comes down to a simple principle: use formats that don't require any specific company to exist. PDF/A will be readable as long as the ISO exists. TIFF has been stable since 1992 and will be parseable for decades more. FLAC's specification is public and implementations are open-source. Plain text will outlast civilization.

The practical cost of preservation is small: convert important files to archival formats, keep the originals alongside them, and check every few years that everything still opens. The cost of not doing this — realizing that your 20-year photo archive is in a format no current tool can read — is enormous and irreversible.

Key Takeaways

Open specification is the single most important criterion for archival formats — proprietary formats die with their companies
PDF/A (ISO 19005) is the only document format specifically designed for long-term preservation
TIFF is the Library of Congress standard for image preservation; PNG is the web-friendly alternative
WAV and FLAC are both excellent for audio preservation — WAV is simpler, FLAC is 50% smaller
MKV with FFV1 video and FLAC audio is the institutional standard for video preservation
DRM-protected content is the highest obsolescence risk — it becomes inaccessible when the license server dies
Plain text (UTF-8) is the most durable format in existence and will outlast everything else

Frequently Asked Questions

Will JPEG files still be readable in 50 years?

Almost certainly yes. JPEG is the most widely supported image format in history — the specification is public (ITU T.81), hardware and software decoders number in the billions, and every platform supports it. The format hasn't changed since 1992 and there's no reason decoders would stop being available. JPEG is safe for long-term storage, though it's lossy, so if you have the original lossless source, keep that too.

Is DOCX safe for long-term storage?

Reasonably safe but not ideal. DOCX is ISO-standardized (ISO/IEC 29500), openly documented, and supported by multiple applications (Microsoft Office, LibreOffice, Google Docs). But complex documents may not render identically across implementations, and some features depend on Microsoft-specific extensions. For archival, convert to PDF/A for fixed-layout preservation, or keep both DOCX (editable) and PDF/A (frozen) versions.

Should I convert my JPEG photos to TIFF for archival?

Only for organizational purposes, not for quality improvement. Converting JPEG to TIFF doesn't restore quality lost during JPEG compression — it just wraps the already-degraded image in a lossless container. The TIFF file will be larger with identical visual quality. If you have the original RAW or uncompressed source, convert from that. If JPEG is all you have, keeping it as JPEG is fine for preservation.

What happens to my DRM-protected ebooks if the store closes?

They become inaccessible. When a DRM license server goes offline, the content it protected can no longer be authorized for reading. This has already happened: Microsoft shut down its ebook store in 2019, and customers lost access to all their purchases. The same risk applies to any DRM-protected content from any vendor. For preservation, DRM-free formats (EPUB without DRM, PDF without DRM) are essential.

How long does a hard drive last?

Consumer hard drives have a median lifespan of 3-5 years under continuous use, though many last 10+ years. SSDs last longer under light use but can lose data if left unpowered for extended periods (NAND flash cells lose charge over months to years). For preservation, don't rely on any single storage medium. Use the 3-2-1 rule: three copies, two media types, one offsite. Cloud storage with geographic redundancy is the most practical long-term solution for most people.

Is MP4 with H.264 safe for video archival?

For personal and small-organization archival, yes. H.264 is universally supported, MP4 is ISO-standardized, and both will be readable for decades. The compression is lossy, so archival institutions prefer lossless codecs (FFV1), but H.264 at high quality (CRF 18 or lower) is visually transparent and far more practical in terms of storage. For a personal video archive, H.264 MP4 is a good choice.

What is PRONOM and why does it matter?

PRONOM is a technical registry of file formats maintained by The National Archives of the UK. It assigns unique persistent identifiers (PUIDs) to every known file format and version, documenting their technical properties, risks, and relationships. Digital preservation tools like DROID use PRONOM to identify and validate file formats. If you're doing institutional preservation, PRONOM is the reference for assessing format risk.

Should I keep files in their original format or convert everything?

Keep the originals AND create archival copies. Never delete a source file because you converted it — the conversion might have lost metadata, reduced quality, or introduced errors. The ideal approach: store the original file as-is, plus a copy in an archival format (TIFF, FLAC, PDF/A). The original preserves everything the creator intended. The archival copy ensures readability if the original format becomes obsolete.

Ready to convert your files?

Use ChangeThisFile to convert between 600+ formats — free, fast, and private.

Start Converting