Initializing... drag & drop files here
Supports: FILE
.tz2 — the typical Linux-distribution layout where one TAR stream is compressed with bzip2 — or Individual Archives to produce a separate .tz2 per input file when you want each artifact independently downloadable and extractable..tz2 by default. If you'd prefer .tar.bz2, .tbz2, or another bzip2-compatible extension for downstream tooling, the TZ2 to TAR.BZ2 and TZ2 to TBZ converters re-emit the same bytes under a different name with no recompression cost..tz2 directly, or grab a ZIP bundle if you produced multiple individual archives.TZ2 — also written .tar.bz2, .tbz, or .tbz2 — is a TAR archive run through bzip2 compression. bzip2 was first released by Julian Seward in 1996 and combines run-length encoding, the Burrows–Wheeler transform, move-to-front, and Huffman coding to hit compression ratios noticeably tighter than gzip on text-heavy data. It is one of the three classic Linux/Unix archive formats alongside .tar.gz and .tar.xz, and remains the default download for plenty of long-running open-source projects. Typical reasons to package as TZ2:
.tar.bz2 is still the canonical source-tarball extension for many GNU and long-running Unix projects (autoconf, bash, glibc, the older Apache HTTPD source releases). Anyone on Linux or macOS can extract it with the built-in tar — no third-party tool required..tz2 keeps file modes (executable bit), symlinks, ownership, and modification times intact in a way that ZIP often does not..bz2 over .gz for archival columnar / log data.| Property | TZ2 (.tar.bz2) |
TGZ (.tar.gz) |
TAR.XZ | ZIP | 7Z |
|---|---|---|---|---|---|
| Compression algorithm | bzip2 (Burrows-Wheeler) | gzip (DEFLATE) | xz (LZMA2) | DEFLATE per-file | LZMA2 / others |
| Typical ratio (text) | ~17% | ~21% | ~14% | ~22% | ~13% |
| Compress speed | Slow | Fast | Slowest | Fast | Slow |
| Decompress speed | Moderate | Fastest | Fast | Fast | Moderate |
| Preserves Unix permissions | Yes (TAR layer) | Yes (TAR layer) | Yes (TAR layer) | Limited | Limited |
| Native on Linux/macOS | Yes | Yes | Yes | Via unzip |
Via p7zip |
| Native on Windows | No (7-Zip / WinRAR) | No (7-Zip / WinRAR) | No (7-Zip / WinRAR) | Yes | No |
| Built-in encryption | No | No | No | Yes (weak) | Yes (AES-256) |
| Random access to single file | No (stream) | No (stream) | No (stream) | Yes | Yes |
Ratio numbers above are approximate and based on the linux-3.18 kernel tarball benchmark (rootusers); your mileage will vary by content type — already-compressed files (JPEG, MP4, MP3) gain little or nothing from any of these.
bzip2 exposes block-size levels -1 through -9. Unlike gzip, the level mostly tunes block size and memory rather than dramatically changing speed. xconvert uses the bzip2 default (level 9, 900k blocks).
| Level | Block size | Typical ratio vs -9 |
Memory needed | Notes |
|---|---|---|---|---|
-1 |
100k | ~5% larger | ~7 MB | Use only on very memory-constrained targets |
-5 |
500k | ~1% larger | ~32 MB | Middle ground, rarely needed |
-9 (default) |
900k | baseline | ~7600 KB compress / ~3700 KB decompress | What virtually every distro tarball ships at |
All four extensions describe the same file — a TAR archive compressed with bzip2. .tar.bz2 is the explicit two-stage form. .tbz2 and .tbz are short forms preserved for systems with 3-character or 4-character filename limits. .tz2 is the rarest of the four and behaves identically. You can rename between them freely; the bytes are unchanged.
On Linux and macOS, run tar -xjf archive.tz2 from the terminal — both ship with tar and bzip2 by default. Modern GNU tar also auto-detects the format with plain tar -xf archive.tz2 (GNU tar manual). On Windows, use 7-Zip, WinRAR, PeaZip, or WinZip — Windows Explorer does not natively understand bzip2. On any platform, double-clicking usually triggers the installed archive manager.
It depends what you're optimizing for. bzip2 typically produces 10–20% smaller archives than gzip on text-heavy content, but compression takes 2–5× longer and decompression is also slower. Pick bzip2 when archive size and bandwidth dominate (releasing a source tarball, storing a backup once, decompressing rarely). Pick gzip when you compress and decompress constantly (HTTP transport, streaming logs, frequent CI artifacts).
xz (LZMA2) usually beats bzip2 on ratio by another 4–6 percentage points on the same input, but compress times are dramatically higher at high levels — xz at level 9 took ~500s on the kernel-source benchmark vs ~83s for bzip2 (rootusers). bzip2 also has a much smaller memory footprint. Choose xz when you want the absolute smallest archive (typical for modern Linux distro packages); choose bzip2 when you need a portable, well-supported format with reasonable compress time.
Single Archive packages every uploaded file into one .tz2 — the standard layout for distributing a project. Recipients extract once and get the entire tree. Individual Archives wraps each input file in its own .tz2, so 10 input files yield 10 output archives. Useful when you want each artifact independently downloadable on a release page, or when uploading per-file to systems that prefer one archive per item.
No. Neither TAR nor bzip2 has a built-in encryption layer — that's a known limitation of the format. If you need encryption, either compress with 7z or RAR (which support AES-256), or pipe the .tz2 through a separate tool such as GPG (gpg -c archive.tz2) before sharing. xconvert produces standard, unencrypted .tz2 archives.
You can, but the size savings will be small or zero. JPEG, PNG, MP3, MP4, AAC, and ZIP files are already entropy-coded; bzip2 has nothing left to remove. The TAR layer is still useful for bundling many files into one and preserving directory structure, but expect the output to be roughly the sum of the inputs plus a small overhead. TAR-only (without bzip2) would give the same bundling at higher speed.
bzip2's Burrows-Wheeler transform is asymmetric: it processes data in 100k–900k blocks, and each block must be reconstructed by inverting the BWT. That work is heavier than gzip's DEFLATE inflation. On the kernel-source benchmark, gzip decompresses at level 9 in ~6s vs ~24s for bzip2 (rootusers). For workloads where you decompress far more often than you compress, gzip is generally the better trade.
There's no hard cap from the format itself — bzip2 streams can be arbitrarily large, and the TAR layer supports multi-terabyte archives via the pax/USTAR extensions. In xconvert, the practical limit is your browser's available memory: very large archives (multi-GB) may stress mobile browsers. For a 100 GB cold-storage backup, run tar c | bzip2 locally; for typical source tarballs, project bundles, log dumps, or document sets, browser-side conversion is the easier path.
Yes — see TZ2 to ZIP for the cross-platform-friendly version, TZ2 to TAR.GZ when you want the same TAR contents but with faster decompress, TZ2 to TAR.XZ for tighter ratios, or Extract TZ2 to pull the raw files out without re-archiving.