Repository navigation
Releases: kikoz85/audioSTAX
Release list
AudioSTAX 0.2.0
AudioSTAX 0.2.0 — tags, cover art and file names for the whole batch.
▶ One-minute video of every screen / Video di un minuto con tutte le schermate: audiostax-tour-en.mp4 · audiostax-tour-it.mp4 (below / qui sotto).
English
New
- Tags for every file at once: type artist, album, album artist, year, genre or comment once and they go into every file of the batch; empty fields keep each file's own tags. The grey hints show the selected file's current values.
- Title from the file name and track numbers in queue order (1/12, 2/12…).
- Cover art: keep the original, replace it with a JPEG or PNG of your choice, or remove it — in MP3, FLAC and AAC (.m4a). It also survives AI noise reduction.
- Rename rules: find and replace (e.g.
_draft→ nothing,_→ space), letter case (abc, ABC, Abc Abc) and new variables{n},{albumartist},{genre},{year}, with zero padding ({n:03}→ 007,{track:02}→ 03). Tags you edit feed the names too. - Output names preview: the real names of the files waiting to be converted, with a warning for files that already exist and for duplicate names.
- New "Tags and cover" chapter in the built-in guide (F1), in all five languages.
Fixed
- 0.1.0 did not copy the cover art of the original file.
Download
| System | File |
|---|---|
| macOS 12+ (Apple Silicon) | AudioSTAX-v0.2.0-macOS.dmg |
| Windows 10/11 x64 | AudioSTAX-v0.2.0-Windows-x64-Setup.exe (per-user installer) or …-Windows-x64.zip (portable) |
| Linux x86_64 | AudioSTAX-v0.2.0-Linux-x86_64.AppImage, audiostax_0.2.0_amd64.deb (Debian/Ubuntu), audiostax-0.2.0-1.x86_64.rpm (Fedora/openSUSE) or …-Linux-x86_64.tar.gz |
| Windows 11 on ARM (Snapdragon, Parallels on Mac) — new | AudioSTAX-v0.2.0-Windows-arm64-Setup.exe or …-Windows-arm64.zip |
| Linux ARM64 (Raspberry Pi 4/5, ARM laptops) — new | AudioSTAX-v0.2.0-Linux-aarch64.AppImage, audiostax_0.2.0_arm64.deb, audiostax-0.2.0-1.aarch64.rpm or …-Linux-aarch64.tar.gz (Ubuntu 22.04, Debian 12, Raspberry Pi OS Bookworm 64-bit, Fedora 35 or later) |
Checksums: SHA256SUMS-macos, SHA256SUMS-windows, SHA256SUMS-linux, SHA256SUMS-windows-arm64, SHA256SUMS-linux-arm64, SHA256SUMS-sources. Updating from 0.1.0: install over it, your settings and AI models are kept.
First launch
- macOS — not yet signed with an Apple certificate: right-click AudioSTAX in Applications → Open → Open (or System Settings → Privacy & Security → Open Anyway).
- Windows — if SmartScreen appears: More info → Run anyway.
- Linux —
chmod +x AudioSTAX-*.AppImage; without FUSE start it with--appimage-extract-and-run. Orsudo apt install ./audiostax_0.2.0_amd64.deb/sudo dnf install ./audiostax-0.2.0-1.x86_64.rpm.
Italiano
Novità
- Tag per tutti i file in un colpo: scrivi una volta artista, album, artista album, anno, genere o commento e finiscono in tutti i file del batch; i campi vuoti lasciano i tag di ogni file. In grigio vedi i valori attuali del file selezionato.
- Titolo dal nome del file e numeri di traccia nell'ordine della coda (1/12, 2/12…).
- Copertina: mantieni quella originale, sostituiscila con un JPEG o PNG a tua scelta oppure rimuovila — in MP3, FLAC e AAC (.m4a). Resta anche dopo la riduzione del rumore AI.
- Regole di rinomina: trova e sostituisci (es.
_bozza→ niente,_→ spazio), maiuscole/minuscole (abc, ABC, Abc Abc) e nuove variabili{n},{albumartist},{genre},{year}, con zeri davanti ({n:03}→ 007,{track:02}→ 03). I tag che modifichi valgono anche nei nomi. - Anteprima dei nomi in uscita: i nomi reali dei file da convertire, con un avviso per i file già presenti e per i nomi doppi.
- Nuovo capitolo "Tag e copertina" nella guida integrata (F1), in tutte e cinque le lingue.
Corretto
- La 0.1.0 non copiava la copertina del file originale.
Download
| Sistema | File |
|---|---|
| macOS 12+ (Apple Silicon) | AudioSTAX-v0.2.0-macOS.dmg |
| Windows 10/11 x64 | AudioSTAX-v0.2.0-Windows-x64-Setup.exe (installer per utente) oppure …-Windows-x64.zip (portabile) |
| Linux x86_64 | AudioSTAX-v0.2.0-Linux-x86_64.AppImage, audiostax_0.2.0_amd64.deb (Debian/Ubuntu), audiostax-0.2.0-1.x86_64.rpm (Fedora/openSUSE) oppure …-Linux-x86_64.tar.gz |
| Windows 11 su ARM (Snapdragon, Parallels su Mac) — novità | AudioSTAX-v0.2.0-Windows-arm64-Setup.exe oppure …-Windows-arm64.zip |
| Linux ARM64 (Raspberry Pi 4/5, portatili ARM) — novità | AudioSTAX-v0.2.0-Linux-aarch64.AppImage, audiostax_0.2.0_arm64.deb, audiostax-0.2.0-1.aarch64.rpm oppure …-Linux-aarch64.tar.gz (Ubuntu 22.04, Debian 12, Raspberry Pi OS Bookworm 64-bit, Fedora 35 o successivi) |
Checksum: SHA256SUMS-macos, SHA256SUMS-windows, SHA256SUMS-linux, SHA256SUMS-windows-arm64, SHA256SUMS-linux-arm64, SHA256SUMS-sources. Aggiornare dalla 0.1.0: installa sopra, impostazioni e modelli AI restano.
Primo avvio
- macOS — non ancora firmato con un certificato Apple: clic destro su AudioSTAX in Applicazioni → Apri → Apri (oppure Impostazioni di Sistema → Privacy e sicurezza → Apri comunque).
- Windows — se compare SmartScreen: Ulteriori informazioni → Esegui comunque.
- Linux —
chmod +x AudioSTAX-*.AppImage; senza FUSE avviala con--appimage-extract-and-run. Oppuresudo apt install ./audiostax_0.2.0_amd64.deb/sudo dnf install ./audiostax-0.2.0-1.x86_64.rpm.
Third-party sources (LGPL) · Sorgenti di terze parti (LGPL)
AudioSTAX includes FFmpeg as a separate program (bin/ffmpeg, bin/ffprobe), audio only (no video decoders), built without GPL or non-free parts. AudioSTAX-v0.2.0-FFmpeg-sources.tar.gz below holds the complete corresponding source: the unmodified official archives, their SHA-256, the build script and its exact configuration. · Il file AudioSTAX-v0.2.0-FFmpeg-sources.tar.gz qui sotto contiene i sorgenti completi: archivi ufficiali non modificati, i loro SHA-256, lo script di compilazione e la configurazione esatta. Upstream:
FFmpeg 7.1.2 (https://ffmpeg.org/releases/ffmpeg-7.1.2.tar.xz) · LAME 3.100 (https://downloads.sourceforge.net/project/lame/lame/3.100/lame-3.100.tar.gz) · Opus 1.5.2 · libogg 1.3.5 · libvorbis 1.3.7 (https://downloads.xiph.org/releases/) · zlib 1.3.1 on Windows (https://zlib.net/fossils/zlib-1.3.1.tar.gz). The Windows ARM64 build also ships LLVM's libc++.dll and libunwind.dll (Apache-2.0 WITH LLVM-exception). All licenses: THIRD_PARTY_LICENSES.txt in the app and in this repository.
❤ AudioSTAX is free: if it saves you time, support it on GitHub Sponsors. · © 2026 Enrico Fanucchi · www.enricofanucchi.com
AudioSTAX 0.1.0
AudioSTAX 0.1.0 — the first release. Batch audio studio: convert, normalize to LUFS, clean up speech and write subtitles for hundreds of files at once. Offline, free, no subscriptions.
English
What's in it
- Batch conversion to WAV (16/24/32-bit float), FLAC, MP3 (CBR/VBR), AAC (.m4a), Opus and OGG Vorbis, from almost any audio format and from the soundtrack of MP4, MOV, MKV, WebM and AVI videos.
- Two-pass LUFS normalization (EBU R128): podcast −16, streaming −14, broadcast −23 or your own targets, with a plain gain that keeps the dynamics and a true-peak ceiling.
- Speech cleanup for whole batches: AI noise reduction (DeepFilterNet) or a light mode for faint hiss, automatic 50/60 Hz mains hum removal, and a leveler that evens out quiet and loud voices.
- Subtitles and transcripts with Whisper (Tiny → Large Turbo, language detected automatically): SRT, VTT, TXT and JSON next to every converted file.
- Apple AudioToolbox AAC on macOS.
- Waveform and measured loudness of every file, parallel jobs, pause/resume, Dock/taskbar progress, file names from tags, a complete built-in guide (F1), five languages.
AI models are downloaded only when you first need them (8 MB for noise reduction, 32–574 MB for subtitles), after their license is shown; nothing else ever goes online and no audio leaves your computer.
Download
| System | File |
|---|---|
| macOS 12+ (Apple Silicon) | AudioSTAX-v0.1.0-macOS.dmg |
| Windows 10/11 x64 | AudioSTAX-v0.1.0-Windows-x64-Setup.exe (per-user installer) or …-Windows-x64.zip (portable) |
| Linux x86_64 | AudioSTAX-v0.1.0-Linux-x86_64.AppImage, audiostax_0.1.0_amd64.deb (Debian/Ubuntu), audiostax-0.1.0-1.x86_64.rpm (Fedora/openSUSE) or …-Linux-x86_64.tar.gz |
Checksums: SHA256SUMS-macos, SHA256SUMS-windows, SHA256SUMS-linux.
Subtitles on Windows and Linux need a processor with AVX2 (every PC since about 2013).
First launch
- macOS — not yet signed with an Apple certificate: right-click AudioSTAX in Applications → Open → Open (or System Settings → Privacy & Security → Open Anyway).
- Windows — if SmartScreen appears: More info → Run anyway.
- Linux —
chmod +x AudioSTAX-*.AppImage; without FUSE start it with--appimage-extract-and-run. Orsudo apt install ./audiostax_0.1.0_amd64.deb/sudo dnf install ./audiostax-0.1.0-1.x86_64.rpm.
Italiano
Cosa contiene
- Conversione batch in WAV (16/24/32-bit float), FLAC, MP3 (CBR/VBR), AAC (.m4a), Opus e OGG Vorbis, da quasi ogni formato audio e dalla colonna sonora di video MP4, MOV, MKV, WebM e AVI.
- Normalizzazione LUFS a due passaggi (EBU R128): podcast −16, streaming −14, broadcast −23 o valori tuoi, con un guadagno semplice che conserva la dinamica e il tetto true peak.
- Pulizia della voce su interi batch: riduzione del rumore con AI (DeepFilterNet) o modalità leggera per il fruscio, rimozione automatica del ronzio di rete a 50/60 Hz e uniforma voci.
- Sottotitoli e trascrizioni con Whisper (da Tiny a Large Turbo, lingua riconosciuta in automatico): SRT, VTT, TXT e JSON accanto a ogni file convertito.
- AAC Apple AudioToolbox su macOS.
- Forma d'onda e loudness misurata di ogni file, job paralleli, pausa e ripresa, avanzamento nel Dock e nella barra delle applicazioni, nomi dei file dai tag, guida completa integrata (F1), cinque lingue.
I modelli AI si scaricano solo la prima volta che servono (8 MB per la riduzione del rumore, da 32 a 574 MB per i sottotitoli), dopo averne mostrato la licenza; per il resto niente va online e nessun audio lascia il computer.
Download
| Sistema | File |
|---|---|
| macOS 12+ (Apple Silicon) | AudioSTAX-v0.1.0-macOS.dmg |
| Windows 10/11 x64 | AudioSTAX-v0.1.0-Windows-x64-Setup.exe (installer per utente) oppure …-Windows-x64.zip (portabile) |
| Linux x86_64 | AudioSTAX-v0.1.0-Linux-x86_64.AppImage, audiostax_0.1.0_amd64.deb (Debian/Ubuntu), audiostax-0.1.0-1.x86_64.rpm (Fedora/openSUSE) oppure …-Linux-x86_64.tar.gz |
Checksum: SHA256SUMS-macos, SHA256SUMS-windows, SHA256SUMS-linux.
I sottotitoli su Windows e Linux richiedono un processore con AVX2 (tutti i PC dal 2013 circa).
Primo avvio
- macOS — non ancora firmato con un certificato Apple: clic destro su AudioSTAX in Applicazioni → Apri → Apri (oppure Impostazioni di Sistema → Privacy e sicurezza → Apri comunque).
- Windows — se compare SmartScreen: Ulteriori informazioni → Esegui comunque.
- Linux —
chmod +x AudioSTAX-*.AppImage; senza FUSE avviala con--appimage-extract-and-run. Oppuresudo apt install ./audiostax_0.1.0_amd64.deb/sudo dnf install ./audiostax-0.1.0-1.x86_64.rpm.
Third-party sources (LGPL) · Sorgenti di terze parti (LGPL)
AudioSTAX includes FFmpeg as a separate program (bin/ffmpeg, bin/ffprobe), built without GPL or non-free parts from these exact sources:
FFmpeg 7.1.2 (https://ffmpeg.org/releases/ffmpeg-7.1.2.tar.xz) · LAME 3.100 (https://downloads.sourceforge.net/project/lame/lame/3.100/lame-3.100.tar.gz) · Opus 1.5.2 · libogg 1.3.5 · libvorbis 1.3.7 (https://downloads.xiph.org/releases/). All licenses: THIRD_PARTY_LICENSES.txt in the app and in this repository.
❤ AudioSTAX is free: if it saves you time, support it on GitHub Sponsors. · © 2026 Enrico Fanucchi · www.enricofanucchi.com