jersou / studio-pack-generator

Convert a folder or a RSS URL to Studio pack zip for Lunii device
MIT License
80 stars 11 forks source link

menu audio files : problem with TTS generation on Windows (WSL instead Windows TTS) #14

Closed schmurtzm closed 1 year ago

schmurtzm commented 2 years ago

Hello,

when generating a pack on Windows 10 with studio-pack-generator v0.1.17 I see this command line : "C:\WINDOWS\system32\wsl.exe" pico2wave -l fr-FR -w "/mnt/g/Studio-Pack-Generator/test/Ourson.item.wav" " . Ourson . "

But theoretically wsl is not a required dependency : "Windows release use Windows TTS instead of picoTTS"

This is annoying to doesn't have audio on each title because without it Studio refuse to convert a pack.

Thank you !

schmurtzm commented 2 years ago

OK understood, it's not really a bug but instead a lack in the readme : studio-pack-generator requires wsl to be installed and libttspico-utils inside wsl. Cf this line in the code.

So yes, after having installed wsl and installed libttspico-utils on it it works as it should. May be the readme should be updated to clarify this point for Windows users ?

But this is a huge requirement to generate tts on Windows, I would recommend to add wsay.exe in the "tools" folder to generate mp3 easily. The French "Julie voice" available on Windows 10 is not too bad.

https://github.com/p-groarke/wsay

MrAlucardDante commented 1 year ago

I can confirm that the issue is real, although the project hasn't been updated in a year.

Maybe someone will fork it to fix that.

For those that aren't that technical to use wsay or that don't want to install a Linux distro just to generate TTS, you can use one of these :

jersou commented 1 year ago

Feel free to fork this project or to submit a PR !

I did not reproduce this bug at the time on my Windows VM...

Do you use the release for Windows here https://github.com/jersou/studio-pack-generator/releases/download/v0.1.17/studio-pack-generator-x86_64-windows.zip ? or with deno directly ?

@schmurtzm :

But theoretically wsl is not a required dependency : "Windows release use Windows TTS instead of picoTTS"

→ yes ! the zip release, from https://github.com/jersou/studio-pack-generator/releases/ use Windows TTS and it does not need WSL.

Another doc in french here : https://github.com/jersou/studio-pack-generator/wiki/Tuto-Windows-FR

The Readme should be updated to specify that it is necessary to use the zip release or to have WSL installed to have the TTS...

Anyway, there is probably a bug in the detection of WSL and Pico2Wave.

MrAlucardDante commented 1 year ago

Thanks for the quick answer.

I used the windows release.

Maybe this bug happens because I have WSL installed.

I have noticed a couple of tweaks/bug that are not related to TTS. I'll open a couple of other issues (in French if you prefer)

MrAlucardDante commented 1 year ago

Here are the logs, hope this helps : ` PS C:\Studio-Pack-Generator> .\studio-pack-generator-x8664-windows.exe "pack" { deno: "1.18.2", v8: "9.8.177.6", typescript: "4.5.2" } { opt: { : [], "skip-image-item-gen": false, i: false, skipImageItemGen: false, "skip-audio-item-gen": false, a: false, skipAudioItemGen: false, "skip-audio-convert": false, v: false, skipAudioConvert: false, "skip-extract-image-from-mp3": false, m: false, skipExtractImageFromMp3: false, "skip-zip-generation": false, z: false, skipZipGeneration: false, "skip-not-rss": false, s: false, skipNotRss: false, "auto-next-story-transition": false, n: false, autoNextStoryTransition: false, "add-delay": false, d: false, addDelay: false, "night-mode": false, t: false, nightMode: false, "seek-story": undefined, c: undefined, seekStory: undefined, "$0": "deno run", "story-path": "pack", storyPath: "pack" } } Try convert pack\test pack\story.mp3 → pack\test pack\story.item.png checkCommand [ "C:\Studio-Pack-Generator\tools\ffmpeg.exe", "-version" ] → KO : ffmpeg version 4.2 Copyright (c) 2000-2019 the FFmpeg developers built with gcc 9.1.1 (GCC) 20190807 configuration: --enable-gpl --enable-version3 --enable-sdl2 --enable-fontconfig --enable-gnutls --enable-iconv --enable-libass --enable-libdav1d --enable-libbluray --enable-libfreetype --enable-libmp3lame --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libopenjpeg --enable-libopus --enable-libshine --enable-libsnappy --enable-libsoxr --enable-libtheora --enable-libtwolame --enable-libvpx --enable-libwavpack --enable-libwebp --enable-libx264 --enable-libx265 --enable-libxml2 --enable-libzimg --enable-lzma --enable-zlib --enable-gmp --enable-libvidstab --enable-libvorbis --enable-libvo-amrwbenc --enable-libmysofa --enable-libspeex --enable-libxvid --enable-libaom --enable-libmfx --enable-amf --enable-ffnvcodec --enable-cuvid --enable-d3d11va --enable-nvenc --enable-nvdec --enable-dxva2 --enable-avisynth --enable-libopenmpt libavutil 56. 31.100 / 56. 31.100 libavcodec 58. 54.100 / 58. 54.100 libavformat 58. 29.100 / 58. 29.100 libavdevice 58. 8.100 / 58. 8.100 libavfilter 7. 57.100 / 7. 57.100 libswscale 5. 5.100 / 5. 5.100 libswresample 3. 5.100 / 3. 5.100 libpostproc 55. 5.100 / 55. 5.100 Input #0, mp3, from 'pack\test pack\story.mp3': Metadata: artist : Zouk (2021) title : Mister Ogre comment : fre:ac - free audio converter https://www.freac.org/ encoder : fre:ac v1.1.6 Duration: 00:11:05.21, start: 0.025057, bitrate: 163 kb/s Stream #0:0: Audio: mp3, 44100 Hz, stereo, fltp, 163 kb/s Metadata: encoder : LAME3.100 Output #0, image2, to 'pack\test pack\story.item.png': Output file #0 does not contain any stream

Generate image to pack/0-item.png checkCommand [ "C:\Studio-Pack-Generator\tools\convert.exe", "--version" ] convert.exe: UnableToOpenConfigureFile colors.xml' @ warning/configure.c/GetConfigureOptions/705. Generate audio to pack/0-item.wav checkCommand [ "wsl", "pico2wave", "--version" ] checkCommand [ "wsl", "pico2wave", "--version" ] Une erreur s’est produite lors du montage de l’un de vos systèmes de fichiers. Veuillez exécuter la commande « dmesg » pour plus d’informations. Generate audio to pack/0-night-mode.wav Une erreur s’est produite lors du montage de l’un de vos systèmes de fichiers. Veuillez exécuter la commande « dmesg » pour plus d’informations. Generate image to pack\test pack/0-item.png convert.exe: UnableToOpenConfigureFilecolors.xml' @ warning/configure.c/GetConfigureOptions/705. Generate audio to pack\test pack/0-item.wav Une erreur s’est produite lors du montage de l’un de vos systèmes de fichiers. Veuillez exécuter la commande « dmesg » pour plus d’informations. Generate image to pack\test pack/story.item.png convert.exe: UnableToOpenConfigureFile colors.xml' @ warning/configure.c/GetConfigureOptions/705. Generate audio to pack\test pack/story.item.wav Une erreur s’est produite lors du montage de l’un de vos systèmes de fichiers. Veuillez exécuter la commande « dmesg » pour plus d’informations. get max volume of file pack\test pack\story.mp3 → OK : 4.5 Db Convert file pack\test pack\3fb09ec7.mp3 → pack\test pack\story.mp3 "C:\Studio-Pack-Generator\tools\ffmpeg.exe" "-i" "pack\test pack\3fb09ec7.mp3" "-af" "volume=4.5dB,dynaudnorm" "-ac" "1" "-ar" "44100" "-map_metadata" "-1" "-y" "pack\test pack\story.mp3" → OK Try convert pack\0-item.png → pack\thumbnail.png → OK create pack-1675201380802.zip add asset test pack\story.mp3 add asset 0-item.png add asset test pack\0-item.png add asset test pack\story.item.png write pack-1675201380802.zip Done (13.81 sec) : pack → pack-1675201380802.zip

jersou commented 1 year ago

Une option --skip-wsl a été ajoutée à la dernière release pour désactiver l'usage de WSL pour le TTS

MrAlucardDante commented 1 year ago

Merci pour ce correctif, je le testerais dès que possible demain.

En passant le lien vers Discord n'est pas bon, c'est un lien vers un salon textuel mais il ne fonctionne pas si l'on est pas déjà membre du serveur auquel il est lié.

Je ne trouve de lien vers un serveur Discord sur la page Github, il pourrait être judicieux de l'ajouter en plus du lien vers le salon

jersou commented 1 year ago

ok, merci pour l'info sur le lien, j'avais un doute.

Je ne trouve de lien vers un serveur Discord sur la page Github, il pourrait être judicieux de l'ajouter en plus du lien vers le salon

je ne comprends pas du coup, le lien qui ne marche pas, c'est pas celui du Readme d'accueil ? ou tu parles du github de studio ?

le bon lien sans doute : https://discord.com/invite/jg9MjHBWQC

MrAlucardDante commented 1 year ago

C’est bien le lien du readme qui n’est pas bon. Merci pour le lien

jersou commented 1 year ago

Je n'ai pas reproduit sur un pc Windows avec wsl installé, mais pas pico2wave, même sans l'option --skip-wsl, le TTS a été généré.

Je clos le ticket, commenter ici si ce n'est résolu.