Openvoice V1 vs V2

[FREDPIM]

Good morning,
I installed OpenVoice V1 to do channel cloning (a channel coming from a recording). This language is in French and he cannot exit it (error message). So I wanted to upgrade to V2 but it looks more like installing a new program than an update. In addition I have a graphical interface in V1 and command lines in V2. I'm not familiar with these programs because I'm a beginner and so I can't really understand the principle..
Could someone explain to me? Thank you for your help

[dosubot[bot]]

Bonjour @FREDPIM! Je suis là pour vous aider avec vos problèmes GitHub. N'hésitez pas à me poser des questions ou à signaler des bogues.

OpenVoice V2 indeed involves a different installation process compared to V1, and it might feel like installing a new program. Here are some key points to help you understand the differences and the installation process:

1. Installation Differences:

   • For V1, you download checkpoint files and extract them to a 'checkpoints' folder.
   • For V2, you download different checkpoint files and extract them to a 'checkpoints_v2' folder. Additionally, you need to install MeloTTS and run a specific command to download language data ^[1][2].

2. Graphical User Interface:

   • OpenVoice V2 does have a graphical user interface available through Gradio, which provides a more user-friendly experience compared to command-line operations. This interface includes text boxes, dropdowns, and audio inputs ^[3].

3. Functionality Improvements:

   • V2 offers better audio quality and native multi-lingual support, which might help with your French channel cloning issue ^[4].

If you're facing issues with the graphical interface in V2, ensure that you have all the necessary dependencies installed and that you're running the correct script to launch the Gradio interface.

To continue talking to Dosu, mention @dosu.

---

^{Help Dosu learn! Give it feedback: Great Response | Irrelevant Answer | Incorrect Sources | Too Verbose | Hallucination | Other}

[FREDPIM]

Thanks @dosu.
What procedure do you recommend I follow to install V2 step by step?
Does V1 have to be installed before V2 or can I directly install V2 without V1?
Thanks

[dosubot[bot]]

To install OpenVoice V2, you can follow these steps:

1. Set up the environment:

   conda create -n openvoice python=3.9
   conda activate openvoice

2. Clone the repository:

   git clone git@github.com:myshell-ai/OpenVoice.git
   cd OpenVoice

3. Install the package:

   pip install -e .

4. Download and extract the V2 checkpoint:
   Download the checkpoint from here and extract it to the checkpoints_v2 folder.

5. Install MeloTTS:

   pip install git+https://github.com/myshell-ai/MeloTTS.git
   python -m unidic download

You can install V2 directly without needing to have V1 installed. The installation process for V1 and V2 is independent after the initial setup ^[1].

To continue talking to Dosu, mention @dosu.

---

^{Help Dosu learn! Give it feedback: Great Response | Irrelevant Answer | Incorrect Sources | Too Verbose | Hallucination | Other}

[FREDPIM]

@dosu
thank you, I am on Windows 10 and I use ANACONDA3, this does not pose any problems?