Many links are available in Menu Bar -- Help/About, such as model download addresses and CUDA configuration. Try clicking them if you encounter problems.

1. Unable to open after double-clicking sp.exe

The software is developed based on pyside6. The main interface uses many Qt components, so loading may be slow, ranging from 5 seconds to 2 minutes. Please wait patiently.

If it still doesn't display completely within a few minutes, and no startup screen is displayed, only a black window appears, the program may have crashed. Check the console for errors. For pre-packaged versions, check if only the upgrade package was downloaded. If so, please download the full package.

If you have tried all methods and waited a long time but still cannot open it, please open the latest log file in the logs folder to check the error information, or submit the file to GitHub Issue or ask a question on bbs.pyvideotrans.com and upload the file.

2. Video clarity decreased

Some operations during translation involve transcoding, which inevitably results in quality loss. If you want to minimize the loss, you can perform the following operations:

1. The original video is an mp4 video encoded with libx264.
2. Set the menu -- Tools -- Advanced Settings as shown in the figure.

3. Translation error

After recognizing the subtitles, a red error occurs during the translation process. This usually indicates a "network connection error" or an "account problem" with the translation channel used.

If you are using translation channels such as "Google", "Microsoft", "Gemini", etc., it is most likely a network connection problem. You need to enable a VPN and fill in the network proxy address provided by the "VPN tool" in the network proxy address textbox.

If you are already using a VPN and still get a network connection error, your proxy may be unavailable. Please correctly fill in the HTTP proxy IP and port number provided by your VPN tool.

If the proxy is confirmed to be available and the error persists, the account is most likely unavailable. For example, Gemini is not available in all countries. You can try switching proxy nodes to other countries.

Gemini, ChatGPT, and AzureGPT channels generally have request frequency limits. Excluding network problems and account problems, the request frequency may have exceeded the limit. In this case, you can open Menu -- Tools -- Advanced Settings and set "Pause time after translation/s" to 30 or a larger number.

3. Low speech recognition accuracy

1. In faster mode and openai mode, using a larger model can improve accuracy. tiny is a small model with poor recognition results, while large-v3 is the largest model and provides the best recognition results. All model download addresses: https://pyvideotrans.com/model
2. If the original video is in Chinese, try using zh_recogn for better results. Instructions: https://pyvideotrans.com/zh_recogn.html
3. Selecting "Keep background sound" will perform pre-denoising, improving recognition accuracy. However, note that if the video is large, do not select this option.

4. Model download address

Model download address: https://pyvideotrans.com/model

5. Availability on Win7

Win7 is not supported.

6. Missing python310.dll

You may have only downloaded the upgrade patch package. Patch packages cannot be used alone. Please download the 1.9G full package first, then extract it. You can then download the patch package to overwrite it.

7. Error during merging ffprobe {}

This is most likely caused by spaces or single/double quotes in the original video name. Try renaming the original video. For example, the original name is D:/UNSW/2024 T2/BIOS 2061/Week 5 Amphibians, reptiles, and birds/_video_out/BIOS2061-5246_00069- Lecture 13 - Birds 1 'Origin of Birds' - Prof. Richard Kingsford- Part 2 - UNSW##BIOS2061-5246_00069- Lecture 13 - Birds 1 'Origin of Birds' - Prof. Richard Kingsford- Part 2 - UNSW.mp4. The video name is complex, containing spaces and single quotes, which can easily cause errors during processing. Please remove spaces and single quotes.

Docker deployment support

Not supported.

OCR recognition of subtitles in the video

This software recognizes human speech in videos and converts it into text subtitles. It does not support OCR recognition of subtitles.

HTTP API call

Currently unavailable. This feature may be added later.

Adding new languages

Not possible. Speech recognition to subtitles relies on the whisper model, which has limited language support. Unsupported languages cannot be recognized.

Software download location

https://pyvideotrans.com/downpackage.html

Model download location

https://pyvideotrans.com/model.html

CUDA is installed, but it still cannot be used

Possible reasons:

1. Built-in CUDA support requires version 11.8 or higher. Check if your CUDA version is too low.
2. The graphics card driver is too old and needs to be updated.
3. cudnn is not installed.
4. The graphics card is not an NVIDIA card or is incompatible.

Multi-role recognition and dubbing support

Not supported. The recognized subtitles do not distinguish between speakers and roles. This can be manually achieved through "Setting row roles".

CLI command-line mode always has problems

CLI updates are lagging. Please use the older version.

Translation error

Please change the translation channel or fill in the network proxy.

Software freezes after double-clicking, unable to start, stuck on the startup screen

The software is large. Please wait patiently. If it still cannot be opened after a long time, try:

1. Closing antivirus software and security software.
2. Ensuring that the software path and directory consist of English characters or numbers and do not contain spaces, Chinese characters, or special symbols.

If it cannot start after overwriting the upgrade package, please download the complete version.

If it is already the complete package, please wait patiently. If it still does not start after more than 2 minutes, try forcibly closing and reopening it.

Supported translations

Currently supports Microsoft Translator, Google Translate, Baidu Translate, Tencent Translate, DeepL Translate, ChatGPT Translate, AzureGPT Translate, Gemini Pro Translate, DeepLx Translate, OTT Offline Translate, FreeGoogle Translate, FreeChatGPT Translate.

Connection error

The Connection error indicates a network connection failure. If a proxy is not filled in the software interface, please fill it in. Directly connecting to ChatGPT/Gemini/Google APIs is not possible in China; a proxy must be filled in. Note that being able to open the corresponding website in the browser does not mean that it can be used in the software. Please enter the correct proxy address in the network proxy input box.

Whole all out of memory

The Whole all out of memory error indicates insufficient video memory. Please use a smaller model, such as tiny or small.

Requested float16 compute type, but the target device or backend do not support efficient float 16 computation

The error indicates that the current graphics card does not support this data type. The solution is to open Menu -- Tools -- Advanced Settings and find:

CUDA Data Type

Change the content to int8_float16.

Then restart the software and run it. If the error persists, change it to:

float32

How to install

No installation is required. After downloading the complete package, extract it and double-click sp.exe to use it.

Why is it reported as a virus or blocked?

This is software packaged using pyinstaller. It is not digitally signed or certified by antivirus software, which may cause false positives. Please add it to the trusted whitelist or disable your security software. Or use source code deployment.

Supported TTS dubbing

edgeTTS / Azure AI / GPT-SoVITS / clone-voice / elevenlabs

Source code deployment instructions

1. The default uses ctranslate2 version 4.x, which only supports CUDA 12.x. If your CUDA is lower than 12 and you cannot upgrade CUDA to 12.x, execute the command to uninstall ctranslate2 and then reinstall it.

pip uninstall -y ctranslate2

pip install ctranslate2==3.24.0

2. You may encounter errors such as xx module not found. Please open requirements.txt, search for the xx module, and then remove the == and the version number after it.

Customer support

None. This is free software with no income or profit, so we cannot provide customer support.

Pricing

This is a free and open-source project. No fees are charged. Translation and TTS interfaces are charged by their respective API providers and are not related to this project.

Commercial use

Individuals and companies can use it freely. However, if you want to integrate it into a commercial project, please follow the GPL-v3 open-source license.