Manim plugin for adding speech to videos.
Project description
Manim Speech
Manim plugin for adding speech to videos.
Special thanks to osolmaz and others who worked on Manim Voiceover, which Manim Speech is heavily based on.
Installation
Manim Speech can be installed via pip using the following command:
pip install "manim_speech[*optional_dependency_sets*]"
Where *optional_dependency_sets*
is a list of optional dependency sets for Manim Speech.
Features
-
Integrate voiceovers into Manim animations.
- Utilize bookmarks to pause for key moments in the voiceover (identical to Manim Voiceover functionality).
-
Easily translate text within Manim animations to various languages with minimal code.
-
Leverage AI-driven services for text-to-speech, speech-to-text, and translation.
Services
Service Name | Optional Dependency Set | Is Local | Text-To-Speech | Speech-To-Text | Translation |
---|---|---|---|---|---|
OpenAI | openai |
No* | Yes | Yes | Yes |
ElevenLabs | elevenlabs |
No | Yes | No | No |
Whisper | whisper |
Yes | No | Yes | No |
AssemblyAI | assemblyai |
No | No | Yes | No |
DeepL | deepl |
No | No | No | Yes |
* The OpenAI service can use any backend compatible with the OpenAI API through the base_url
parameter, some of which are local (e.g., LocalAI).
** For now, the optional dependencies for the Whisper service cannot be installed automatically. To use the Whisper service, run pip install openai-whisper
.
Notes on Chinese Translations
OpenAI Translator: Please specify Traditional (zh_tw
) or Simplified (zh_cn
) when using the OpenAI translator. Only specifying zh
has no guarantee on the script used, although tests have indicated that GPT-4o strongly prefers Simplified Chinese in such cases.
DeepL Translator: The DeepL translator does not natively support translating to Traditional Chinese, so translating to Traditional Chinese requires the chinese
optional dependency set to be installed. Only specifying zh
has the same effect as specifying zh_cn
.
Usage Examples
Creating a basic scene with a voiceover with Manim Speech:
import manim
import manim_speech
from manim_speech.services import openai as openai_service
class MeaningOfLife(manim_speech.VoiceoverScene):
def construct(self) -> None:
self.set_tts_service(openai_service.OpenAITTSService()) # Remove this line if you want to manually record voiceovers.
self.set_stt_service(openai_service.OpenAISTTService()) # Only required if you use bookmarks.
with self.voiceover("What is the meaning of life?") as speech_data:
txt = manim.Text("The meaning of life is 42.")
self.play(manim.Write(txt), run_time=speech_data.duration)
self.wait_for_voiceover()
self.play(manim.FadeOut(txt))
The same scene, but translated into Traditional Chinese:
import manim
import manim_speech
from manim_speech.services import openai as openai_service
class MeaningOfLife(manim_speech.VoiceoverScene, manim_speech.TranslationScene):
def construct(self) -> None:
self.set_tts_service(openai_service.OpenAITTSService()) # Remove this line if you want to manually record voiceovers.
self.set_stt_service(openai_service.OpenAISTTService()) # Only required if you use bookmarks.
self.set_translation_service(openai_service.OpenAITranslationService()) # Remove this line if you want to manually translate text.
self.translate(__file__, "meaning_of_life", "en", "zh_tw")
_ = self._
with self.voiceover(_("What is the meaning of life?")) as speech_data:
txt = manim.Text(_("The meaning of life is 42."))
self.play(manim.Write(txt), run_time=speech_data.duration)
self.wait_for_voiceover()
self.play(manim.FadeOut(txt))
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
File details
Details for the file manim_speech-0.2.1.tar.gz
.
File metadata
- Download URL: manim_speech-0.2.1.tar.gz
- Upload date:
- Size: 10.8 kB
- Tags: Source
- Uploaded using Trusted Publishing? Yes
- Uploaded via: twine/5.1.0 CPython/3.12.4
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 97365ec5d84676b9123b2c6810bd37a49ae6d0ec712e3a00c79400eef75e30b3 |
|
MD5 | 1406eca9233dd8c99e9ab9d9e7a0defc |
|
BLAKE2b-256 | 42029d430edee9cd9775bdb69b42fe90f39f38b194b4c07576d04b8202db86bd |
File details
Details for the file manim_speech-0.2.1-py3-none-any.whl
.
File metadata
- Download URL: manim_speech-0.2.1-py3-none-any.whl
- Upload date:
- Size: 14.0 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? Yes
- Uploaded via: twine/5.1.0 CPython/3.12.4
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 6bdf47b6cc6fe2f0c99e3c6a180f6c22166b2750bba7bb8f4323b9fc92451941 |
|
MD5 | dd34fffc30dab5fd05058e220bb61bab |
|
BLAKE2b-256 | fab4c5198b6d5b2419f4718d283e4c6d7c779de028f1380529318f55568b841f |