[Programming learning] Make a little progress every day, the road to programming learning: a word to MP3 production software, word walkman production tool v1.0.1
2021/03/04 The author's words:
The software was written by the author according to his own needs at the time when he was studying English at the university in 2016. Since preparing for the postgraduate exam in 2017, the software has never been maintained or updated. Until now, there are occasional netizens who use e-mail to inquire about problems encountered in the use of the software. The author has repeatedly stated that he will give up the maintenance and update of the software because he may never plan to use the software again. Therefore, since the software was released in 2016, except for a bug fix in early 2017, I have never cared about this software again. In view of the fact that after the software was released on the major software download platforms, the evaluation was still good and many people like it, so I decided to continue to maintain and update. Due to years of disrepair, the hard disk warehouse has eaten a lot of dust, and the software source code has long disappeared. If it hadn't been for the old account of Baidu Netdisk and the only backup, the source code would have disappeared in the ocean of data. The software was originally written in Python2.7+PyQt4. Due to the backward incompatibility of Python, if you continue to update, you need to follow the trend and rewrite it again with Python3+PyQt5.
=======================================
The software is FFmpegwritten based on powerful free and open source software and is only for English learning.
Foreword (2016)
Six months ago, the author searched the Internet for software related to the production of word audio files, and finally found a software called Wordstorm, which has this function, but it is not free and the cost is not low. But in the end I still bought it because I had a fancy for this feature. The flagship version is probably more than 260, I can't remember exactly how much. . .
These days, I just had nothing to do, so I started to study the production of word-to-speech. I tried it for the first time FFmpegand learned from scratch. Because I didn't have any foundation, it was quite laborious to operate.
Brief introduction of software features
The function of this software is very simple: Words to MP3 .
The quality and efficiency of word-to-speech is still very important. Most of the energy in the process of writing software is to solve these two problems.
quality
The sound quality mainly depends on the quality of the real voice library and TTS engine. The live voice library uses Wei-style live voice, and the audio quality produced by the software is fairly decent in terms of sound quality.
effectiveness
In terms of production speed, it is not necessary to synthesize through the TTS engine every time because of the way of saving information recording. As long as there is a voice file that has been generated before, it can be called directly. It is not necessary to call the TTS engine to synthesize repeatedly, especially for the synthesis of Chinese speech, because In the production process, it takes a lot of time for TTS to convert to voice. Therefore, the production speed of the software is relatively ideal, and it should be about the same as the production speed of word storm.
● Voice The
human voice is preferred for word reading. If the voice of the word is not found in the voice library [SpeechLib], it will be synthesized by TTS.
● Subtitles
LRC subtitles are accurately synchronized, subtitles are synchronized with words, including translations, and the translations can be edited and customized by yourself.
● Word reading
uses Webster’s real-person pronunciation by default. TTS synthesis is called when the voice file is not found. Of course, sentences are also available. If sentences are used, TTS synthesis is called. The key to the synthesis of voice quality is to have a good TTS engine. If you read a word aloud 1-10 times, and read it more than 10 times, I don’t think it is necessary. If you think this is necessary, you can send me an email and let me change the maximum number of readings.
● Translation reading aloud The
reading translation can also be Dict\MyDict.txtdefined by the dictionary data file , so there is no need to worry about the quality of the translation.
● Dictionary (Interpretation) The
dictionary can be customized, the default dictionary data is Dict\MyDict.txt, and the text encoding is adopted utf-8.
You can also follow the given dictionary format (see Dict\MyDict.txtformat) to edit or create your own dictionary, audio translation aloud and subtitle translations are to the dictionary file.
Main interface:
The software uses only a few examples for simple testing, and fails to predict the errors that will occur in the future. If you encounter a bug in the future, you can send an email to notify me, and I will fix it in time.
– The software can be distributed and used freely.
-If you like the software, you can click the sponsorship box in the menu bar to scan the code to sponsor me .
– Email: [email protected]
– By Caviar--2017/01/08
Update log
2017/01/13 First update.
download
Link: https://pan.baidu.com/s/19HKmS7lrTfjBf1Qy4celIQ
extraction code: 09wa
Use tutorial
The settings that need to be made when using it for the first time.
1. Set up a live voice package
SpeechLib.rarDecompress the voice file from the live voice package to the SpeechLibfile under the main program directory .
2. Set up the TTS voice engine
After installing the Chinese and English TTS voice engines, click 设置> 语音、引擎设置, select the Chinese and English TTS engines, and click 设置为默认引擎.
3. Return to the main interface, click 击开始制作, and wait for the MP3+LRC file to be completed. Each parameter can be set according to your own needs.
4. outYou can view the generated MP3+LRC file in the output folder .
