用于创建音高时间线的简单工具_python_代码_下载

# Hachiko [](https://pypi.org/project/hachiko-bapu/) [](https://github.com/duangsuse-valid-projects/Hachiko/wiki) <div align="center"> <img alt="Hachiko" src="Hachiko.png"></img> </div> Simple tool for creating pitch timeline, this program divides midi creation into `pitches` and `timeline` part. When creating timeline, press <kbd>A</kbd> to give position/duration information, and use <kbd>S</kbd> to split different notes directly when holding <kbd>A</kbd> ```python # Twinkle Little Star [45, 45, 52, 52, 54, 54, 52, 50, 50, 49, 49, 47, 47, 45, 52, 52, 50, 50, 49, 49, 47, 52, 52, 50, 50, 49, 49, 47, 45, 45, 52, 52, 54, 54, 52, 50, 50, 49, 49, 47, 47, 45] ``` > Tip: press <kbd>K</kbd> in pitch window, and input code in console (__it's recommended to launch this application in console__) The name of the project is *Hachiko*, inspired by the golden yellow Akita dog - ハチ公, which is in homophonic with "扒公" (means melody extraction "耳 Copy" or "扒谱"). ## Installing ~~There's no need for system-wide installation, just use the script `hachi.py`~~ Since version 1.1, setting up the installation using `setuptools` is required. > For Windows installation guide, please read [this wiki](https://github.com/duangsuse-valid-projects/Hachiko/wiki/Windows-Installation) ```bash python3 setup.py install # can be --user or sudo ``` or get it from PyPI: ```bash pip3 install hachiko-bapu #v or, for the latest version pip3 install --upgrade git+https://github.com/duangsuse-valid-projects/Hachiko ``` System library [FluidSynth](https://github.com/FluidSynth/fluidsynth) is required to run this application. ### Old installation-free version See [release 1.0](https://github.com/duangsuse-valid-projects/Hachiko/releases/tag/v1.0), get a zip(or tar) and uncompress: ```bash pip install --user -r requirements.txt python3 hachi.py # use midnotes.py to replace srt2mid.py print-notes ``` ## UI Control / Basic Routine Hachiko is self documented, so just use the program. ```bash hachiko -h ``` > NOTE: For the first time using GUI, you can spend more time learning hot keys Once `puzi.srt` is generated, you can use `srt2mid puzi.srt` to transform it into MIDI file Btw, you can use pitches from extrenal tool (like [Synthesizer V Editor](https://synthesizerv.com/) or [MidiEditor](https://www.midieditor.org/)) extracted by `srt2mid print-notes puzi.mid` instead of built-in approach Btw, there's also an option to use [MELODIA Algorithm](https://github.com/duangsuse-valid-projects/audio_to_midi_melodia) to extract pitches directly from music ## Tool `srt2mid` and `lrc_merge` [srt2mid.py](hachiko/cli_tools/srt2mid.py) can be used to make conversation between SRT / MIDI File format Output filename is determined automatically from input path, and SRT representation of MIDI track will be timeline of integer(note pitch)s. The default mode, "from", means "from srt to mid", and when extracting lyrics from mid file you have to use "back-lyrics" instead. ```plain Usage: srt2mid [ from/back/back-lyrics/print-notes ] files... ``` [lrc_merge.py](hachiko/cli_tools/lrc_merge.py) can be used to merge words-based lyrics into sentence-based lyrics ```plain usage: lrc_merge [-h] [-dist n_sec] [-min-len n_sec] [-o name] [-sep word_seprator] (path / 'lrc') ``` + `dist` max distance for words in same sentence, default `0.8` + `min-len` min duration for the last word in sentence (only when `lrc` input is used) Execute `lrc_merge -h` to see full details