Blokkily是一款实验性数字音乐工作站，采用模式优先的设计理念1。该软件通过单一事件模型驱动追踪器、钢琴卷和步进网格三个同步编辑器，为用户提供统一的音乐编辑体验1。
在插件支持方面，Blokkily将CLAP作为必需的原生插件格式而非兼容性附加组件，同时也支持VST3和SoundFont格式1。其框架独立的音乐模型支持音符、和弦、转位、扫弦、微时序、概率、齿轮和参数锁定等功能1。用户可通过四个可演奏界面进行编辑，包括钢琴键盘、七种等距网格布局、九种琶音弦调和和弦垫1。
该软件集成了LLM作曲助手，支持本地Ollama模型、Gemini或MiniMax端点1。Blokkily提供MCP服务器功能，允许AI代理读取会话、写入音符、控制播放、设置仪器和截图，实现对运行中窗口的实时编辑和操作1。软件支持导入导出Standard MIDI Files和WAV/FLAC/AIFF等音频格式，采用无分配实时传输机制，实现无中断编辑和低延迟回调1。
Blokkily在GPL许可证下发布，主要依赖Qt 6.5+、FluidSynth、RtAudio、RtMidi、libsndfile和libsamplerate等技术栈1。
Blokkily is an experimental digital audio workstation that prioritizes pattern-based music creation, featuring a unified event model driving three synchronized editors: a tracker, piano roll, and step grid 1. The software natively supports CLAP as its primary plugin format alongside VST3 and SoundFont, while offering comprehensive music production capabilities including multi-track mixing, LLM-assisted composition, and MCP server functionality 1.
The music model operates independently of any specific framework, accommodating notes, chords, inversions, arpeggios, microtiming, probability, swing, and parameter automation 1. Users can interact with the system through four playable interfaces: a piano keyboard, seven isometric grid layouts, nine arpeggio patterns, and chord pads 1. For composition assistance, the platform integrates LLM capabilities supporting local Ollama models as well as Gemini and MiniMax endpoints 1. The MCP server enables AI agents to read sessions, write notes, control playback, configure instruments, and capture screenshots for real-time editing and manipulation of running sessions 1.
The software supports importing and exporting Standard MIDI Files and audio in WAV, FLAC, and AIFF formats 1. Built on Qt 6.5+ with dependencies including FluidSynth, RtAudio, RtMidi, libsndfile, and libsamplerate, Blokkily operates under GPL licensing due to its incorporation of the Rubber Band library 1. The implementation features lock-free real-time transport without centralized routing, enabling uninterrupted editing with low-latency callbacks 1.
评论
还没有评论，欢迎留下第一条。