duangsuse::Echo
770 subscribers
4.42K photos
135 videos
583 files
6.71K links
import this:
美而不丑、明而不暗、短而不凡、长而不乱,扁平不宽,读而后码,行之天下,勿托地上天国。
异常勿吞,难过勿过,叹一真理。效率是很重要,盲目最是低效。
简明是可靠的先验,不是可靠的祭品。
知其变,守其恒,为天下式;穷其变,知不穷,得地上势。知变守恒却穷变知新,我认真理,我不认真。

技术相干订阅~
另外有 throws 闲杂频道 @dsuset
转载频道 @dsusep
极小可能会有批评zf的消息 如有不适可退出
suse小站(面向运气编程): a19a0b
Download Telegram
https://www.sciencefocus.com/future-technology/blob-opera

https://news.ycombinator.com/item?id=25441592

https://github.com/0x2b3bfa0/python-blobopera #tool 歌剧

How do you get them all to sing at the same time?

I don't get it! Can someone explain why/how this is special? Seems like it is only a matrix of like 40 states per blob(4元音10音高) and that the lower blobs just follows with the same state? Of course well executed but..?

🌚准确的说,有人考据过了一个blob的演唱矩阵是5x25的。 DDSP 可微分

有工具技能就像拿py做计算器一样超高效

https://github.com/OverlappingElvis/blob-opera-midi/blob/main/src/midi-to-blob.js#L143
duangsuse::Echo
#music b23.tv/BV1Fg411F7Xk 😂 阿b有二十多个blob歌剧 还做了 MIDI 播放功能! viola bird 应该也有?😱 可惜webui不支持一个一个blob的录,只能录一个音轨。这个lab还是很顶级的 ,HN top10 的水平 他们最牛的地方就在于:把极其硬核的前沿技术(机器学习、DDSP、实时音频合成、WebGL),包装成了一个连三岁小孩都能无脑玩的沙雕玩具。 鸵鸟项目本质上是一个连续的二维输入:X 轴控制运弓方向和力度(Bow),Y 轴控制指板音高(Pitch)。
表面上是在玩泥巴/拉大提琴,背地里其实是 AI 实时音频生成的教科书级 Demo。
Hackers里面,这种一拳超人一样的博客和demo,top是哪些?

ciechanow.ski
neal.fun
Fabrice Bellard
Inigo Quilez

这些人简直是断层式的,最重要的是,他们全都非常善于表达。好像那不是钻研一样!

​对于绝大多数程序员来说,写代码是“搬砖”或者“解题”;但对于 Fabrice Bellard 或 Inigo Quilez 来说,C 语言或者数学公式就是他们的“母语”和“画笔”。

Neal Agarwal(无限合成的作者)做那些沙雕游戏,本质上就像一个拿着极其昂贵的乐高积木(大语言模型、高并发架构)在客厅地板上搭城堡的小孩。

Linus Torvalds 写过一本自传叫《Just for Fun》(只是为了好玩)。这完美概括了这群 Hacker 的心态。
​为什么看起来不像在“钻研”或者“苦读”?
因为他们真的在玩!

这种现象通常被称为 "10x Developer" (十倍程序员) 的降维打击。但是他们确实没有什么课题或项目
This is fantastic. The overall experience is magnificantly done. As a Frontend Developer i am really impressed by the sharpness of the UI and the funcionality the interface makes possible. Does anybody know what the Techstack is or where it is posted how the built this?

- gfxgirl on Dec 17, 2020

Why does need a techstack? It just looks like custom WebGL and custom Web Audio.😁

这种把硬核技术转化为“人话”和“直觉交互”的能力,其实也是现在 AI 时代最稀缺的技能。
https://g.co/gemini/share/c7b6a2fd0557
“雕像本来就在石头里,我只是把多余的部分剔除掉。” #math 数学和验证
#今日挑战*42:伪光追与声学克拉德尼花纹(Chladni Figures & Standing Waves)
https://codepen.io/duangsuz/pen/azmLOmo

本挑战模拟固体薄板在特定频率振动下形成的克拉德尼花纹(Chladni Figures)。
薄板的振动模态由驻波方程描述。当板受到激振时,某些区域(波节线,Nodal Lines)的振幅为零。散布在板上的微小颗粒(如沙子)会因振动加速度而被推离高振幅区域,最终自发聚集在波节线上。

放弃了复杂的偏微分方程求解,直接利用驻波的解析解。梯度的计算被简化为差分近似,且完全移除了 if/else 逻辑,运动方向纯粹由波函数的导数决定。
单一粒子看似在做随机震荡,但数千个粒子的集体行为揭示了隐藏在空间中的对称性。交互的本质不是控制粒子,而是改变空间的“势能景观”,让数学结构在物理实体的堆积中被看见。
#dalao https://lovasz-local-lemma.github.io/showcases/symbolic_math_showcase/index.html

I'm a research-driven software engineer specializing in computer graphics and GPU-based simulation, with over seven years of experience in physically based rendering, shader development, and GPU pipeline architecture. My work spans both offline renderers and real-time systems, integrating academic techniques with high-performance, practical implementations across graphics, simulation, mathematics, and interactive visualization.

I see no rigid boundary between research, engineering, and art. Advanced techniques become meaningful to me only when they are transformed into concrete, interactive systems; and when rigorous internal design is paired with equally thoughtful visual presentation. I enjoy translating research ideas into robust, well-engineered implementations, repurposing techniques across domains to create expressive visual effects, and designing interfaces that make complex behavior intuitive rather than opaque.