Наталя Хандусенко AI Eng 17 April 2026, 15:43

A developer vibecoded a sound visualizer for his hearing-impaired brother to play shooters. The application, with a bunch of settings, was created in 4 hours, and he didn't even have to take his eyes off the game.

Developer Oleksandr Vozny loves playing shooters with his younger brother, who has congenital hearing impairments. Sometimes communication during the game became a real challenge, so the IT guy had the idea to create a sound visualizer. So one evening, while he was playing, Claude was creating an application for Windows on the second monitor. Oleksandr shared in detail how it happened.

Leave a comment

A developer vibecoded a sound visualizer for his hearing-impaired brother to play shooters. The application, with a bunch of settings, was created in 4 hours, and he didn't even have to take his eyes off the game.

Developer Oleksandr Vozny loves playing shooters with his younger brother, who has congenital hearing impairments. Sometimes communication during the game became a real challenge, so the IT guy had the idea to create a sound visualizer. So one evening, while he was playing, Claude was creating an application for Windows on the second monitor. Oleksandr shared in detail how it happened.

The development process looked like a real techno-performance: Oleksandr played with his brother, and on the second monitor he simultaneously “communicated” with Claude. He formulated the context, polished the prompts, and immediately threw them into the IDE terminal. He tested the result instantly — right in the game, watching how the new overlay reacts to sounds in real time.

This is what the process timeline looks like:

04:30 — first commit;
04:53 — Oleksandr records a video for his brother of silently destroying the first enemy using radar;
04:58 — sends the .exe file to his brother for testing and collects feedback from him;
08:03 — the final stable version of the radar is ready, with Opacity settings for the radar and points and lines, sound capture sensitivity, radar scaling and positioning, the ability to select monitors and sound types Stereo 2.0 | Surround 5.1 | Surround 7.1, stable operation of the radar with each type of sound, the ability to Import / Export / Save configurations;
08:30 — the brother with the visualizer does more enemy kills and more total damage than Oleksandr. He adds that “without the info, this would have been practically impossible, and this is despite the fact that we played without chat”;
08:37 — brother demonstrates how he adjusted the radar to make it convenient.

The next day, Oleksandr and Claude updated the visualizer, adding support for artificial intelligence. Claude wrote a Python script to train the AI and simplified the process of collecting data for different directions and types of sound as much as possible. However, the first experience showed that 2000 short audio files are not enough for high-quality AI training, so there is still work to be done with the data.

“My brother is very happy with the result, as am I. Claude, if you configure it correctly and set correctly formulated prompts, produces very impressive results. In 4 hours, you can create a Windows application from scratch with a bunch of settings, while playing a game at the same time,” Oleksandr concludes.

A Ukrainian created an audio player that transcribes and translates audiobooks right while you listen

Kyiv students have created a hearing diagnostic device that is suitable even for infants and can be used in frontline conditions