I’m currently working on improving the FM sound quality for the next update.
As part of this effort, I set up a real hardware playback environment using a board called “Project RE:birth,” which allows me to run actual YM2612 FM sound chips.
Using Battle Hanafuda’s music data, I compared:
The current emulated audio
The real hardware output (via Project RE:birth)
After analyzing the waveforms and listening closely, some interesting differences became clear:
The emulation has clearer and more defined low-end, making it easier to hear individual tones
However, it tends to sound slightly muffled, likely due to a stronger low-pass filtering effect
The real hardware has better attack and presence in the mid-to-high frequencies
Based on these findings, I’m now exploring ways to combine the strengths of both approaches—
aiming for something like a “real hardware+” sound within the emulator.
This work may slightly delay the next update.
That said, achieving an FM sound that goes beyond the real hardware is a deeply interesting challenge for me,
and I’d like to prioritize it, regardless of how niche it may seem.
There are already many excellent FM emulation solutions out there.
However, many of them use GPL-based licenses, which can be difficult to integrate into commercial projects.
This implementation is planned to be part of my MIT-licensed game engine “VGS-X”, making it easier to use in commercial environments.
https://github.com/suzukiplan/vgsx
Battle Hanafuda is currently being developed as a launch title for VGS-X.