Featured BMS

I discovered this BMS during my usual browsing and first found it without a BGA. Using Vizzy.io, I created four unique scenes to match the music\\\’s genre changes. The first features an Ireland-inspired forest for the Celtic elements. The second is a dubstep-inspired scene with an anime man wearing headphones and a matching T-shirt. The third shows an anime girl playing bass guitar in a more general Anime Rock theme. The fourth is one of my favorite scenes from The Melancholy of Haruhi Suzumiya ((涼宮ハルヒの憂鬱)) completing the set.

To match the visuals to the BMS autoplay, I exported the audio from Magix Vegas 18 as a .flac file after removing the video layer. I then uploaded it to the Vizzy.io scene editor and created each scene to directly reflect the songs\\\’ unique and unexpected genre changes.

The project is strong overall but not without minor flaws. The animated text title is missing a space in some scenes, and a tiny fade from Magix Vegas was left in after cutting clips.

In Vegas, I synced each 1.7GB Vizzy.io export on the timeline with audio playing. At each genre change, I cut and muted video sections I did not want, leaving the desired scene visible. This method allowed me to mix and match scenes effectively. It is computationally heavy but gave the best results for creative flexibility.

I made this video. They’re genuinely fun to make, and relatively easy once you get into the flow.
I often come across really elite BMS like this one that don’t have a BGA, which makes them a great fit for this kind of project. For this track, I chose a dark, gritty, red “hell-like” visual style to match the song’s dark, hard techno character.
There are quite a few steps involved, but I still really enjoy doing this, even well into my 40s.

Workflow:

Vizzy.io (local):
I start by finding an existing project and heavily modifying it. I add different tracks to the project and see how they look and react visually.

Next, I extract the audio from the autoplay video.
For アクジ – Don’t Stop GO (BMS edit), I add the autoplay recording into a Vegas timeline, ungroup the audio and video, delete the video track, then remaster the audio slightly (custom compressor, EQ +80dB +3, 52Hz +4).
I then render the audio as output.flac.

vizzy.io:
I add the remastered FLAC to the project.
Then I tune the beat-reactive elements to specific frequency ranges.
I create a new song icon using derivatives from the artwork included with the .bms.
I also start a collaboration with KILLER NATION at this point, who provided a lot of help.

After that, I re-animate event timings, create custom events for the song’s breaks, and listen on repeat for a good hour, constantly adjusting timings until everything feels right.

Finally, I export the video from vizzy.io to my computer — about 6.8GB. 😅

Vegas:
I import the Vizzy export, create a custom mask for the autoplay video (notefield only), set compositing to blend/multiply with a slight feather, align and sync the tracks, and add credits where due (the BMS artist is already well represented).

I render in HEVC, 1920×1080 at 59.94fps, then upload.

Phew!