DanceDanceRevolution A (AC)

The first six beats all scroll by in either 55 or 56 frames each, so there's probably no slowdown there. Specifically, it takes between 332 and 334 frames for all six beats to scroll by, meaning the BPM is somewhere between 32.33 and 32.54. Combined with the first video's estimate which placed the BPM between 32.43 and 33.03 from the first two beats only, it looks like the true BPM is just about 32.5.

After that, the next three beats take 60, 65, and 72 frames to scroll by respectively. This means that not only does the song slow down starting from the first long freeze, but each of the three long freezes slows it down even more. My estimates (subject to rounding error) have the BPM dropping to 30, 27.7, then 25.

For reference, 32.5 BPM is half the speed bag scrolls at. Even if you played Over The "Period" on 8x, the beginning would be slower than Max 300. And we're supposed to have been required to read this at 1x speed, given how it immediately follows Max.(period) and the speed mod change buttons are disabled. And you can't even turn Cut on. And you can only play this song once every six Extra Stages at most, following an 18 you have to AA, following a Phase 2 you have to unlock, following a chart you have to read on 900 BPM Reverse, following a deceptively tricky 17 with 440 BPM 8th note crossovers you have to AA, following five other Extra Stages. *exhale*

(inb4 Konami actively forces 1x on this song the same way Cut is already forced off and the same way 1.5x Reverse was forced on for the first Encore, and the same way they changed the rules on Valkyrie dimension)

Did the usual frame breakdown of the video, and it turns out that the 8th note flashes become critical to analyze the new parts of the song we've just seen. I don't know why they're there, but I'm glad they are.

The cabinet lights are unfortunately largely not visible in the video, so I went by steps instead. Since they're all Marvelouses or Perfects anyway this should still be reasonably accurate.

As usual, the first six beats take 332 frames to scroll by (~32.5 BPM) and themselves don't vary much, suggesting no slowdowns. The seventh, eighth, and ninth beats take 59, 66, and 72 frames, which is within the margin of error from the previous video where they took 60, 65, and 72 frames.

The tenth beat (the fourth and last freeze) can't be measured directly because of the stop before it ends, but here's where the 8th note flashes come in. You can clearly see the flash in the middle of the tenth beat, and it takes 40 frames from the beat's beginning. This means it takes between 39.5 and 40.5 real-time increments of 1/30 sec for half a beat to pass, which bounds the BPM of the last freeze between about 22.2 and 22.8. Since the third freeze was ~25 BPM, this means that all four freezes slow down the BPM.

For simfile makers out there, it takes 29 frames from the flash (beat 9.5) before the freeze ends. This suggests that the end of the freeze is on beat 9.875 or one 32nd note's position before the next step. The stop itself arrives 6 frames later, suggesting beat 9.9375 or a 64th note before. The stop lasts 38 frames or ~1.267 sec.

Without the cabinet lights it's very hard to gain much of what the speedup is like. After it begins, the arrows are only visible for 1 frame before they halt and the FAILED screen comes up. I suppose it would be possible to count how many pixels the arrow at the bottom of the screen moves per frame and compare it to how fast the arrows move at the beginning of the song, but that's an exercise in tedium I'd rather not get involved in. What we do have is, again, that lovely 8th note flash marking the middle of the eleventh beat. It's briefly visible right between the "I" and "L" of "FAILED" during the shutter-closing animation. This comes 7 frames after the player hits the Great. However, from the video and the movement of the note at the bottom of the screen, we can safely assume that the player hit the Great either 1 or 2 frames early, which means that it takes between 5 and 6 frames for half a beat to pass during the speedup. Rounding both ways, we can bound half a beat between 4.5 and 6.5 frames, which yield BPMs of between ~138 and 200. A rather wide range, but the best we can do given what we have, and enough to know that playing above 4x speed might be risky.

Massive props to whoever thought to turn on the measure guides on the left; it made everything so much easier since the chart halts before the second step after the stop can be reached.

It's still very hard to estimate but not nearly as bad as before. Whatever chart the cabinet lights are following seem to have steps too fast for the lights to exactly follow. It takes 9 frames from the moment the steps start speeding up on the right to the moment the measure guideline passes the target zone on the left. Estimating between 8 and 10 frames bounds the speedup's BPM between 180 and 225. Still a rather wide range. I'll try to look at the cabinet lights more, but it's hard to track.

The second beat past the stop just barely crosses the target zone before the FAILED shutter covers it, but that's all we need. Thanks to the 6x speed it's visible that 9 frames per beat is almost an exact number, which would place the speedup at an even 200 BPM. There's also another BPM change that comes right before the speedup and after the stop, which lasts no more than 2 frames and will probably be impossible to measure without pixel counting.

DDR이 하고싶어요... 아 주7일은 역시 넘 빡센가