AC DDR A: Difference between revisions

21 bytes added ,  16 September 2017
no edit summary
(→‎General Information / Changes: unknown exactly how DDR A fixed the problem; they could be polling more than once per frame, or still polling once per frame but rounding the timing windows away from center, etc.)
No edit summary
Line 38: Line 38:
* Successfully clearing a Shock Arrow now has the O.K. judgement and the combo increase appear immediately instead of being delayed a half-beat. (This may indicate that the timing windows on Shock Arrows have been changed, but this is not confirmed.)
* Successfully clearing a Shock Arrow now has the O.K. judgement and the combo increase appear immediately instead of being delayed a half-beat. (This may indicate that the timing windows on Shock Arrows have been changed, but this is not confirmed.)
* Timing windows have been adjusted to remove the "magic BPMs" problem that existed in previous games:
* Timing windows have been adjusted to remove the "magic BPMs" problem that existed in previous games:
** As the Marvelous timing window is exactly 1 frame, and previous games measured timing windows by whole numbers of frames only, the resulting roundoff error would cause the Marvelous window to drift back and forth throughout the song, making obtaining a Marvelous Full Combo prohibitively difficult. (For example, a song with a BPM of 160 has a beat every 22.5 frames, so even a frame-perfect player aiming for a Marvelous Full Combo would have to hit some beats 22 frames apart and some beats 23 frames apart, as the Marvelous windows would be anchored on/rounded to whole numbers of frames.) The only exceptions to this problem were songs whose BPMs were evenly divisible by 3600 (such as 225, 200, 180, and 150), which would have an exact whole number of frames in between beats and be immune to the roundoff error problem. Effectively, songs with these "magic BPMs" were the only ones where it was reasonable to obtain a Marvelous Full Combo. This game fixed the problem, though it is unknown whether inputs and timing windows are now measured at the subframe level, or whether the rounding formula for the timing windows now rounds away from center (which would also fix the problem without having to change how inputs are polled).
** As the Marvelous timing window is exactly 1 frame, and inputs and timing windows are measured by whole numbers of frames only, the resulting roundoff error would cause the Marvelous window to drift back and forth throughout the song, making obtaining a Marvelous Full Combo prohibitively difficult. (For example, a song with a BPM of 160 has a beat every 22.5 frames, so even a frame-perfect player aiming for a Marvelous Full Combo would have to hit some beats 22 frames apart and some beats 23 frames apart, as the Marvelous windows would be anchored on/rounded to whole numbers of frames.) The only exceptions to this problem were songs whose BPMs were evenly divisible by 3600 (such as 225, 200, 180, and 150), which would have an exact whole number of frames in between beats and be immune to the roundoff error problem. Effectively, songs with these "magic BPMs" were the only ones where it was reasonable to obtain a Marvelous Full Combo. This game fixed the problem, though it is unknown whether inputs and timing windows are now measured at the subframe level, or whether the rounding for the timing windows now rounds away from center (which would also fix the problem without having to measure subframe inputs or change how inputs are measured).
* The music select screen now resembles that of the [[SOUND VOLTEX Information|SOUND VOLTEX]] series.
* The music select screen now resembles that of the [[SOUND VOLTEX Information|SOUND VOLTEX]] series.
** Sorting by Difficulty now follows other BEMANI games, with each Level folder containing all difficulties rated with said Level.
** Sorting by Difficulty now follows other BEMANI games, with each Level folder containing all difficulties rated with said Level.
114

edits