forum

[invalid] [Proposal] Allow different timing between difficulties in a mapset if necessary

posted
Total Posts
25
Topic Starter
olc
Currently, the ranking criteria states the following:
Uninherited timing points must be the same in every difficulty of a beatmap. Each point must have the same BPM and offset in each difficulty.

In a lot of songs with variable timing, different layers of music are out of sync with each other. Because low difficulties often follow simpler layers than high difficulties, to ensure the best possible timing, the difficulties would need to diverge. The same thing applies with guest difficulties following different layers, as different mappers usually have different interpretations of the song.

I propose making this a guideline, adding a clause that excludes cases involving songs with unsynced layers:
Uninherited timing points should be the same in every difficulty of a beatmap. Each point should have the same BPM and offset in each difficulty. Uninherited timing points can only differ between difficulties in variable timing cases where the instruments are unsynced in order to allow separate difficulties to follow separate instruments.
wafer
seems logically sound
Ralkinson
absolutely agree
lewski
Fisky
YES
Topic Starter
olc
community/forums/topics/1094337 doesn't deal with the layer of music being followed, the user just didn't know you could snap sliders through red lines via SVs

The solution proposed in community/forums/topics/1167387 seems like a major bandaid. Using odd snap divisors to get the note on the millisecond you want seems like an RC-friendly way of placing unsnapped notes, and it's usually impossible to get equal gaps between several consecutive notes with this method.
quila
agree
Izzeee
agree
laura-
YES PLEASE.
THAT_otaku
Based
[[[[[[
+1
YuEast 2018
For different layering, simply use another grid when editing and revert the grid (to what most people feels) can help to solve this. Because osu uses miliseconds as the absolute position for every object, it doesn't matter when changing the grid snap. That's the case which osumania have many of it, especially for rare snap "dump" (e.g. something like 1/14).
Topic Starter
olc
bump
quila
If this proposal is also accepted, the wording on this one should be updated to reflect that, since it's another case where some red lines are only needed on one diff
Ryu Sei
Strongly disagree because we already have this rule:
  1. Hit objects must be snapped within less than 2 ms of any timeline tick. AiMod will report these issues, as well as rare false positives. False positives mainly occur on slider ends and reverses, and should be verified manually or with other tools. Objects in a musical section requiring unsupported beat snap divisors (e.g. 1/11) can either:
    1. Remain unsnapped, as long as they align with the intended beat snap divisor.
    2. Be snapped through a temporary change in BPM.
The suggestions given from the rule already satisfies the issue you're having on.
lewski
the allowance in that rule quite explicitly only applies to "sections requiring unsupported beat snap divisors" which is not the same thing as different layers of a song being out of sync with each other so it really doesn't help with the issue at all
Ryu Sei
Isn't that basically mapping II-L's songs in nutshell...

I believe BNs know if it's completely acceptable if easier diffs has majority of its hit objects 'out of snap' due to beat simplification. Hence it's unnecessary to make multiple difficulties with different uninherited timing points.
lewski

Ryu Sei wrote:

Isn't that basically mapping II-L's songs in nutshell...
no, this has nothing to do with intentionally complex rhythm

in this context "unsynced layers" refers to the layers genuinely being out of time with each other, usually because the song is unquantised

in practice, this results in situations where sounds that land on the exact same beat in the sheet music (or otherwise in terms of music theory) don't happen at the same time in the actual recording

for example, let's say a song has a bass drum and a trumpet. both play a note that's intended to land on a downbeat, but because humans aren't perfect, the sounds actually happen 25ms apart. the layers are "unsynced" at that point in the song.

Ryu Sei wrote:

I believe BNs know if it's completely acceptable if easier diffs has majority of its hit objects 'out of snap' due to beat simplification. Hence it's unnecessary to make multiple difficulties with different uninherited timing points.
the entire point of the proposal (and certain previous proposals) is that this ISN'T considered acceptable
Sparhten
gonna revive this thread with a +1, as the previous reasoning while sound by pishi in the first post, and the latter talking about snappings, still dont encompase edge cases, and i see no harm in adding it as a guideline.
-Koish-
+1, could make low-diff creation more interesting
Ephemeral
For the sake of comparison, is anyone able to provide two short sample difficulties illustrating the kind of timing changes this would encompass? The proposal does sound fairly solid at a glance, but being able to see what kind of changes are wanted in the actual would help a fair bit.
Dabbe_01

Ephemeral wrote:

For the sake of comparison, is anyone able to provide two short sample difficulties illustrating the kind of timing changes this would encompass? The proposal does sound fairly solid at a glance, but being able to see what kind of changes are wanted in the actual would help a fair bit
This could work well for hybrid sets for example, like here - beatmapsets/1663982#osu/4007934 (I heard this one explained to suffer from the very accurate timing needed for mania, because some areas were beyond unfair to acc on std). Applicable bc different gamemodes may warrant different levels of timing leniency
-White
Bump on this. Polyrhythm maps exist that would warrant different timing depending on the mapper's preference on layer to follow. Would do well to bundle with community/forums/topics/1862002?n=1
Pope Gadget
Completely agree that an amendment needs to be made here.

I'm currently in a dilemma with the chart I've been trying to rank most recently, which is a multi-mode set containing mania and standard difficulties. The timing for the song is incredibly complex, which works well for mania but makes the standard difficulty (according to some high-level standard players) nigh impossible to play without simplification. The full discussion bringing it up can be found here: https://osu.ppy.sh/beatmapsets/1663982/discussion/-/generalAll#/4095633.

The simplest solution here, to me at least, is to change the rule to a guideline and add a clause explaining when it would be reasonable to go against it, such as:
  1. differences in playability between modes
  2. differences in which instrument each difficulty focuses on if the syncing of the instruments is noticeably off

Ephemeral wrote:

For the sake of comparison, is anyone able to provide two short sample difficulties illustrating the kind of timing changes this would encompass? The proposal does sound fairly solid at a glance, but being able to see what kind of changes are wanted in the actual would help a fair bit.
I'm happy to do so for this one since I imagine I'd have to to get this set back off the ground anyway, but whilst I do that I suppose just imagine the mania timings without any red lines between beats (at least for the denser/140bpm~ parts), and BPM values readjusted to match.
Nao Tomori
Proposal denied by peppy, archived.
Please sign in to reply.

New reply