forum

Taiko ScoreV2 Discussion

posted
Total Posts
255
show more
Conor
finishers for some reason can be 300'd in 2 different ways. they're kinda like 300/MAX notes on mania but on taiko you get awarded with the same points for both outcomes. this is how it works in scorev1 and i'm guessing that's how it works in scorev2 because i can't think of anything that would change it.

i believe it has something to do with the ms difference between your 2 key presses. if you hit them close enough to each other you get a MAX hit instead.
5urface

Conor wrote:

finishers for some reason can be 300'd in 2 different ways. they're kinda like 300/MAX notes on mania but on taiko you get awarded with the same points for both outcomes. this is how it works in scorev1 and i'm guessing that's how it works in scorev2 because i can't think of anything that would change it.

i believe it has something to do with the ms difference between your 2 key presses. if you hit them close enough to each other you get a MAX hit instead.
I actually investigated this a while back, within a hit window of about +-15ms you get the Finisher hit awarded without it showing as a real Finisher on screen and in stats. For me it only awards the real Finisher hit if I hit both notes on the same frame which is framerate dependant and therefore stupid. When I limited my framerate to 120 fps I could consistently hit Finishers that showed up in stats as well.
Loctav
Resolution scaling is having the sideeffect of altering the scroll speed on different resolutions, so I am currently trying to figure out a solution for that.

Kiai time/Go-Go time is currently not giving any benefits, but that might change. Let us figure out the bases first before we go into multipliers and other elements.

The display bug on fnishers, I dunno, it seem to be gone but now it is back, so I am a bit puzzled. But at least it is registering the input and doesn't break combo in case of it showing just the normal hit.

I am not sure about HP, I feel that right now it is not scaling properly, being way too easy as of currently.

HDHR turns out to be pretty easy once you play the fixed version, so I might consider buffing HR to be really worth to be on par with DT, because right now, HDHR is just cheap in comparison to DT. (as in, DT demands way more for the same score bonus, e.g. in terms of hit window)
Yuzeyun
Is there a rounding error on per-note score?
Map composition: 873 circles (a few big notes, but they're not the question), 0 slider, 0 spinner.



Edit: I think I know what caused the +159 points:


Side note, I don't know if this is just me, but the scrolling speed was faster. Is this the inception of ratio-based scrolling speed scale?

———

About big notes, have you considered that not every player has keyboards that register big notes (either perfectly or with the delay)? I mean, having misses due to having hardware that don't register big notes correctly should not be acceptable in tournament setting, as this passively tells "have a correct keyboard to play". I see no one willing to do that in the middle of a tournament, and even less if they're too young.

———
Conor
whoa you're right about the low frame rate

i always thought the display was based off ms difference because auto plays with a hit error of +/- 0.00 and gets perfect hits on them all

i can agree with the HP being too easy right now

the fixed HDHR is simply HD with higher OD (but DT should have a higher score multiplier than HDHR imo)
k3v227

Loctav wrote:

HDHR turns out to be pretty easy once you play the fixed version, so I might consider buffing HR to be really worth to be on par with DT, because right now, HDHR is just cheap in comparison to DT. (as in, DT demands way more for the same score bonus, e.g. in terms of hit window)
Perhaps HDHR could have x1.1 instead of x1.12?
karterfreak
We discovered a bug (visual) where non frame perfect finisher hits will give +2 combo instead of +1. Frame perfect finishers will give +1 as expected. Score isn't affected as far as I can tell.

Might be related to visual bug about finisher score / hit graphic not being displayed correctly if not hit frame perfect (including result screen display having the same issue)
Loctav

I already found out about that and requested a fix for it

scroll speed currently alters with used resolution, so we need to rethink this and make resolution stop having any impact on scroll speed or visible notes altogether.
frukoyurdakul
There is a SV change (well, less SV to be precise) happening at the end of the sliders (drum rolls). I think it's not intentional and needs to be fixed.
karterfreak
Doesn't it kinda make sense for higher resolutions to have a faster scroll if you're aiming for them to have identical amount of notes on the screen? 16:9 is obviously wider than 4:3 so 16:9 notes should scroll faster shouldn't they? What are you guys aiming to do here?

Also sliders visibly slow down at the end now.

edit: ninja'd by frukoyurdakul

edit x 2: SLIDERS SLOWING DOWN IS ALSO HAPPENING IN SCOREV1
_yu68
Isn't scroll speed changing intended?

I think same scroll speed as 4:3 should be applied even at 16:9 resolution.
5urface

_yu68 wrote:

Is scroll speed changing not intended?

I think same scroll speed as 4:3 should be applied even at 16:9 resolution.
That would be unfair for 4:3 screen users as they would see less notes at the same time and it would be horribly unfair on higher SV
Same scroll speed on different screen aspect ratios could only be achieved if the length of the taiko bar was limited on wider resolutions.

I think the double hit finisher/not registering as real finisher bug happens as follows: Taiko registers a hit on a finisher immediately as the first button is pressed. If the second button is pressed on the same frame it registers as proper finisher. If the second button is pressed later but still within the acceptable time frame, it will add another hit, add the score of the second hit and internally count the finisher as hit by two buttons. But as the finisher was already registered as not being hit with two, but one button it wont revert that.
To fix this you would have to change how hit registration works on finishers, only registering the single/double button press after the 30ms timeframe for successfully hitting both buttons has passed.

Currently hitting the finisher buttons on different frames also seems to award more points:

Auto SS score with perfect finishers
My SS score with finisher hits on different frames
_yu68

5urface wrote:

That would be unfair for 4:3 screen users as they would see less notes at the same time and it would be horribly unfair on higher SV
Same scroll speed on different screen aspect ratios could only be achieved if the length of the taiko bar was limited on wider resolutions.
Can't players change resolution by option?

The disadvantage by 4:3 monitor is like the disadvantage by pc specs.
Should not restrict 16:9 players due to such problems.

Osu!taiko models "Taiko no Tatsujin", 16:9 scroll is closest to it.
So osu!taiko should follow with it.

To maintain fairness, It is best to slow 4:3 scroll on scorev 2 with 16:9 scroll as reference.
Nwolf
osu!taiko is not Taiko no Tatsujin. It doesn't have to follow it and scorev2 is a step in the direction of REALLY not following TnT mechanics.
_yu68
Anyway, I think it's better to slow 4:3 scroll speed
XK2238
played for a bit, can confirm


1. result totals show correctly despite combo screwing up. I'm not seeing how the finishers are 4x worth though; for example, in the first screenshot, regular notes were worth 1200~1350, while finishers had only around 2000 or 2200 (the first note being a finisher had ~1300, wat). Also, the >1m thing.
2. kiai time still has its +20% absent.
3. HR is slower now! E.g. in 4:3 reso, the circles would still overlap a bit on the borders, compared to being exactly adjacent like it was before. Haven't looked at 16:9 reso, but some people say it's relatively even slower than in 4:3, although very slightly. Also, HD visibility area got scaled along with the reso width?

there's most likely more to be found along the way, but so far it's going pretty well. Isn't it nice to have the same on-screen time of notes across different widths though? I mean, widescreen has been quite overpowered in non-HD/HR ever since the widescreen update back in late 2012 or early 2013.

---------

alright, go on
Loctav
yeah, but right now you can factually shift your scroll speed with resolution, which sort of makes HR and HD trivial. We are revising our changes here.
Bramble
HDHR turns out to be pretty easy once you play the fixed version, so I might consider buffing HR to be really worth to be on par with DT, because right now, HDHR is just cheap in comparison to DT. (as in, DT demands way more for the same score bonus, e.g. in terms of hit window)
Other than the part where HDHR is cheap in comparison to DT, this statement just genuinely confuses me :? What do you mean by "buffing HR to be really worth to be on par with DT"?

EDIT:
Upon completion of a drum roll, a miss will be given if the player has hit less than 15% of the ticks, a 100 will be given if the player has hit between 15% and 30% of the ticks, and a miss will be given if the player has hit more than 30% of the ticks.
"and a miss will be given if the player has hit more than 30% of the ticks" Please tell me this is a typo -w-
karterfreak
So something that should be pointed out...

Scrolling differences per mod
Nomod = ScoreV2 is faster
DT = ScoreV2 is faster
HR = ScoreV1 is slightly faster?

All three of these should be unchanged from V1 if you based things off 16:9 and rescaled them to 4:3 (to make up for 4:3's smaller width). What we have now with ScoreV2 is Nomod and DT both being harder while HR got easier. Kinda throws the balance of the mods out of whack a little bit.

On the positive side of things... OHGODYESTHANKYOUFORHDHRPLSMAKEAVAILABLEINSCOREV1ASAPKTHX.
Bramble
So you're saying that scroll speed for HR will be slower in V2?
karterfreak

BrambleClaw wrote:

So you're saying that scroll speed for HR will be slower in V2?
It is slower! Not by much but it's definitely slower.
Edgar_Figaro
Interesting I'll need to test out what BPM I can play effectively with HR with new scroll speed change to see "how much slower"
Bramble
Well I'm all for slightly slower HR, maybe I can finally pass 200 BPM lol. I'm a lot more open to V2 with these changes that were updated, I haven't tried it yet though so I can't really say. I'm really glad we changed how finishers work, just because people's hardware would make the game unplayable (not me personally), so keep up the good work guys! (Still confused about things I mentioned before though)
Full Tablet
As for being able to change the scroll speed by changing resolution, the best solution is making the play area proportions standard and constant regardless of resolution (the same way the osu!mania play area scales with resolution with constant skin settings). That way, the scroll speed and time each note is on the screen is constant.

animexamera
I am really happy to see that you people who make this scorev2, that nobody wants, are still able to listen to others after all.

I still think misses on double notes are stupid, as they make certain maps and playstyles impossible.

also making maps impossible are drum roll´s and spinner´s HP drain and HP drain overally

i dont have any issues with the resolution stuff (because it actually makes sense in comparisson to the mania copies (why would you even make taiko like mania))
_yu68
I agree with Full Tablet.
Topic Starter
smoogipoo

Full Tablet wrote:

snip
So I've seen this suggested multiple times now, but this is impossible to achieve. You have one independent variable - the resolution (r), one uncontrolled (but constant) variable - the hitobject time (t), and two dependent variables - the speed of hitobjects (s) and the density of hitobjects per unit time (d).

You cannot control both s and d simultaneously - or at all. The simplest one is speed, which can be modeled as
s = r / t
Where you can see that a change in r results in a change in s, which can only be adjusted by a change in the uncontrolled variable t.

Density likewise can't be controlled, but it can be hand-wavily modeled by:
d = r / s
= r / (r / t)
= t
But t is uncontrolled.

Your "How it currently scales" box controls speed and density by keeping resolution constant.
Your "How it should scale" box is impossible as the change in density requires a change in time.

I've thought about it but the only conclusion I've come up with is to crop the area as ScoreV1 does.
k3v227
Two painless suggestions:

- Forget all of this resolution nonsense and just add this feature to taiko. Problem solved.

- Also add the note-randomization mod that mania has too :)
Topic Starter
smoogipoo

k3v227 wrote:

Two painless suggestions:

- Forget all of this resolution nonsense and just add this feature to taiko. Problem solved.

- Also add the note-randomization mod that mania has too :)
I have considered the first suggestion, but concluded that it would break HR forevermore. So that's not happening.

Second one is not a feature request for consideration right now.
k3v227
You sniped my edit so I'll just respond with it instead :P

smoogipooo wrote:

k3v227 wrote:

Two painless suggestions:

- Forget all of this resolution nonsense and just add this feature to taiko. Problem solved.

- Also add the note-randomization mod that mania has too :)
I have considered the first suggestion, but concluded that it would break HR forevermore. So that's not happening.

Second one is not a feature request for consideration right now.

Disable the ability to Decrease Speed if you use HR to preserve the integrity of the mod.

Also that's fair about the randomization mod, but that seems like a no brainer to a lot of the community.
Full Tablet

smoogipooo wrote:

Full Tablet wrote:

snip
So I've seen this suggested multiple times now, but this is impossible to achieve. You have one independent variable - the resolution (r), one uncontrolled (but constant) variable - the hitobject time (t), and two dependent variables - the speed of hitobjects (s) and the density of hitobjects per unit time (d).

You cannot control both s and d simultaneously - or at all. The simplest one is speed, which can be modeled as
s = r / t
Where you can see that a change in r results in a change in s, which can only be adjusted by a change in the uncontrolled variable t.

Density likewise can't be controlled, but it can be hand-wavily modeled by:
d = r / s
= r / (r / t)
= t
But t is uncontrolled.

Your "How it currently scales" box controls speed and density by keeping resolution constant.
Your "How it should scale" box is impossible as the change in density requires a change in time.

I've thought about it but the only conclusion I've come up with is to crop the area as ScoreV1 does.
What I meant was always keeping the proportions of the playfield constant, varying which percentage of the area of the screen is covered by it depending on the screen proportions. This way, density and speed (percentage of the screen covered per second) of the objects is always the same regardless of screen proportions.

Maybe these pictures illustrate what I mean more clearly:
karterfreak
I think a big help would be to make playfield area constant at least for width. Both of the below images are the same width, but the widescreen one is by default scrolling faster right now because the playfield between them is scaling differently.



vs

Shirai-
I think scorev2 should exist under the following circumstances
1. It should only be used on competitive games (since competitives would require mastery and perfection)
2. It should only stay as an option, since the score cap is different it should be implemented as an unranked mod to be played only on competitive or for fun, training, etc.

Contents

!Competitives

1. Talking about the "freemods" on tournaments
Mods should still always retain their rightful bonuses aside of 1m score cap

2. Fail bignote=miss/combo break, this one is allright to be implemented but picking a map for it has become harder as a challenge

3. Sliders and Spinners should stay as they have always been it should stay as bonus

4. I read words about removing the slider ticks. I dont think its a good idea since the bonus is abusable

!Should be applied onto casual plays

1. HDHR changes, since the one we know is really broken and as we have noticed that more players will come to "farm" this 2 mods especially HD players//if you know what I mean. I think it is a good implementation but I think that you will need to re calculate the "numbers" when both of this 2 mods are up

2. The consistency HR's SV on different resolutions

3. Widescreen HD does not revert back to 4:3

Now, wouldn't it be more balanced that way

P.S. not of all that I read it since there were too many posts
_yu68
It is inappropriate to make scrollspeed fast on widescreen.
It is disadvantage for 16:9 players because notes interval becomes difficult to read.(Well, this is inconsistent with my other post, so I withdraw it.)

Like told by other in here, it is best to change the 4:3 screen layout so that the same time notes as 16:9 is displayed.
Topic Starter
smoogipoo

_yu68 wrote:

It is inappropriate to make scrollspeed fast on widescreen.
It is disadvantage for 16:9 players because notes interval becomes difficult to read.(Well, this is inconsistent with my other post, so I withdraw it.)

Like told by other in here, it is best to change the 4:3 screen layout so that the same time notes as 16:9 is displayed.
This is what I'm working towards right now, will hopefully be able to push an update in the next few days but it's quite difficult :P
TimmyAkmed

_yu68 wrote:

It is disadvantage for 16:9 players because notes interval becomes difficult to read.

Like told by other in here, it is best to change the 4:3 screen layout so that the same time notes as 16:9 is displayed.
I kinda disagree here for the exact same reason, as a 4:3 player i'm just unable to read anything on widescreen resolution due to the note density on screen. While yu68 and probably all the widescreen users/players coming from TnT can probably easily read 16:9 / 16:10. There are also alot who plays only on 4:3.

The only maps where I feel like widescreen is required to be readable are Converts or Taiko diff (with 1.6 Sv while using DT)

In the end 4:3 or widescreen is a matter of preferences and i don't understand why widescreen should be considered as the "good" resolution.

The actual Scorev2 make the note density on screen lower on widescreen and higher on 4:3. I find it pretty good as it is now but the only problem is that there are now maps that became unreadable at all even if "Mostly converts" and Unranked very high SV maps.
Nofool
^ same here, as a mostly 4:3 resolution user i would then be disadvantaged.

Why would you unify all résolutions in the first place? The current system seems better. Also a good amount of players (including me) use to swap resolutions depending on the SV of the map.
Ak1o

Tasha wrote:

So something that should be pointed out...

Scrolling differences per mod
Nomod = ScoreV2 is faster
DT = ScoreV2 is faster
HR = ScoreV1 is slightly faster?
I just screened the density of some notes on a map with HR, first time with ScoreV2 and second time without.

To me it seems that HR on ScoreV1 is slightly slower than on ScoreV2. Idk if it is the maps fault or not, so I will test this on some other maps as well.
It's not much, but there is a difference.

EDIT: I tested this on maps with various BPM (140, 180, 215, 240), the change stays as little as on these screenshots.

Open both images in new tabs for the best experience.


karterfreak
@Ak1o: checked this again on a different map as the original screenshots I used to come to that conclusion showed otherwise... This time Scorev2 was faster (very miniscule difference anyways) so idk anymore, this is probably something relating to SV but it's incredibly minor

Nofool wrote:

Why would you unify all resolutions in the first place? The current system seems better. Also a good amount of players (including me) use to swap resolutions depending on the SV of the map.
I think the problem is more that equal width resolutions don't have an equal scroll for their playing field, not so much that you can use lower resolutions to have lower SV. Either way gonna wait and see what smoogipooo introduces here.
Bramble

Ak1o wrote:

I just screened the density of some notes on a map with HR, first time with ScoreV2 and second time without.

To me it seems that HR on ScoreV1 is slightly slower than on ScoreV2. Idk if it is the maps fault or not, so I will test this on some other maps as well.
It's not much, but there is a difference.
Well that's definitely an issue then :/ Is this getting any attention yet? It does seem like it's intended for ScoreV2 to have a slower HR speed than V1, so...Unless I'm mistaken of course, in which case is fine too, I don't mind. But holy shit, I haven't seen it yet but HDHR fix? Apply to V1 please? ;w;
show more
Please sign in to reply.

New reply