# IIDX LR2 beatoraja differences

## Gauge types

Notes

• T = Total value in BMS
• n = number of notes
• a = a function of total note count in IIDX, but depends on chart. See graph below.
• all units in %
raja/AEASY T/n T/n (T/n)*0.5 -1.5 -3 -0.5 -
IIDX/AEASY a a a*0.5 -1.6 -4.8 -1.6 -

raja/EASY T/n T/n (T/n)*0.5 -1.5 -4.5 -1 -
LR2/EASY (T/n)*1.2 (T/n)*1.2 (T/n)*0.6 -3.2 -4.8 -1.6 -
IIDX/EASY a a a*0.5 -1.6 -4.8 -1.6 -

raja/NORMAL T/n T/n (T/n)*0.5 -3 -6 -2 -
LR2/NORMAL T/n T/n (T/n)*0.5 -4 -6 -2 -
IIDX/NORMAL a a a*0.5 -2 -6 -2 -

raja/HARD 0.15 ※2 0.12 ※2 0.03 ※2 -5 -10 -5 Yes
LR2/HARD 0.1 0.1 0.05 -6 ※1 -10 ※1 -2 ※1 Yes
IIDX/HARD 0.16 0.16 0 -5 -9 -5 Yes

raja/EXHARD 0.15 ※2 0.06 ※2 0 -8 -16 -8 No
IIDX/EXHARD 0.16 0.16 0 -10 -18 -10 No

raja/DAN ※3 0.15 0.12 0.06 -1.5 -3 -1.5 Yes
LR2/DAN 0.1 0.1 0.05 -2 -3 -2 Yes
IIDX/DAN 0.16 0.16 0.04 -1.5 -2.5 -1.5 Yes

raja/EX-DAN 0.15 0.12 0.03 -3 -6 -3 No
IIDX/EX-DAN 0.16 0.16 0.04 -3.0 -5.0 -3.0 No

raja/EXHARD-DAN 0.15 0.08 0 -5 -10 -5 No

These are for 7key and 14key modes; other modes in beatoraja have different values.

• ※0 When the gauge is low, the rate of decrease is lowered; see the two sections below.
• ※1 In LR2, low T and low n leads to increased damage when on hard gauge
• ※2 In beatoraja, exhard and hard gauge increases at a lower rate when total is lower (under 250). If T is under 100 (like ★4Liberte) the gauge never increases.
• ※3 Very important: most DAN courses are tagged with gauge_lr2 to force enable beatoraja to use the LR2 DAN gauge instead.

### IIDX a-value explained

Remaining Life LR2 IIDX beatoraja
BD PR 空P BD PR 空P BD PR 空P
60+ -6 -10 -2 -5 -9 -5 -5 -10 -5
50 -4 -8 -4
40 -3.5 -7 -3.5
30 -3.6 -6 -1.2 -2.5 -4.5 -2.5 -3 -6 -3
20 -2.5 -5 -2.5
10 -2 -4 -2
• In IIDX and LR2, the gauge decreases at a lower rate when the gauge is at or below 30%
• In beatoraja, this is more of a gradual decrease from 50% and down
• According to the latest release of LR2oraja, LR2 damage reduction begins at 32% and not at 30%, and you get a stage fail at 2%. This has not been confirmed by other sources but good to keep in mind that there may be slight differences.

Remaining Life LR2 IIDX beatoraja
BD PR 空P BD PR 空P BD PR 空P
30+ -2 -3 -2 -1.5 -2.5 -1.5 -1.5 -3 -1.5
30 -1.2 -1.8 -1.2 -0.75 -1.25 -0.75 -1.5 -3 -1.5
25 -1.2 -2.4 -1.2
20 -1.05 -2.1 -1.05
15 -0.9 -1.8 -0.9
10 -0.75 -1.5 -0.75
5 -0.6 -1.2 -0.6
• Again, most DAN courses are tagged to enable LR2 DAN gauge in beatoraja.

## Timing window

For notes:

PG GR GD BD 空PR
IIDX (most charts) ±16.67 ±33.33 ±116.67 ±250 ?
LR2 easy ±21 ±60 ±120 ±200 +1000
LR2 normal ±18 ±40 ±100 ±200 +1000
LR2 hard ±15 ±30 ±60 ±200 +1000
LR2 very hard ±8 ±24 ±40 ±200 +1000
raja very easy (125%) ±25 ±75 ±187 +275 -350 +500 -150
raja easy (100%) ±20 ±60 ±150 +220 -280 +500 -150
raja normal (75%) ±15 ±45 ±112 +165 -210 +500 -150
raja hard (50%) ±10 ±30 ±75 +110 -140 +500 -150
raja very hard (25%) ±5 ±15 ±37 +55 -70 +500 -150

All units in milliseconds. Positive means before the note, negative is after the note.

Notes:

• For IIDX,
• exact numbers for excessive POORs in IIDX are not precisely known. It can happen before or after each note.
• charge notes have different rules
• A POOR happens if a button is not pressed before the LATE BAD window ends.
• In LR2, excessive POORs are only possible before a note; not after.

Detailed rules for beatoraja (based on easy judge):

• Scratch notes: add -10 +10 to each end of note timing window
• Release window of long notes: significantly larger (±120, ±160, ±200, {+220, -280})
• Release window of long scratch notes: add -10 +10 to each end of long note window
• For other judge windows: VERYEASY is 25% wider than EASY, NORMAL is 25% narrower, HARD is 50% narrower, and so on. Integer math is used to scale (discard decimal value).

## Green number, white number, LIFT

### IIDX

SUDDEN+ white number measures how much of the screen is covered by the lane cover, out of 1000. For example, the value of 227 means the 22.7% of the screen is covered.

LIFT number works the same way; it’s how much of the screen is covered by the LIFT, out of 1000.

SUDDEN+ and LIFT numbers are independent (one does not affect the other).

Green number is 10 times the number of frames a note is visible for, based on 60 frames per second. For example, GN of 270 means it’s visible for 27 frames, so `27 frames / 60 frames * 1 second = 450ms`.

(Running the game as 120Hz does not change any of the math above)

### beatoraja

LIFT number works the same way as IIDX; it’s how much of the screen is taken up by LIFT, out of 1000.

SUDDEN+ white number, without LIFT, also works the same way as IIDX, how much of the play field is taken up by the lane cover, out of 1000.

SUDDEN+ white number with LIFT is different from IIDX. The WN will now be adjusted to compensate for the fact that LIFT made the play field shorter. You first do the LIFT calculation, which means the remaining play field has been reduced from 1000 to (1000-LIFT). SUDDEN+ white number is percentage of remaining play field being covered by the lane cover.

For example, let’s say you have 200 WN, 100 LIFT in IIDX, which means the lane cover is 20% of the total play field and LIFT is 10% of the total play field. In beatoraja, you would need to use 222WN and 100 LIFT to achieve the same thing, since `200 / (1000 - 100) * 100% = 22.2%`.

Green number is slightly different as well. By default beatoraja will use raw milliseconds (e.g., 573 means a note is visible for 573ms)… but some skins will also show the equivalent IIDX green number at the same time. For example, LITONE will show the beatoraja-green-number in BLUE text, and IIDX-green-number in GREEN text.

### LR2

LR2 is complicated, and the math depends on the skin. See this page for the full math and conversions.

## Concluding Remarks

If you only care about EX-score, beatoraja is ever so slightly harder than LR2, since its PGREAT window is 2ms narrower while GREAT window is the same.

Excessive POORs work differently in LR2 and beatoraja, which can make a notable impact on life gauge.

If you are playing BMS DAN courses, the difference between LR2 and beatoraja is very minor, since the gauge calculation will follow LR2 algorithm in both cases. It’s difficult to make a 1:1 comparison, however; while beatoraja has a wider GOOD window, but excessive-poor calculation is stricter.

If you are playing difficulty tables, your experience will differ in beatoraja vs. LR2, since difficulty tables are rated using LR2 algorithm. Tables are usually rated on easy or normal gauge of LR2. Even if you ignore the differences in timing, no direct comparison can be made between LR2 groove gauge and beatoraja groove gauge, since their calculation methods are different. On survival gauges though, beatoraja tends to be easier than LR2 due to the penalty fall off in the lower part of the gauge.

If you like to use LR2 timing and gauge for everything, check out LR2oraja.

## References

https://lntakeshi.hateblo.jp/entry/2017/05/19/002127

https://github.com/exch-bms2/beatoraja/blob/master/src/bms/player/beatoraja/play/GaugeProperty.java

https://github.com/exch-bms2/beatoraja/blob/master/src/bms/player/beatoraja/play/GrooveGauge.java

https://github.com/exch-bms2/beatoraja/blob/master/src/bms/player/beatoraja/play/JudgeProperty.java

https://ch.nicovideo.jp/mirai_s_bemani_blog/blomaga/ar1389382

http://stairway.sakura.ne.jp/bms/delay.htm#input

https://hitkey.nekokan.dyndns.info/diary1501.php#D150119