Talk:Delay
From FFXIclopedia, the free Final Fantasy XI encyclopedia
Contents |
FPS of Servers?
Uh, FPS (frames per second) in the context of gaming is typically a measure of how fast an environment is being drawn by a video card. The FFXI servers, in performing all battle-related calculations, should not be rendering anything, and thus it doesn't really make sense to talk about them "falling below" any kind of FPS value. I'm pretty sure that even if an FFXI server's CPU was at full load and the network was at full utilization, as long as your weapon is drawn, you're facing the mob, and the mob isn't moving, then as far as the server is concerned, you are swinging at the correct rate (Delay/60). --Ichthyos 05:16, 1 September 2006 (EDT)
60 delay/second is the rounded average from timed video
Mithrael from ffxionline did the measurement
and he got this result:
Anyway, you can see the results in the HTML. The slowest was 58.85311871, or rougly delay/59. The fastest was 60.11583012, which as you say should technically rounded up to delay/61. I think you'll see that it's pretty convincing that delay for melee weapons is indeed 60/second.
as far as using FPS as measurement... I'm not really sure if I should agree with it.
Because FFXI servers don't draw, they simply do I/O and data (packet) processing--VZX 02:05, 2 September 2006 (EDT)
Ranged Delay
I got it how ranged delay works.
source and discussion :
Blue Gartr |
Allakhazam --VZX 15:19, 4 March 2007 (EST)
I used to believe that 1s was equal to 60 delay. That was until I did some tests on dual-wield/Haste and found the number for 1s to be closer to 59 than 60 delay.
I've recently done a test where I had a LV1 job with a ceremonial dagger swing non-stop at a mob for one hour, and the result for that test was that 1s was equal to 58.84 delay. If 1s were 60 delay, then there would have been an error of 70s, which is way bigger than the margin of error that connection lag and human reflexes could produce. It could have been the server lagging but I doubt that's the case since other tests I've done show that 1s = 59 delay too.
When observing my fps rating with Windower, I've noticed the highest stable rating it can go is 29.4. This happens even on a black screen or just having my character facing a wall in a lag-less area. 2x 29.4 is 58.8 which would make sense if the server really ran at 58.8-58.99 fps and the client at half the rate.
About ranged delay, I posted similar findings in this post, and with (1s=58.85delay) the formula in the post would change to Time = Delay / (2 * 58.85) + 2 + reshoot time.
I'm still not sure what ammo delay is used for. I speculate that shooting is similar to using spells/items except for not having a 'casting bar', and that ammo delay may have something to do with '/ra recast time'. Maybe this shooting 'recast time' is a fixed number and it would explain the 3rd period of being able to melee but not shoot yet. The ~1s cool down period in the KI post was estimated by watching the number of shoots I could fire in between 2 melee swings with different weapons.
--JKL 02:33, 6 March 2007 (EST)
Yeah, I heard that from long time ago that actual melee delay is closer to 58-59 delay/second.
May I know which job were you using for the test for ranged delay? Because I still find 110 to be a more fitting number for upper bound, regardless my rapid shot proc too often.
The delay putting back the weapon, however, varied too widely from 1.6~1.9s (nevermind what I wrote about 1.7s~1.8s on the article)
--VZX 19:49, 6 March 2007 (EST)
I used NIN and WAR for the tests mainly because I didn't want rapid shot to kick in. I obtained the numbers for the ranged delay equation by solving Time = Delay / X + Y for X and Y. For this, I had to get data from weapons of different delays.
There is a way to obtain the duration of the 'put the weapon away' period regardless of the speed at which the server is running. You can measure the time it takes to perform N melee rounds and N ranged attacks. Then do the same but this time you interrupt the N ranged attacks by moving your character. Divive the difference of the 2 measurements by N, and that's the duration of the animation.
Time_A = N * Melee_Delay + N * (Aiming_Delay + Put_Weapon_Away_Delay) Time_B = N * Melee_Delay + N * (Aiming_Delay) Put_Weapon_Away_Delay = (Time_A - Time_B) / N
Note that 'N * Melee_Delay + N * (Aiming_Delay)' in Time_A and Time_B cancel each other when you subtract them.
--JKL 06:00, 7 March 2007 (EST)
I thought ranged weapon delay/second would be much faster than 110 delay/second too, but when I use Kaman+1 (delay 350. 3.18 seconds according to my calculation, 2.97 seconds according to your calculation), occasionally, this macro doesn't work
/ra <t> /wait 6 /wa <t>
If it's close to 2.97 seconds (or round it up to 3.0s), then /wait 6 suppose to work everytime, but if we use delay(in s) = weapon_delay/110 + 3.0s = 6.18s, it'll explain why some of the /wait 6 doesn't work.
--VZX 17:53, 8 March 2007 (EST)
If ranged attack delay is calculated differently than melee attacks, all of the bows, crossbows, and throwing weapons have their DPS listed incorrectly. --Raithen 18:38, 13 July 2007 (MDT)
- Yes. In fact, not only DPS, the TP return for ranged weapon (also ammo) also irrelevant, because for compound ranged weapons (crossbows, guns, and bows) TP return is based on both ranged weapon delay + ammo delay and may give you different result if you calculate it separately.--VZX 22:35, 13 July 2007 (CDT)
--- "2. After shooting → putting back ranged weapon : This delay usually takes 1.7 second to 1.8 second. However, occasionally, it can take fluctuate from 1.6 second to 1.9 second ". The delay after shooting is infact affected by the ammo delay, you can test this out easy using high delay and low delay ammo pieces (Crude arrows vs demon arrows or grand knight arrow, sleep bolt vs Gold musketeer's bolt, or just a bow vs gun). You can also try this on throwing weapons such as a boomerang, you'll notice there's practically no delay after catching it, while projectile weapons require some time before you can shoot again. I want to edit that part but I figured some discussion should be done first :P --Radiënt 15:35, 12 April 2009 (UTC)
Calculator Removed
Link is dead. If there's a new place hosting it, feel free to update. --Ami 02:06, 25 November 2008 (UTC)
Ammo only?
I know it's not a big deal. But for the sake of completeness can we have a paragraph on thrown items? Do they fall under the 100 or the 110 rule? --AnonRamuh 20:39, 9 June 2009 (UTC)
