Barbarian capstone seems to give a hidden haste boost

[eulogy098]

my mate and i (me 20 WF Barb, he 20 WF Fighter- both THF) decided to put this entire speculation to rest. Us two, and our WF Sorc friend zoned into TOD (just for the sake of it being a combat zone) and had a bit of a swing contest. Both of us recording with fraps as well as watching each other, we find that with absolutely no buffs he and I swing at the exact same speed (we both have our respective capstones)

HOWEVER--------------

When our sorc buddy casts Haste upon us- my Fighter friend very clearly swings faster. We are in the process of counting the exact swings but it is visable and clearly inarguable. Once we finish our counts and youtube finishes its upload i will post our videos.


Basically, i am totally and 100% positive that this seemingly impressive Barb Capstone Speed Increase is bunk as it does NOT stack with Haste. It does, however, exist. Capstone Barbarians are attacking as fast as Capstone Fighters if both are totally unbuffed. The Barbarian boost does not stack with haste, the fighter one does.

Draw your own conclusions on if this is a significant game changer.

[Shade]

my mate and i (me 20 WF Barb, he 20 WF Fighter- both THF) decided to put this entire speculation to rest. Us two, and our WF Sorc friend zoned into TOD (just for the sake of it being a combat zone) and had a bit of a swing contest. Both of us recording with fraps as well as watching each other, we find that with absolutely no buffs he and I swing at the exact same speed (we both have our respective capstones)

HOWEVER--------------

When our sorc buddy casts Haste upon us- my Fighter friend very clearly swings faster. We are in the process of counting the exact swings but it is visable and clearly inarguable. Once we finish our counts and youtube finishes its upload i will post our videos.


Basically, i am totally and 100% positive that this seemingly impressive Barb Capstone Speed Increase is bunk as it does NOT stack with Haste. It does, however, exist. Capstone Barbarians are attacking as fast as Capstone Fighters if both are totally unbuffed. The Barbarian boost does not stack with haste, the fighter one does.

Draw your own conclusions on if this is a significant game changer.

Good to know im not the only one with to perform such a trivial test and draw a sensible conclusion.

Really sorry guys but comon, this isn't a hard test to perform so take it for what it is, accept our results, or hey, go do it yourself, it takes 5 minuits.

[Nick_RC]

Nice verification Eulogy. I like it as the non pures dont suck in comparison.

[Greydeath]

My question is how in the hell did the entire purist barb population miss this for so long?

It was no doubt Shaded

[Vanshilar]

my mate and i (me 20 WF Barb, he 20 WF Fighter- both THF) decided to put this entire speculation to rest. Us two, and our WF Sorc friend zoned into TOD (just for the sake of it being a combat zone) and had a bit of a swing contest. Both of us recording with fraps as well as watching each other, we find that with absolutely no buffs he and I swing at the exact same speed (we both have our respective capstones)

Eh the question is not whether or not the barb capstone's speed increase is the same as the fighter's, it's whether or not it's giving a speed increase at all. Thus the proper comparison is not hasted fighter with capstone vs hasted barb with capstone, it's hasted barb without capstone vs hasted barb with capstone.

So I did the same sort of thing as last time. I don't have a pocket mage handy, so I used the madstone boots instead. Started the madstone, swung around until it expired, then ran out to barb trainer, got the capstone, ran back, reactivated madstone, swung around again. This time though I needed to swing in a tavern since I only have 1 madstone and needed it to recharge. This time I recorded it in 10 frames per second rather than 20, hope that helps, although as I've mentioned before, my system isn't set up very well for video recording. Once again it's a continuous video to decrease the chance of other factors such as (for example) server load at different times of day or something.

The link to the video is here:

http://www.youtube.com/watch?v=DhPnx9x5qBI

The results were (based on the video):
No capstone with madstone: first attack (attack round 0) at frame 44 (0:00:04.4), last attack (start of attack round 52, i.e. 51 completed rounds) at frame 1225 (0:02:02.5). This is 51*4 = 202 completed attacks in 118.1 seconds, or 103.6 attacks per minute.
Capstone with madstone: first attack (attack round 0) at frame 1886 (0:03:08.6), last attack (start of attack round 55, i.e. 54 completed rounds) at frame 3062 (0:05:06.2). This is 54*4 = 216 completed attacks in 117.6 seconds, or 110.2 attacks per minute.

This means that taking the original 85.8 attacks per minute as the base speed, the percentages (in attacks per minute) are:

Barb w/o capstone: 100%
Barb w/ capstone: 109.8%
Barb w/o capstone, madstoned: 120.8%
Barb w/ capstone, madstoned: 128.4%

So the barb capstone is providing an attack speed boost that stacks with madstone. Feel free to do your own testing if you think this isn't the case, since my video capturing setup isn't that great.

As for fighter capstone vs barb capstone, I would recommend testing for a longer period of time unhasted to see if they're the same, since they may only differ by a percent or two. I have no idea what the actual *tested* THF increase attack speed is for a fighter capstone.

[Monkey_Archer]

The link to the video is here:

http://www.youtube.com/watch?v=DhPnx9x5qBI

Thanks for posting.

With capstone, starting at 1:58 or 1:59 on timer i saw ~ 28 full animations in 1 minute (112 attacks per minute)

My fighter video shows the exact same attack speed with capstone. (starting at 1:55 on mine)

So its pretty clear that both capstones provide the same bonus, with the same stacking.

So why doesn't my fighter get a +2 strength bonus too? is my capstone bugged?

[Nick_RC]

Thanks for posting.

With capstone, starting at 1:58 or 1:59 on timer i saw ~ 28 full animations in 1 minute (112 attacks per minute)

My fighter video shows the exact same attack speed with capstone. (starting at 1:55 on mine)

So its pretty clear that both capstones provide the same bonus, with the same stacking.

So why doesn't my fighter get a +2 strength bonus too? is my capstone bugged?

So wierd people are getting dif results... guess video footage wins tho.

[vVAnjilaVv]

Are you guys still discussing this?!?!?!?!....barbarians are awesome....'nuff said....get over it!

[Vanshilar]

Thanks for posting.

With capstone, starting at 1:58 or 1:59 on timer i saw ~ 28 full animations in 1 minute (112 attacks per minute)

My fighter video shows the exact same attack speed with capstone. (starting at 1:55 on mine)

So its pretty clear that both capstones provide the same bonus, with the same stacking.

So why doesn't my fighter get a +2 strength bonus too? is my capstone bugged?

The question I want to ask is, why is your video so smooth? We're both using CamStudio yet I have problems with repeated frames causing choppy motion and stuff. It's not an FPS issue since I see it smooth while I'm recording it, but on playback things get all choppy. I've tried full screen and windowed (I usually play in windowed mode) and it doesn't seem to make a difference, both get the repeated frames. I record at 10 FPS while the usable data is more or less like 4-5 FPS due to all the repeated frames.

Anyhow, I tried testing with a capped fighter guildie, but get wildly different results. Without the capstone, his attacks/min was 87 without madstone and 109 with madstone (so somewhat faster than my barb's 86 and 104, respectively), but when he got the capstone, on the three tests, his attacks/min were 97.6, 92.9, and 88.0, while my own speed stayed pretty consistent (94.1 and 94.6 for the two that I swung with him). Fighter capstone plus madstone was even stranger, with one test giving 101.9 attacks/min and another giving 117.6 attacks/min (the second one, where I swung with him, I did 109.7). I'm at a loss as to why the speed differs so much. It wasn't a lag spike; each attack was simply slower by those fractions of a second.

It's possible that it has something to do with equipment or something. He's a stalwart defender (if that has a hidden attack speed side effect) but it wasn't active. What's weird is one test with both of us capped and madstoned, he got 117.6 attacks/min and I got 109.7 attacks/min, but immediately after, un-madstoned, he was swinging at 88 attacks/min while I did 94.6 attacks/min. I don't know if it's server issues (a la DPS lag where other player's attacks aren't being sent consistently to you) but it looks like general weirdness is the order of the day.

[Monkey_Archer]

The question I want to ask is, why is your video so smooth? We're both using CamStudio yet I have problems with repeated frames causing choppy motion and stuff. It's not an FPS issue since I see it smooth while I'm recording it, but on playback things get all choppy. I've tried full screen and windowed (I usually play in windowed mode) and it doesn't seem to make a difference, both get the repeated frames. I record at 10 FPS while the usable data is more or less like 4-5 FPS due to all the repeated frames.

Anyhow, I tried testing with a capped fighter guildie, but get wildly different results. Without the capstone, his attacks/min was 87 without madstone and 109 with madstone (so somewhat faster than my barb's 86 and 104, respectively), but when he got the capstone, on the three tests, his attacks/min were 97.6, 92.9, and 88.0, while my own speed stayed pretty consistent (94.1 and 94.6 for the two that I swung with him). Fighter capstone plus madstone was even stranger, with one test giving 101.9 attacks/min and another giving 117.6 attacks/min (the second one, where I swung with him, I did 109.7). I'm at a loss as to why the speed differs so much. It wasn't a lag spike; each attack was simply slower by those fractions of a second.

It's possible that it has something to do with equipment or something. He's a stalwart defender (if that has a hidden attack speed side effect) but it wasn't active. What's weird is one test with both of us capped and madstoned, he got 117.6 attacks/min and I got 109.7 attacks/min, but immediately after, un-madstoned, he was swinging at 88 attacks/min while I did 94.6 attacks/min. I don't know if it's server issues (a la DPS lag where other player's attacks aren't being sent consistently to you) but it looks like general weirdness is the order of the day.

Which codec are you using? I downloaded DivX and that made a huge difference.

My Cam studio settings are:
Quality: 20
Set key frames every 30 frames
Capture frames ever 16 milliseconds
Playback rate 60 fps

Divx settings:
1080 HD Profile
1-pass
Bitrate 7000 kbps

In game i had to reduce resolution to 1080 x 720 (the highest resolution divX 1080 HD profile can handle at 60 fps)


I would blame different results on latency. Just having 50-100 ms latency both ways could result in a whole swing difference. Trying to watch another player's actions is difficult as latency fluctuation can make them appear to swing faster or slower then the server is registering it. This is why video is often the best method to compare swing speeds as the video is only of one machine and takes latency out of the equation.
Still, that much fluctuation in swing speed seems odd to me... might be good to have multiple videos to confirm swing speeds...

[ThunderTank]

Any other tests upon the capstone aspd stacking with haste/madstone???
Would like to see a 100% reliable one.

What if u use a vicious weapon on the aussicarex portal, counting the dmg you deal to yourself after 1 minute?
With dungeon scaling should be 1dmg/swing , therefore giving you the undeniable swing/minute result.
Not sure how effective that method would be. A friend of mine came up with the idea.

[Shade]

I'll do one with proper quality, at playable fps before I tr in a week or 2.

The game isn't even remotely playable at 10fps, so I can't really trust results shown at those fps. I seem to get widly different results depending on my FPS. Even settings such as if I enable vsync seem to effect it.

There's something weird going on here.. We'll figure it out.

Far a recording goes.. You really need to record at at least 30fps to get real results that are clear.
For capture I'd suggest either Fraps or Xfire.
Fraps isn't free but I believe Xfire is, and is much superior to camstudio as it can capture full lossless video on the fly with less overhead.

For encoding there's really only 1 really solid app out there:
Virtualdubmod (mod of virtualdub, bit older but I find it works better)
Best Codec: x264vfw (far superior to divx)

Settings:
Filters - Add. Resize your video to either 1280x720, or 1920x1072 for best quality on youtube, regardless of your chosen game recording resolution. Mode: Lancoz3
Codec Settings:
Bitrate nice and a high around 8000-10,000

[eulogy098]

The absolute only way to really settle it is to have the same person test both the Fighter and the Barb on the same computer, at the same frame rates. The fact is that things like framerate seem to have a very dynamic impact on the perceived attacks per minute.


Suggested test and the only one that anyone should accept as valid:

One person must have both a capstone fighter and a capstone barb and ideally of the same race
Strip off all equipment except madstone boots and Two Hand weapon (same weapon for both toons)
Zone into an area with no other players (combat zones)
Swing away for the duration, and than count every swing
Close DDO completely and than repeat test on the other character
Both tests done with at least 30 FPS

Comparing one persons character to another on a different computer, with different latency and frame rate and who knows what other variables shouldn't be considered accurate.

[Vanshilar]

What if u use a vicious weapon on the aussicarex portal, counting the dmg you deal to yourself after 1 minute?
With dungeon scaling should be 1dmg/swing , therefore giving you the undeniable swing/minute result.
Not sure how effective that method would be. A friend of mine came up with the idea.

The problem with this method is that you can also do vicious damage on glancing blows, at the glancing blow proc rate. Since it's probabilistic, you'd get different amounts of vicious damage from glancing blows on every test, thus throwing off your results. (This is for THF btw, I think you should be able to use this technique for other weapon styles.)

Regarding frame rate (FPS), note that there are two: the game's own frame rate (how fast the game is processing stuff) and the video grabber's recording frame rate. Your attacks would be based on the game's frame rate (since it's the game timing when to send the attack signals), which you can see in the upper left corner of the video. It stays fairly constant throughout. It doesn't particularly matter what the video grabber's recording frame rate is as long as it's fast enough to see each attack. By the way, the game plays fine when recording, i.e. I don't notice any perceptible difference between recording and not recording, it's the video capture that's choppy. There's no need to time the attacks down to the length of a single attack; there are about two minutes of attacks, and it's fine to take the average time of the attacks over this time to get your attacks/min. Thus a lower capture speed is fine as long as you can still resolve each attack (in order to count how many have elapsed). Obviously though a slower playing speed will affect your attacks per minute, which I tested before here:

http://forums.ddo.com/showthread.php?t=213564

The absolute only way to really settle it is to have the same person test both the Fighter and the Barb on the same computer, at the same frame rates. The fact is that things like framerate seem to have a very dynamic impact on the perceived attacks per minute.

Remember there's two different issues:
1) Whether or not the barbarian capstone gives an attack speed bonus compared to without it.
2) Whether or not such a bonus (if it exists) matches the fighter capstone bonus.

I think you're going for testing #2, which is perfectly fine, but to do that first, you should first know if #1 is true or not. My tests were done in one continuous video (so there shouldn't be many other factors such as testing at different times of day when the server load might be different, or video settings might've changed, etc.) to best compare #1. For #2, any fighter can actually do the same, i.e. one continuous video of swinging without the capstone, then getting the capstone, and swinging again. This would allow us to compare attack speeds with and without capstone. I presume both attack at the same speed without capstone (although, of course, I don't know if this has been tested before).

[Monkey_Archer]

Regarding frame rate (FPS), note that there are two: the game's own frame rate (how fast the game is processing stuff) and the video grabber's recording frame rate. Your attacks would be based on the game's frame rate (since it's the game timing when to send the attack signals), which you can see in the upper left corner of the video. It stays fairly constant throughout. It doesn't particularly matter what the video grabber's recording frame rate is as long as it's fast enough to see each attack.

Technically there are 3 "rates". The game itself (or any game) will run separate from the graphics rate (fps). You can test this quite easily. Find a static target, hold your attack button and turn off your monitor 0 fps yet the game still runs and processes every attack. (not an entirely accurate example but you get the point) You could technically run the game and the video at 2 fps (matching the approximate 2 attacks per second) and still get fairly accurate results.

NOTE: a drop in game performance can cause a drop in graphics performance, which may affect swing speed. But you cant tell if that happens just by looking at fps.

[ThunderTank]

Im no part of the number cruncher's team and video making team... Too lazy to do any of that.
But I did opened 2 clients, one with a level 20 THF WF Barb with the capstone, and the other with a 18 Barb/2 Fighter THF WF Barb.
Without haste and madstone, the pure build w/ capstone was clearly faster. With haste casted, the pure build was also slightly faster. Procd madstone+haste , and the attack speed difference was barely seen while twitching... But was still there, while swinging and standing still...

[Kaeldur]

Just for the sake of it: get a non-pure barb and a pure capstoned barb and test the difference between them... (I would do it if I had a pure 20 barb...). I know it doesn't make much sense since you can just test the same 20 brb w/o capstone, but neither does this capstone haste bonus anyway...

[Vanshilar]

Technically there are 3 "rates". The game itself (or any game) will run separate from the graphics rate (fps). You can test this quite easily. Find a static target, hold your attack button and turn off your monitor 0 fps yet the game still runs and processes every attack. (not an entirely accurate example but you get the point) You could technically run the game and the video at 2 fps (matching the approximate 2 attacks per second) and still get fairly accurate results.

Well, I'm not sure if what's displayed when you activate the frame per second counter (ctrl-F) is the game's internal state update rate or its video processing rate (how often it's sending a new graphics frame to your video card). However, I assume those will be fairly close proportionally, i.e. if one lowers then so will the other, all other variables being constant, etc. My point though was that this is different than the video capture rate of 10 FPS, which is part of CamStudio's settings and doesn't have much to do with the game's own rate. You can see in the video that the FPS counter stays pretty similar with and without the capstone to show that the amount of client loading when the test was done was not a factor.

In terms of just how low a video capture FPS you need to get accurate attack speed data, I was just swinging continuously without moving (i.e. non-twitch). Thus, I wasn't actually counting every attack, but every attack round (4 attacks), which occurs approximately every 2 seconds. Essentially, then, I was capturing a signal that has a frequency of around 0.5 Hz; from sampling theory, then, which states (essentially) that your sampling frequency should be more than twice that of the signal that you're sampling, I only need to record at a frame rate of 1 FPS or faster to get accurate results. Of course it's desirable to be a bit faster because the theory is for a sinusoidal signal and my character goes through different animations throughout an attack round (not just back and forth), plus generally you don't want to do things at the theoretical minimums (or maximums); however, 10 FPS, or 5 FPS considering the repeated frames, is more than adequate to resolve when each attack round occurs, and indeed each individual attack. Faster than that may be prettier (less choppy), but won't affect your ability to count the number of swings, and takes up more hard drive space.

Just for the sake of it: get a non-pure barb and a pure capstoned barb and test the difference between them... (I would do it if I had a pure 20 barb...). I know it doesn't make much sense since you can just test the same 20 brb w/o capstone, but neither does this capstone haste bonus anyway...

The problem with a non-pure barb would be the introduction of other variables. The reason why I tested with a level 20 pure barb (other than, well, that's all I have to test with) with and without the capstone is to decrease the number of other variables that might influence the test. What I mean is that it's possible, for example, that by taking some other class you might have inadvertently taken some other enhancement that affects attack speeds. Testing with a pure barb without the capstone and then immediately thereafter (in fact, within the same video) with the capstone removes those other variables, since the only variable that changed is the existence of the capstone (or, depending on how closely you want to look at it, the number of Action Points spent). Thus it eliminates other factors that might influence attack speeds from consideration.