PDA

View Full Version : Animations-fix



Red Spot
08-23-2007, 16:07
{removed the old topic/info}
{updated again 28/8}

Ok the descr_skeleton file is now officially driving me nuts ...

Some changes seem to take effect others seem hardcoded, results are pretty tempermental.
Trying to change a halberd_sec. from being a hybride-2hswordman to being a hybride-2haxe doesnt seem to take effect, the men dont even change their charge in any visable way.
Removing the taunts gives me as result ... yep, taunting men ...

I'd like to believe I've made some improvements to the 2handed axe and sword in particular the axemen as they really came up short compared to the 2hsword and halberd_sec. but I'm actually doubting if there is a "real" difference so minimal it is.
Am currently stopping any tempering with the file as after like 4-ish days of doing small battles with diff. settings I feel I'm right where I was when I started ...


G

SigniferOne
08-23-2007, 18:23
Just fyi, the "slow" in all slow animations only refers to their running speed. All other animations are completely the same between the regular and slow version. You can confirm this yourself by looking at descr_skeleton.txt, and finding the regular animation, and then the slow variant below it.

Differences in combat could be more attributable to stats of the unit, or some sort of delay in attack that's specified in export_descr_unit.txt.

FactionHeir
08-23-2007, 18:46
Just fyi, the "slow" in all slow animations only refers to their running speed. All other animations are completely the same between the regular and slow version. You can confirm this yourself by looking at descr_skeleton.txt, and finding the regular animation, and then the slow variant below it.

I haven't verified that, but I think there may be some truth in both of you two's statements.



Differences in combat could be more attributable to stats of the unit, or some sort of delay in attack that's specified in export_descr_unit.txt.

That's a negative. All units have the same delay and even changing that number doesn't seem to significantly impact combat ability.


That said, I'll have a look at the file and play around a bit to check things out.

Red Spot
08-24-2007, 00:55
seen the file, never payed any attention to it, did now you mentioned it ... thanks! ;)

anyway I cant do anything but agree with you after seeing the file, interesting that test show somewhat diff. results ...


G

Red Spot
08-24-2007, 20:28
Seems I made a mistake, SigniferOne it seems you are right, its not so much the slow 2handed animations, I just quickly put them aside as bugged/slow and so did I do with the regular 2handed axe/sword animations, they dont really give other results, they just all suck ..:P

Though I still cant help feeling the slow sword (single handed) do own the regular 1handed swordmen, the battle seems pretty much even though more often than not the slow ones are the one with some men standing up after the battle ...


Somewhat different, I've been trying to understand the animations now I know where to find them (thanks again ;)) but I'm having a hard time figuring out the use of the 2nd animation, like with pikemen the mtw2_pike makes them hold their weapon, it controls the fighting/walking/defending animations, so what exactly does the mtw2_pike_primary do??
If changing the pikeman's "...._primary" animation it makes no visable diff. though giving some 2handed axe/sword/halberd like the pike_primary animation they "seem" to do slightly better vs cavalry, but there is no visable diff. nor any change in stats ...


G

KnightErrant
08-29-2007, 06:19
Hi, This one is an interesting one as regards animations
since several people have asked about it before. SigniferOne
is quite right that the "fast" and "slow" anim families only change
the run animation but otherwise inherit all other animations from
the base family. (C++ jargon there.) Did a quick look at the
2H_axe family. The regular run animation has this for the bone_pelvis
traversals during an animation cycle:


-0.0001573533
+0.1655251682
+0.3312083185
+0.5012405515
+0.6773928404
+0.8578948975
+1.0417801142
+1.2317030430
+1.4250081778
+1.6127607822
+1.7877503633
+1.9571866989
+2.1337628365
+2.3261723518
+2.5239360332

These are the z coordinate movements of bone_pelvis over 15 frames
at 0.05 seconds per frame. (z is out of the screen or the forward
movement of a model.)

For the "fast" run animation you have:


-0.0001573517
+0.2150593549
+0.4082388878
+0.6062413454
+0.8276299834
+1.0409616232
+1.2517681122
+1.4909939766
+1.7162835598
+1.9277546406
+2.1349267960
+2.3421463966
+2.5537605286
+2.7737166882
+3.0139348507


Making the assumption CA works in metric where possible
this is 2.5 meters per 0.7 seconds versus 3.01 meters per
0.7 seconds. For us hopelessly English unit users this
is roughly 8 miles per hour versus 10 mph. For reference,
us older joggers can do 8 mph without military gear for a few
miles. 10 mph is a 6 minute mile. There aren't that many
people that can do that for more than a mile or two and
that's without gear. (Elite Marathoners do 5 minute miles.)
I think CA's "fast" anim is probably all you could realistically
ask for.

Bwian
08-29-2007, 09:05
KE is correct... only the RUN animation changes for the fast version of an animation. If you look at the text in descr_skeleton, you can see the 'fast' and 'slow' variants of an animation.

First, they reference the base animation. So 'fast' knifeman has a parent 'knifeman'. This tells it to use an animation from the parent set if none is listed in the specific set.

In the 'fast knifeman' you only see a basepose and run animation defined. This is used in place of the parent animation. The basepose is needed to define a startpoint, but that seems to be all.

Same is true for fast mace, sword etc.

When it comes to actual combat, though, I am not at all convinced that the displayed animations actually have any real bearing on combat. Some of the weapon animations have an 'effect' animation that is applied to the opposing figure. This seems to take it out of the battle until the animation finishes. I spotted this when I made a 2 handed skeleton that was 4 times scaled... certain 'kill' anims stretched the corpses to 4 times height! Not all anim sets have this. and it's primary purpose seems to be to give more variety and response to the attacks. It may have an underlying effect on the battle calculatons.

Also...I am not sure how the contact spheres and range of engagement stuff works. That can also have a big impact on combat effectiveness. The first strike may not be applying to the weapon it should.

Red Spot
08-29-2007, 09:13
thats just it, the animation doesnt give much changed results, take for instance the a slow_sword and a fast_sword they'll do about as good, you cant really say 1 is better than the other, but in the end slow_sword "always" slaughters fast_sword IF AI-vs-AI ...

There is more to it than just animations, the game makes infantry IF they are faster runners back out of combat and re-charge the slower men, but this only works in such in way they make sure they will lose the fight ..

I've went from disliking the slow animations (biased about it) to liking it in favour of the AI, the AI knows a lot better to handle slow runners than it does handle fast runners.

Anyway, thanks for the replies guys ..;)


G