Default player animations

#1 If the game crashes, my guess is that it can't find the root animation. I have heard that that will cause the problem. Have you set up a .cs file with the TSShape constructor datablock? If not, it's easy if you're using the default animations, just go into the player.cs file (in data/shapes/player), and change "player.dts" to whatever your character's dts is called. I think that should work...

#2 Tried using the default naming convention aready (i.e my replacement model was named player.dts also). As such i dont need to change the *.cs file .

As regards the root animation, you dont need to export your own root animation, do you?
..i have tried with and without doing this already...but some clarification would be good..
Do i have to export a root animation with the character in the exact same default pose? ..or do i need to at all as the dts file will just use the default root pose already there?

#3 Why dont you just use the default player, delete the mesh(s) and merge yours
in. Then apply sking modifier add Detail level, and export.

You dont have to export any animations 'cause the DSQ's are already made.
Dont forget the unmessdts or you wont see your Character.

At first, all I wanted to do was just see my character animate with the default
animations. But now Im at the stage where thats not enough and I need to
export all fresh sequences myself.

The great thing about mergeing with the defauly player, is that the default already works
so it gives you good guidelines to work with in.

Surge

#4 Tried using the default naming convention aready (i.e my replacement model was named player.dts also). As such i dont need to change the *.cs file .

As regards the root animation, you dont need to export your own root animation, do you?
..i have tried with and without doing this already...but some clarification would be good..
Do i have to export a root animation with the character in the exact same default pose? ..or do i need to at all as the dts file will just use the default root pose already there?

#5 (double post??? ... soz)

...i tried that... it didnt work to well.. partially because the player models are being made by other people and hence are in fairly random poses.. it would mean pushing the vertex around, meaning i would probably mess up the look o the model... way to many models to have the modellers redo/repose them

i have tried instead using the exact naming convention used in the default player file... i even got to the point of having all anims play in TST, but not ingame..yet i can export my own animations ok, and have gotten a few player models working ingame with my own animations... thing is my animations skills are not upto it yet and i wanted to used the default player animations as mockups till i get skills, so as not to slow the mission setup process etcetera

does it matter what pose the actual dts file i export is in? .. i still feel something related to the root pose may be the culprit


i have yet to use unmessdts as up to now i have only used 1 detail level (the main mesh) for the purposes of quick testing, no autodetail...
this has never seemed to be problem as the models exported with my own anims worked fine and ingame....

as a sidenote, can u add detail levels manually to your player model using snapshots or must u use autodetail?

To Surge: i take it you have been successful in using the default player animations ingame yourself? ..i ask to make sure you are telling me first hand information as i have spent a fair time on this problem already and want to dispel red herrings, or chinese whispers

#6 If you want your shape to load the default SDK player_*.dsq sequences and look/function proper; then your player.dts shape and it's rig MUST be in the default rifle holding 'root' pose, and have the properly named Nodes in the exact location and orientation/rotation. Otherwise, any shape can load the sequences[had a HORSE.dts with proper names...;), and it assumed a human pose, veeeeerrry creepy] with proper naming convention, the pose it assumes will be based off the 'root' pose of the dts shape[I think]...good luck dealing with mulitiple shapes in custom poses.

I've gotten my own shapes to load the defaults , 'players' exported from both Max and Milkshape. I've even produced 'player' rigs, in Milkshape3D, with the maximum amount of mountPoints allowed by default code[8]; and it loaded all the default sequences quite nicely. It can be done, it's a bit tricky, but it will work.

Rex

...also something I've noticed and I 'think' it's a 'quirk' of the default engine. When at the 'root' sequence/idle and holding a scripted weapon in a armThread 'look*' pose, player looks great holding it's shiny weapon. Now when you start to move, the 'player' transitions from the 'root' sequence to the 'run' sequence and while running; the 'player' is now holding the shiny weapon in the default rifle pose. I can't seem to get an answer if it's possible to get the 'player', 'running' with ANY other armThread 'look*' sequence blended as to make it look better for pistol and small hand-held items/weapons...??? I've noticed this as I'm putting together a lot of hand held items, pistols, etc....and it's starting to look odd!?!

#7 ..i think you have answered this already rex.. but just to clear it in my own mind...
i set the root pose in the DTS file and not in ROOT.DSQ?

i am using these as mockups so aslong as they function and look reasonably humanoid im ok, especially at this stage of trying.. so do i HAVE to be 100% exact with placement?.. i thought i could just rig my model to a skeleton with the same name, nodes and structure and i'd be ok...

sorry if some of these questions sound dumb..and thx again

#8 If you want your custom rig to exactly follow the stored .dsq data, yes, the exact placement/orientation is necessary....it's the 'law', lol.....!


Not 'dumb' at all...;). I get confused a little myself sometimes. The 'root' pose, to me, is the pose[keyframe] the dts shape is in in the Scene Root, or if your package permits, the Zero Frame. This is what the engine examines to see if a Node has had any transform data applied to it. This is not the 'skinning' pose[ala DaVinci, or similar], but after the skinning process and how you want your shape referenced in the game.

The 'Root' sequence, is based upon the 'root' pose/frame and the rig passes thru this pose at some point[beginning/end] to allow the engine to syn the animations. This is usually a cyclic animation and most if not all Nodes are included an animated, a few exceptions in the defaults.

So, everything seems to be based upon this one key pose/frame, anything different will have the stored values in the dsq files applied to this 'reference' position and that's why players assume inhuman poses when different posed rigs load the stored values.

I hope I made it a bit clearer and please, someone more knowledgeable about the TGE animation process correct me if I'm wrong.