• 🏆 Texturing Contest #33 is OPEN! Contestants must re-texture a SD unit model found in-game (Warcraft 3 Classic), recreating the unit into a peaceful NPC version. 🔗Click here to enter!
  • It's time for the first HD Modeling Contest of 2024. Join the theme discussion for Hive's HD Modeling Contest #6! Click here to post your idea!

REFORGED MODELING- PROBLEMS TO SELECT HD MODELS.

Status
Not open for further replies.
Level 4
Joined
May 19, 2020
Messages
319
I have two models in my art library with cursor detection issues on them, which makes it difficult to make it selectable (whether in WE or in-game) and when at some point we finally manage to select it, it doesn't show responses to actions of sound, whether from "What", "YesAttack", "Pissed"... What was or was lost in the model?

The models are these:
or

Note: I noticed that if I add any custom portrait to the model, the unit audios will work, but why are there other fully custom models that don't need a new portrait to work the voice? There are numerous here at Hive running the unit's audio with just the custom model's protrait camera, without the need for a "_portrait" file renamed to the model. That sounds pretty weird.
In this way the problems of voices are resolved, but the difficulty of selection still remains with them.
Ps. I have reviewed your .MDL codes, but so far no solution.
 
Last edited:
Level 29
Joined
Jul 29, 2007
Messages
5,174
I had discussions with Retera about what exactly is used for selections. It was a long time ago, but if I remember correctly it was the collision shapes for units, while buildings have some sort of pixel-accurate-but-not-really selections (e.g. try clicking around the branches and leaves of a Tree of Life).

As you noticed, it is not required to have an external portrait model - if your model has a camera and portrait animations (e.g. "Portrait" and "Portrait Talk"), the game will use it as a portrait directly.
Having external portraits is more of an optimization because they can be much simplified versions of the model, and thus require less resources to render in-game.
In reality in a game that regularly renders hundreds to thousands of models and tens to hundreds of units, I am not really sure why this was even implemented, but maybe it made a tiny difference in 2002 🤷‍♀️
The HD portraits are another story - they are all 1:1 copies of their models with portrait animations instead, so every portrait actually just adds more required memory and splits the rendering while offering no benefits (just one of the smart decisions that came with Reforged...)

Sounds not working are also an HD thing - I know that you must point to a valid facefx file and that your model must have the bones to support it, or something along those lines. You can ask in the RMS discord where it was discussed and answered already in the past.

/Edit

Replied by Bogdan - the models need a facefx chunk regardless if it points to a valid facefx file, and this is only for soundsets that have facefx data, while soundsets without it will work on any model.
 
Last edited:
Level 4
Joined
May 19, 2020
Messages
319
So, as for the sounds of the models I was referring to, the sound actions of the portrait work only by the personalized model itself, we would have these, for example:
They don't need an isolated “_portrait” file, they use the same default Portrait_Camera existing in the model, they don't have the default portrait animations (Talk 1, Talk 2…) added to the model, all works normally. Maybe my edits have changed something regarding FaceFX and Bones (as you stated above), maybe that's what ties up the audio part of the unit… but that's okay, that's not a big issue to be “fixed” . Although it's important to know!
In HD Portraits, there are lighting particles that differentiate them from the standard models, in addition to the animations themselves. But nothing that prevents us from using a particle point with predetermined light effect, being generated only in the intervals of portrait animations.. Lol

The biggest problem so far is the difficult model selection, it can be a silly solution inherent only to the Reforged models or some undue alteration by the RMS itself, going unnoticed.
I can even redo it from scratch. And by the way, I usually transform to SD because I have more experience and ease in handling animations using MdlVis and then return to HD using RMS, to then do the imports that generate combinations of different Geosets and Textures, in addition to Animation transformation between different HD models, there must be some missing link in that, maybe...
By testing each of the steps I use, I can decipher at what point the models suffer this reasonable loss in selection, and then switch to other methods thereafter. But I only intended to resolve this selection issue directly within this modified model or try to unravel what would cause it. For the SD format there is no such conflict, which in general, just confer the extensions or the collision boxes to resolve.

EDIT: Among everything I've read about “model selections”, I haven't identified any weirdness in the .MDL codes. I've tested the different Collision Shapes (whether boxes or spheres), inserted SD Model Collisions (is curious thing is that collisions for HD don't seem to have much effect). I changed some Bounds radius, consulted the extents limits, even checked if it doesn't have any relation with some main or origin attach nodes… everything seems to be in agreement, even “recalculating the extents” in the RMS, there seem to be no positive effects.
 
Status
Not open for further replies.
Top