Hi again, I’m having some troubles with the fbx size.
I don’t understand why when I import and fbx model that is 10mb size when it upload the .json weights over 70mb or 90mb. There is some kind of attributes that I’m not taking into account?
I’m getting a little bit conffuse about the sizes and the optimization maybe you can give me a hand.
When I publish my project into playcanvas, it says “40mb”, but when I test it with Chrome developer tool, it says “31mb” transfered and “159mb” resources, where does this diferente sizes come from? Which one should I take care of?
Also this is kinda weird, when I load te developer tool, it shows me that some assets that are NOT in the scene, but are in the project ( for testing porpuse I’ve upload differntes fbx of the same asset ) are loading, is it an error or the game always load all the asset beside they are in the scene or not? It also happens whit enable and disabled assets.
And one more question… It would be best practice to load 1 mesh of the windows ( for example ) and copy and relocated it in all the windows of the building to get better load times? Or it’s just better to have all my windows collapsed into one big mesh? Because relocate all windows in the PlayCanvas is gonna be slower. I suppouse that with furniture this is the way to go, is it?
Yes, that’s Playcanvas having gzip compression enabled by default, and reducing drastically the size of your app. All builds hosted on the Playcanvas server have that.
Yes, all assets available in your project that have the preload flag enabled will be downloaded during loading, even if they aren’t used in scene. So just disable the preload flag if you aren’t using them at all.
Yes, in terms of performance and load time, it’s best to have 1 model asset of a window and reuse it in your scene. If that’s not possible in your art creation pipeline, then using a single model it’s still an option just beware of the increased loadtime.
If you go for the first option, one single window model and multiple entities, make sure to enable batching to get the best runtime performance: