Quote: "@raven why in the hell would you animate a model before you skin(texture) it just makes no sence, as 1st off you have no guarantee your uv map is perfect maybe you could use some checkered grid but there could be some hard to see problems, the only advantage i can see in your pipeline is if you are working with a group of people ie a modeling team skinning team etc etc, then the workflow becomes more quicker but if your a one man modeler then your pipeline is inneficient unless of cource you want the untextured model ingame to test it or something but i believe it is better to have a finished model in a game rather than a unfinished one"
The pipeline I've used for years has served me very well to save *alot* of time both solo and in a team.
A Checker pattern will *always* show you where your texture stretches, it doesn't matter what you do.. as Checkered images are simply Grids with alternating colours. From this you can quickly see where warping occurs.
I have often found I'll map something, and then texture it... only to find out that while animating it, the map stretches in places that I do not wish it to.. as such I have to remap it to suit, which won't affect the texture is the remapping is only minor; but often when you move something this means space must be compensated and you end up rearranging the entire map in order for everything to fit with the maximum amount of space.
If this occurs then you also have to completely remake the Texture, or at the very least rearrange it in such a way that can cause quality loss. Why risk hours of pixel work for a matter of seeing your model looking good while animating it? Seems totally irrational to me.
Quote: "you are delusioning man, below 5000 polys nowdays is low poly, sega saturn times are long gone."
Really? .. lets have a look at the titles on the market shall we?
Half-Life 2:
Low Mesh = 1,500
Medium Mesh = 3,000
High Mesh = 6,500
That's for the people models, most systems except pretty expensive high-end ones can only handle the Medium Range... while an X-Box version of HL2 is due... the models will all be running on Low Mesh.
Let's go for another example, something purely console based. How about a title I have access to right now.
Starfox Adventures, Fox McCloud is 3,500 polygons; most of the Dinosaurs in each area are between 2,500-4,000 polygons.
Oh I know here's a good one. Dead or Alive 3 for the X-Box, this machine can really push the polygons extremely; considering it is purely environments onf 2 fighters at any given time really.
Characters generally weight in at just under 6,500 polygons.
The fact is, 90% of the people on this site right now and a majority of other forums have no sodding clue what polycounts are viable for games.
When I'm making a game mesh... I consider 8,000 to be High Polygon, that is the limit for a general character, medium would be around 4,000ish and Low would be 2,000ish.
If you try and place a group of 4-5,000 polygon characters in a game like Starfox Adventures, in those environments with heavy use of shaders you are going to kill the speed quicker than a rabbit strapped to a nuclear bomb!
I mean Jesus what the hell do you think the hardware you using is actually capable of... Considering that even Unreal3, which runs at 15-20fps on STATE OF THE ART HARDWARE; only uses models that are 10,000 polygons.
You cannot think of just the models, but everything in the environments around them; Hell even Resident Evil only uses 8,500 polygons models and that uses almost exclusively 2D backgrounds!
Seriously, get a clue. Cause right now, you honestly don't have one.