Hi-poly vs Low-poly

[yoyohobo665]

First of all, this poll is intended soley for people who make 3D models for mods for Oblivion, but the topis is open to all comments.

 

-Definitions-

 

Hi-poly: A large usage of traingles used to create an extremely smooth model.

 

Low-poly: A lesser usage of said triangles to create a basic shape for a 3D model.

 

-Examples-

 

Hi-poly: New Frostmourne Sword -37,579 triangles.

 

Low-poly: Adonnays Elven Weaponry - average of 2K

 

I used swords for the examples. The average sword in vanilla Oblivion consists of 1.5K trainagles (excluding the sheath). Both examples have very beautiful weapons- the difference? Adonnay stuck to the rules that are needed to be followed by game artists such as those at Bethesda that actually make swords for Oblivion. Where as jojjo bent these rules to make a VERY nice sword. I don't know the reasoning behind either of their decisions (low or hi-poly) and I'm not here to tell which one is better. I'm here to simply ask what my fellow 3D modelers choose, and their reasoning behind their choices. I for one try to keep my models as low a poly count as possible, because I intend to become a game artist when I graduate from college. Making models for Oblivion mods is great practice. Anyway, along with the poll I ask these questions-

 

1.) Why do you make hi-poly/low-poly models?

 

2.) Which do you think should be more widely used?

 

P.S. I understand that the Frostmourne Sword is much more detailed in comparison to Adonnay's swords, but the point of the topic isn't affected by that because they were just random examples, and not the epitome of their entire classes.

[LHammonds]

Generally, I would say low-poly for 99% of the stuff in gaming. Textures, Normal Maps, Vertex shading and mesh smoothing can add a LOT of detail to a basic model.

 

Swords however are a bit of an exception since they are onscreen and in close up view most of the time.

 

The blade and crossguard might need a bit more detail than normal but the pommel and grip can be very low-poly and not be noticeable since they usually have to be in odd positions to see them.

 

It is always best to start your model off as a low-poly model and increase the complexity as desired. The biggest reasons to increase geometry are smoothing curves and places that move due to animation (shoulders, elbows, knees and mouths for example).

 

I've heard it said that if the geometry you are adding it not noticeable in a shadow, then you are adding too much geometry and should use textures for that level of detail.

 

LHammonds

[DogtoothCG]

huzzah for low poly!!! If you smooth to soon and try to edit it then, youll find yourself in a world of fun (sarcasm)but higher poly models are easier to animate if you rig them sligthely incorrectly

[ihateregisteringeverywhere]

You can also create a high poly mesh with plenty of detail, then a low poly model of the same thing and project the high poly into a normal map for the low poly. So doing both will leave you with a low polymodel with a high level of detail :P.

 

Huzzah for normal maps!