Polycount too high or to low?

Polish_Wizard

Hello everyone, I hope you're all having a wonderful day! I've been immersed in modeling for my game, but I've hit a bit of a snag. This game is an action-adventure, akin to Tomb Raider and Uncharted, where players explore ancient locations, discover treasures, and engage in battles with enemies, among other exciting elements.

As I navigate through levels in the game, I find myself in need of creating some assets as you do, to fill in the empty spaces. However, I'm uncertain about the ideal poly limit for these assets. I'm contemplating whether it's advisable to maintain a moderate poly count—neither too high nor too low—given the advancements in modern technology and the capabilities of UE5. If anyone has insights or recommendations on the matter, I would greatly appreciate your input. Thank you!

The assets in question

Greek Warier Statue

Size/ 6m in hight

• Vertices count : 13,715
• Triangoles count : 27,715

The statues in this room play a crucial role in advancing to Act 2. Upon successfully completing a puzzle, the player will need the statues to move their hands holding ther soreds to, enabl access to an object securely locked behind a gate. The room itself spans dimensions of 35x35x25m, and within it, four additional statues await. These statues are strategically placed to assist the player in solving another intricate puzzle, paving the way for progression to the next stage of the adventure.

Alex_J

It seems reasonable to me, I don't see wasted geometry.

If you go through with a fine comb you might find some edge loops here or there that you could delete but I don't think that is going to pay you back meaningfully for the time.

I think rather than try to guess what is perfect, try to find the max and min. To find the max you can use really high resolution assets and make a big scene to test on your target hardware. Try to determine how much shit you can accommodate before performance becomes an issue.

To find the min, you reduce resolution of your models and textures as low as you can before the visual quality goes below what you wanted.

For a tiny team or solo developer I think staying much closer to the min is generally best. There are too many things to do, solving performance issues isn't where I'd want to focus too much time. But you just have to identify where you want to spend your time on per project basis.

edit: also be aware that the intent for nanite is that the high resolution mesh replaces normal maps. So if you ended up using a normal map (or several) for these statues, you would want to compare the cost of that compared to a high resolution mesh.

Epic put out some comparisons showing how nanite is supposed to be the better option. But I think you have to look at total cost of what that could mean for development, not just the the end user.

Polish_Wizard

Alex_J

Thank you so much, Alex, I truly appreciate it. I've been immersed in this project since 2022, and it's only been four months since I completed the research, scripting, and initial blocking out/planning of the level design. I've recently delved into the modeling phase, and while I'm eager to make progress, I want to ensure I don't make any decisions that might come back to haunt me later on. Your advice is spot on dwelling too much on potential issues as a solo developer at this stage might not be ideal. I'll keep pushing forward and address concerns as they arise. Thanks again for your support!