Hello !
I’m making a prototype of a simulation game in space, and I wanted to know if there was a negative impact of reducing the scale of all my games by 10.
My game will be design for mobile so I’m thinking a little about performance.
If I thought about that, it’s because certain maps will be quite large, with objects that will need to be seen from very far away while maintaining a movement/perspective effect. Leaving the possibility for the player to travel great distances.
The advantage of reducing the scale of everything by 10 would allow us to avoid using too large numbers, or having a camera that looks from very far away.
I could have distant and very imposing objects while having “normal” figures to manage.
Is it possible that this is having a negative impact on the game in other ways that I wouldn’t have thought of?
I’m not an expert in physics, materials, effects, light etc… or others…, so I don’t know if this could have a negative impact on anything?
Thanks for your help! I appreciate it very much
Have a great day! 
in the same way that having large numbers can reduce precision, so can having too small of numbers. you could do it and depending on detail your mileage may vary, but you may see vertexes not moving as smoothly. i would test it yourself though for your case, obviousness can be up to camera distance as well as other factors. i’m not sure if /10x is a big enough margin to matter
Thanks for helping me !
That’s the problem, is /10 enough to be considered something negative?
I’m doing some small tests but I don’t have the possibility or the time to push the test thoroughly only to realize in the end that I made a mistake in my reasoning much too late…
the test should be doable, just make a very simple scene with some models and /10 in scale. see if anything looks amiss and if not you’re probably fine. note though /10 might have precision problems for things like physics and transforms etc as well so proceed with caution
My feeling is that this wouldn’t really help to improve precision. You are basically just moving the decimal point. A floating point number is similar to having a number with for example 3 digits and a variable decimal point. So you can either have the numbers 1, 2, 3, … 100 or 0.1, 0.2, 0.3, … 10.0 or 0.01, 0.02, 0.03, … 1.00. Moving the decimal point doesn’t give you more distinct numbers. If you want to count to 100, you can’t have anything after the decimal point.
Usually, you’d use camera relative rendering to solve this issue:
https://docs.unity3d.com/Packages/com.unity.render-pipelines.high-definition@8.0/manual/Camera-Relative-Rendering.html
1 Like