The "What Are You Working On?" Thread

Also regarding edge chamfers for most panel gaps, 'best' (in terms of balance between amount of tris and simpleness to create) way is with a simple support loop.

This works for almost all standard panels, only ones with characteristically large gaps need a proper chamfer imo.

And I say 'best' because you could technically do better with weighted normals etc but this works fine and it what Kunos use for almost all of their models over the years.

upload_2019-7-14_22-53-40.png


Works well for any smaller (less than 1cm-ish) radii, whether they are convex or concave.

upload_2019-7-14_22-56-40.png


upload_2019-7-14_22-59-28.png
 
Depends how you use it. Different modellers use difference styles, but I think most of us do use it in the early stages to create the initial general shape, but then collapse and optimise/add details a bit later.

Here is an example of one of mine, in the early stage - this is the pre-subsurf model:
unknown.png


This is it a bit later, after subsurf applied and optimised/detailed manually added:
unknown.png

That's still something I really struggle with, coming from a parametric CAD modeling background. That "point of no return" terrifies me still :roflmao:

That being said, @Pessio is already surpassing my capabilities by miles when it comes to poly modeling. Nice stuff, mate!
 
That "point of no return" terrifies me still :roflmao:
Yeah it did for me too in the beginning, but I'm used to just making copies (either of the object or even the whole file) before moving forward in any drastic fashion. Countless times I've had to hop back to an older copy to fetch a pre-'whatever'd object to either fix or just use in another way.

But i do think its necessary if you want to make use of subsurf in your workflow, otherwise its just not possible to build in the right level of detail whilst maintaining a decent tri count.
 
coool! i'm working with same thing right now! is a transparent thing, with color on "details"?
Not really sure what you mean, transparent?
The scratches etc on the floor are just painted on the main diffuse, and also on the alpha channel to block the 'detail' texture, leaving the scratches white, regardless of the detail texture colour.

Here is an example of the diffuse:
upload_2019-7-16_14-19-10.png

And the alpha channel for the diffuse:
upload_2019-7-16_14-19-37.png


And then the meta detail is just something like this, which multiplies over diffuse (for areas black on the alpha channel):
upload_2019-7-16_14-20-35.png
 

Attachments

  • upload_2019-7-16_14-19-25.png
    upload_2019-7-16_14-19-25.png
    832.8 KB · Views: 152
The car is too fast, in terms of lap time and straight line speed. Also there are loads of alternative gearbox ratios in the 1964 Austin/Morris Mini Cooper S 1275cc homologation paper which arent there ingame.

I only tested a couple of laps at Monza Junior and did a 1:00.1 in the Mini and similar top speeds to the other cars in class. I've been testing those cars at multiple tracks for something else so here are there-or-thereabouts lap times at that track (which, while not a top speed track, probably doesnt favour the Mini.)

u2l.jpg


Why not get one of the experienced AC physics guys to help out if they are interested? As you said yourself, you wanted to make the most realistic Mini in any racing game.
 
Thanks.
In reality, i made phisics too be good on paper for power, aero and suspensions. Then adjusted the performance with tyres.
I did that after looking a good amount of nick swift videos, checking his laptimes, and also checking these:

http://www.mini7.co.uk/laprecords.php

searching to balance my mini between SEVEN and MIGLIA class.
So my mini match good times, really close to reality ( a little faster on brands than nick swift best, but in race )

Btw i think gta in game is less less speedy than IRL, but i suppose is related to tyres.
 

Latest News

Are you buying car setups?

  • Yes

  • No


Results are only viewable after voting.
Back
Top