The W.I.P. Thread

Status
Not open for further replies.
Yep, looks like they photo-match then spline cage, then err, model over it :D

Very nice work.

But it's amazing how much work one car is, never mind many to make up a whole race series!

Good luck to them getting it all in GTR2. Even if the mesh density isn't cranked right up, as long as they use the quality shaders and so on then the visual look should still be very very nice :D

Dave
 
60k polys with a low-poly interior for the exterior view appears to be what the team at AC are using for their cars so far...

I'd try keep the main body about 25k in that case since it's only a small car too.

Looking good so far though :D

Dave
 
Yep, looks like they photo-match then spline cage, then err, model over it :D

Very nice work.

But it's amazing how much work one car is, never mind many to make up a whole race series!

Dave

How to make it sound nice and easy! :D I believe it's one dude, by the way. Plus, check out some of the early interior shots:

Autobacs-Garaiya-wip-002c.jpg


That's not spline work as far as I can tell, although I'd love to know it is done! I'm constantly in awe of the talent of some of the modders and modelers out there, just magnificent. Sincerely hope we see them in AC :thumbsup:
 
Bram Hengeveld
I'm glad you like it.

@ Mr Whippy
I'm one man band when comes to modeling those machines
And the workflow you describe is correct.
Thank you very much for the comment,I really appreciate it.

@ BanjoMaster
This is an older interior that was done by looking at pictures.
Same with shiden.

I used camera matching to determine supra's interior though
as well as some other tricks in photoshop.
Thanks a lot for the comments,and for featuring my work in this thread

I would love to see them in AC as well.
 
Yep, looks like they photo-match then spline cage, then err, model over it :D

Very nice work.

But it's amazing how much work one car is, never mind many to make up a whole race series!

Good luck to them getting it all in GTR2. Even if the mesh density isn't cranked right up, as long as they use the quality shaders and so on then the visual look should still be very very nice :D

Dave

Also meant to say, that Shadow's meshes (supra at least because it's the furthest developed) look to be of a comparable standard to the kunos models, to my eyes at least. If you look at the kunos models, they seem quite efficient. It's the texturing and lighting engines that really make them look awesome, and that should be applicable to all modders.
 
Those Kunos models have very tight edge loops at panel edges though. In some shots they look sharper than others so it's almost like they do a super-high polygon version for close up shots, but then suffer with stretched normals at panel edges in lower poly versions of the same kinda car panels further away.

As they say in the trade, wireframes! Until we see them we can't really say too much.

Basically the imagery I'm seeing feels like Forza 4 stuff with a very high density model for super-close up screenies (but possibly not seen much in actual game use!?)



In any case, Shadow's work is very nice, and even without the (you could say pointless) super-high def meshes for photo-mode type shots, they will look great and play great in AC, so I hope we see them there!


Cheers Shadow on confirming the workflow. Do you use ImageModeler to set up the initial cameras/imagery, and then move over to a proper 3D app for splining?
I currently do that, and then I lay out some splines in one camera view, move to another camera view and then move the spline knots *only* in the axis direction of the camera which I drew the initial splines from. This seems to get me close enough, and then I refine from other camera views to get it almost perfect.

A slow and tiring workflow... I can't imagine many commercial ventures will pay an artist to do that when their economies of scale can invest in a laser scanner for example, or pay for models already made, or whatever else.

Obviously in the worst case it works, and it works well... just it's not quick! Maybe I'm just slow at it haha...? I am a perfectionist though so maybe that is my problem.

Dave
 
I do the matching manually in my 3d app.
Due to my approach, Supra (the most further progressed car and my favorite machine) might end up being heavier than any game model so far.

I want GT500 models to first be complete,then proper lods will add efficiency.
so the "perfect" model will be only for time attack. (and years later with more proccessing power might end up on the grid XD

GT300 models will be really modest. they will not have those crazy details.
The reasons are: lack of sufficient data,time/money.

Regarding camera matching,More or less I do the same as you.
Might take me from half a day to 2 max. depends on the pictures and the model.I do not find it tiring though.

There are other factors that increase the time needed
For Supra (and others)there isn't much info available freely on the net. I really had to study "race car anatomy",
bought high res pictures and a scale model.correct stuff etc...

It is still a slow process,but very rewarding.

About laser scanning,I don't like it.
I want to see a model and know that someone cared for it and also see the way he approached it.

Scanning mona liza will not make you da vinci.It is just about bragging rights
Last but not least you may check the blog.I've been working on the internal structure of the car^^
 
Just a small update from me. The brochure that I said about earlier has now grown to nearly 50 pages and the look has improved dramatically. I have several magazines with road tests of the Rover Cooper in which will be scanned and put into the brochure, as well as two Rover brochures - yep, that's right, a brochure in a brochure!

Hopefully when the tech demo is released I will be able to start looking at the physics of the car and getting used to AC's architecture. Fingers crossed... :)

EDIT: Also I plan on including some semi-slick track day compound tyres for those of you who might want to race this car.
 
I do the matching manually in my 3d app.

Regarding camera matching,More or less I do the same as you.
Might take me from half a day to 2 max. depends on the pictures and the model.I do not find it tiring though.

There are other factors that increase the time needed
For Supra (and others)there isn't much info available freely on the net. I really had to study "race car anatomy",
bought high res pictures and a scale model.correct stuff etc...

It is still a slow process,but very rewarding.

Are you using 3DS Max internal matching?

Do you correct for camera distortions or just choose imagery carefully?

Agree on the other factors. I'm building a Eurofighter Typhoon here and it's a damn tough aircraft to get right.

I have my own photos, photos off the web, and a scale model, and I'm still struggling.

But I like it because it's a challenge to get right. Just having a CAD model would kinda take the fun out of the modelling challenge.

But from a commercial point of view it's not really an amazing skill since a lot of the time for decent projects you will get resources to help you along, and if it's not a decent project, then the quality doesn't have to be that good and budgets won't justify the passionate model making process.
So it'd be rare I guess to do the work this way commercially. Still a great personal project/exercise though imo!


Amazing and inspirational work :D

Dave
 
Are you using 3DS Max internal matching?
Do you correct for camera distortions or just choose imagery carefully?

I use softimage,and try to choose the best pictures
By camera distortion you mean lens distortion near the edges of the pic?

Agree on the other factors. I'm building a Eurofighter Typhoon here and it's a damn tough aircraft to get right.

Sounds like a lot of work,good luck with it :)



I have my own photos, photos off the web, and a scale model, and I'm still struggling.

But I like it because it's a challenge to get right. Just having a CAD model would kinda take the fun out of the modelling challenge.


That was exactly my point in the other thread.Its always more rewarding to know that you can build your favorite machines on your own.

But from a commercial point of view it's not really an amazing skill since a lot of the time for decent projects you will get resources to help you along, and if it's not a decent project, then the quality doesn't have to be that good and budgets won't justify the passionate model making process.
So it'd be rare I guess to do the work this way commercially. Still a great personal project/exercise though imo!

So far I had to use matching in every commercial work I did so far.
sometimes even search for better pictures myself.
I'm used to use that technique a lot so i can complete a cage within a day most of the times.
FOr me its really worth it.
I have also seen others to use it like simbin,I think its a very big advantage to have matching
in your skillset.

Amazing and inspirational work

Thanks a lot,I really appreciate it :)
 
Status
Not open for further replies.

Latest News

Are you buying car setups?

  • Yes

  • No


Results are only viewable after voting.
Back
Top