Cars Ford P68 (F3L) W.I.P.

Ever since I was a kid, I always thought that the GT40 was the most beautiful racecar ever, and the P68 a close second.
Well, everybody knows of the GT40 - and it has been in so many games that I lost count.
I've never seen the P68 in any game.

So, with my very-basic-very-noob-just-starting-3D skills, I want to fix that.
(I've modeled half a jet fighter in Maya once, and 1/3 of a Dodge Charger in Max, before giving up.)
Thas was some time ago - I think I'm more patient now. And trying a different program that (so far) seems easier to get into.

Here's the humble begiinings - I hope you guys are patient, because I'm learning as I go.
P68.jpg
P68_2.jpg
 
Last edited:
really cool idea. Do you use high subsurface smoothing on the model? If I remember correctly the current "limit" Kunos proposes is 200k tris overall for LOD_A
 
The screenshot I have taken are with subdivision surfaces applied. You don't want to see the cage model ;)
There's still quite some distance to 200k so I hope I can manage it.

It's a lot of back and forth though. (That's what she said!)
 
I can see some improvements. I'm learning how to draw curves in photoshop, so I could show :D

But it looks like you really can do it !

Go for tires and rims. I have done XJ13 rims first week (they could use some more love :D)
 
Yeah, there's plenty room for improvements. I started on the headlights, and the rims. It's some tricky ****, that's for sure - all those sexy curves, and trying to avoid pinching. So progress is a bit slow today.
 
Yeah, I have been working with blender, and even having perfect tutorial it took ~8hours to do rim and tire.

Later wheels and various details such as headlights makes it easier to recreate shape, only if those details are not very depending on body mesh.

But you don't have too much of those objects on your car. It is actually good, because besides interior, your only real challenge will be precision. I'm sure you can get it right, lights, seams, windscreens and air intakes should be good yard sticks. Headlights probably are going to be very tricky.

Are you playing with different FOVs to match reference images ?

I'm modeling Mercedes W125 now, and I think I have applied subdivision too early. I lacked patience, now it is harder :D And it is always very hard to decide when shape of the body is good enough to walk off, because, you'll have to once before (obviously) applying mirroring, and second after completing UV map.
 
Are you playing with different FOVs to match reference images ?

I'm modeling Mercedes W125 now, and I think I have applied subdivision too early. I lacked patience, now it is harder :D And it is always very hard to decide when shape of the body is good enough to walk off, because, you'll have to once before (obviously) applying mirroring, and second after completing UV map.

I haven't played with reference images in the viewport yet. So far I have managed with images on a second screen. (And blueprints - but I think I'm about done with those, they lack detail, and not very accurate.)

Yeah, I heard that the subdivision should wait until you absolutely can't go further with the current detail.

I'm looking forward to seeing your Mercedes... Funny I thought about making that one, too. :D
 
Yeah, blueprints often are "blueprints". Even if they would be exact they still wouldn't be a blueprints. Blueprint in my head is something what would have not only precise top, side, front and rear views, an a lot of ortho cuts from the front and maybe side - a lot. Also a lot of dimensions, especially in th cuts.

I use reference images like that too. Unless they are from very front, very side, very top, very rear, and fits as a background image.

Oh I remember now... I used to do a lot of renders carefully trying to match photos, it helped to get acknowledged with shape a lot. It is challenging not to ruin shape look from other angles when fixing one, and for so curvy cars 1 vertex off by 1 mm and it is visible.
 
The thing with subdivision modeling is, you line up the middle of faces to the surface you want to create and it rounds off the edges. When you're just modeling it straight, you line up the vertices/edges to the surface you want to create. So you can't switch from one to the other without changing the overall shape of the model.

Photomatching will get you good proportions, it can't produce an accurate end result the way laserscans or whatever can, and for most photos you shouldn't expect a perfect match (lenses all have slightly different distortion that the viewport usually can't match) but it'll at least look right. For example if you don't have a real-world measurement of the exact width of some trim piece, you can at least get it visually the same.
 
Blueprint in my head is something what would have not only precise top, side, front and rear views, an a lot of ortho cuts from the front and maybe side - a lot. Also a lot of dimensions, especially in th cuts.
Yeah, but that's the problem (challenge?) with old cars. Most blueprints are wonky and handdrawn.

Maybe I should try to contact that guy that built a replica. I'm sure he has more details.
Here's a link to that forum, if anyone is interested - it's a fascinating read.

http://www.gt40s.com/forum/wings-wheels-keels/12585-ford-f3l-david-piper.html
 
Nice, looks like a lot of information there ! :)

Is it same one ? Looks rather exact, I wonder if it would still look the same after ~175 hours of precision body modeling :D Looks very honest. Probably would.


We had contacted Neville, owner of Building The Legend project. Contacting them can be helpful, but it depends on luck a lot. Imagine somebody asking for valuable data which you have been collecting for years.

I think you can get body look right anyway, it is all about adjusting curves till it is there, you eventually will be simply getting closer. Blueprints aren't necessary, when photos from various angles exits. All you have to do is to do what photo scanning software does, just with your eyes and mind :D
 
Crazy ! The way you cope with your mesh is better than me. Looks very clean. If you'll keep subdivision till you're there where you'll be happy, then it is going to be really nice mesh ! Is it really your first modeling ? :D
 
Yeah - well, as I said I did half a plane and a bit of a Challenger before giving up.
I've had plenty of time looking at UV's though - worked as 2D artist/ texturer for years, so I've got a bit of concept of flow and all that.

Edit: Thanks for the kind words by the way. So it seems I'm on the right track.
 
Last edited:
Back
Top