Interesting MGS4 dev techniques

[Clayonite]

If youre interested in 3D, or games development in general, you might find this interesting.
I know a lot of Japanese developers prefer XSI over Max and Maya, and Capcoms one of them [unless theyve changed].
It seems pretty good anyway.

http://www.softimage.com/products/xsi/customer_stories/metal_gear_4/default.aspx

[cewolf]

wow ... that hearts .. A LOT

here is a Q

Bonneville T100 is a motorcycle made by the illustrious British manufacturer, Triumph. It also appears in the game. To create the data for this motorcycle, the team performed modeling using calibration (image-based modeling).

The team attached measurement markers on a motorcycle that they borrowed from Triumph. They took 16 photographs at 45-degree intervals in a concentric pattern. At the same time, the sound team recorded live-action engine, switch and starter motor sounds.



they did take 16 pic and the they used the images as reference to build the model
how is it done exactly ... and what are the measurement markers that they have used , is it the small red dots placed on the bic
http://www.softimage.com/products/xsi/customer_stories/metal_gear_4/images/tri_im_01.jpg
http://www.softimage.com/products/xsi/customer_stories/metal_gear_4/images/tri_im_02.jpg
and does the images has to be taken at 45 deg

pleaaaase answer me ....details .... details .. I WANT DETAILS , I AM DIEING HERE AAAAAAAA
and by details i mean step by boring step

thanks in advance

zico out

[seth ranson]

I don t know for sure of course, but I m quite sure these dots were used for the creation of the model.
You can see the wireframe has the same dots at the same places.
For the 45ø: As they wrote it I guess that means they took a photo, moved by 45ø then took a picture again. You d only get 8 pictures like this, but I then I think the bottom part could be done like this again, resulting in 16 pics. So the 45ø aren t really necessary I guess, rather a result of how many times you take a picture. If they took 32 pictures it would have been 22.5ø. Could be, however, that I misunderstood that and the 45ø are between the camera view and the ground, which would be useful, because you get more accurate information than from photos that s taken from almost above (if you imagine pictures taken from 89ø , they would hardly be different).
I hope this keeps you alive until Clayonite comes here and gives you a proper explanation.
A step by step guide is probably difficult, because there are probably small variations depending on who does the modeling, so you can t really tell how exactly someone did it, unless you re him. (uh, that sounds strange)

What I miss in this arcticle is that nothing s mentioned about the sounds of the game. Though it would have to be a lot longer then I guess. And it would have nothing to do with softimage... (and it s supposed to be advertisement for them after all...)

[cewolf]

thanks for your explanation

but i figured that already ... what i wanted to know is the setup , the real life camera setup and how
to apply it in the 3d world ... the key for this is to match the virtual cameras with the real cameras
if not done correctly , the result will be bad , very bad , very very bad , very very very bad , very very ( ok, we got it ..stop it )
also , how to mach the in images in the software and make sure that the object has the same size in all the of the pictures
is there a specified camera lens to be used ( is this is how you wright lens ??) should the lens in the virtual camera match the real ones .
that kind of setup

and what about all these lines placed on the floor in the 1st pic , i think it is part of the setup that helps in matching the
cameras ... what do you think

but thanks for your explanation

some one ... help ...HELP


zico out

[seth ranson]

These lines should be important for having a relation between lengts in reality and the ones in the virtual room, I guess. They probably had the same lines in the program before they took the photos. Surely they re also useful for checking if the point of view matches the one of the camera.
As far as I know a lens determines how big the picture will be in the end, right? I remember an option for changing that in one of these 3D programs when I tried to do something with it, so I guess that s merely a math problem. As well as the position of the camera in the virtual and the real room.
I have no idea how exactly you get the dots (btw: there was some term for them I think, but I forgot it) in 3D room however, the program COULD have an automatic function for that actually. I doubt that these people sit there and do calculations for every dot, would be easier to create the model manually (like, setting every dot on your own) then.

[Clayonite]

Theres a couple of different ways to model from reference. Usually, if you use pictures from angles as opposed to profile shots is because youre using splines, rather than placing verts. Its a pretty old school technique as its a bit tedious, but it does make things slightly more accurate, and youll get a better mesh flow than if youre starting out with a box. Well~ Thats not entirely true. Its all preference, really. But to put it short: Its only to make it more accurate. The base mesh is modelled in profile [and probably just half of it, as you normally symmetry those kind of objects].

EDIT:
Yeah, you can easily scan real life items and make them into 3D props. But those models are useless in game engines as theyre way too high poly and doesnt have a good mesh flow, in which will influence how itll appear on screen in terms of animation and lighting. Some people use the scanned models as base instead of photo reference, though, so its still good for something I guess. But again, its all preference.