• Bookmark hashtag #RDLMS on Twitter, Facebook and Instagram
  1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

WIP 2000 Mazda RX-7 Type RZ

Discussion in 'Racer Cars' started by Stereo, Dec 23, 2012.

  1. This car is now released!
    Progress on the things I'd have liked to improve was approximately 0, so I'm just going to call it final for now with the new wheels and ini.

    7zip Download Mazda RX-7


    It's been a long time - a very long time - since this was released. My files suggest it was 8 1/2 years ago that it was last exported to Racer.

    So, new decade, etc. I'm working on the actual model for the first time in a while.

    In this pic the CoG wasn't adjusted
    Yes, the right side's also labeled wrong. I'm bad at left vs. right. (this is with CoG fixed, doing 0.9g cornering)

    I also resplit the materials, which let me do something I wanted to for a while - matte front spoiler. In general I'm improving the materials too, because the old one had some odd splits to allow depth-sorting.

    The rear spoiler had one corrupted sections of UVmapping, I'm not sure why but that's finally corrected (thus the test texture)

    TODO list:
    DONE - model pop-up headlights (not likely to be animated, just on-off. Also due to how lights were originally modeled, not gonna affect the base vehicle so they'll be slightly odd.)
    DONE - add illuminated (ie. emission) light objects for front, brakes (hopefully including car body illumination)
    - illuminated reverse lights
    DONE - move RZ badges into the decal texture (had their own shader)
    DONE - re-confirm the overall body dimensions - it's 1760mm wide, but the wheels seem tucked a little far in now.
    DONE - figure out if the wheel models are the right width, given track seems narrow
    DONE - LOD for wheels
    DONE - import+export the interior model to change material names and possibly realign it with outer model
    - blur models for the wheels
    DONE - redo brake models (caliper left and caliper right are the same model ???)
    - upsize main texture to 2048x2048, possibly double other texture sizes too
    - rename textures to match material names where possible, for greater harmony
    DONE - gauges back to the RZ's white-face km/h only ones
    - add small side lights, front
    - add fins inside front intake
    - bumpmap/specmaps for interior
    DONE - bumpmap sidewall, tread
    - model wipers so they can be animated? unlikely
    - would be nice to do a body LOD that guts the interior too, unlikely
    - basic nocg shader

    Final TODO list:
    - remove unused sections from shader
    - remove unused textures from folder
    - locate window2_f (and other custom shaders) in the car folder
    - test on clean install of latest 090 RC.
    Last edited: Nov 5, 2013
    • Like Like x 7
  2. Is it safe to reply now, with all the invisible edits? ^^

    I'm assuming there won't be any updates to the meshwork itself, like the side marker lights, sidestep contour, front intake strakes...?

    When it comes to the body dimensions there's more to it than box size. According to the official brochures, width of 1760mm is without mirrors in any case. Depending on the positioning of the wheels models - offsets in mesh or car.ini? - there seems to be a bit too much lateral overhang, or too narrow trackwidths. Best to always countercheck vehicle box dimensions against wheelbase and trackwidths to catch out oddities early on. Whether a vehicle total length is off by 10mm is a fairly moot point, but you can easily eyeball the same distance when it comes to wheels-in-fender placement if it's off in any direction.

    Wiper animation would be nice in general, but the limitations in the movement Racer currently still has means it's a battle you can hardly win - always going to look odd floating or clipping somewhere.

    Speed limiter is original equipment and should be included, in my opinion - easy enough to deactivate such a script as an enduser if desired anyway.
  3. I mostly do invisible edits cause nobody replied :p Nothing removed from the post, just adding points as I think of them, so I have one list to refer to.

    Minor mesh additions (like the intake strakes, side markers) are doable, anything that has to be edited (like the not quite accurate side panels) isn't as easy to change, since I'm working off a version that's already mirrored and remapped. There are a few minor details that can be fixed with textures though (panel gaps between front/rear rockers on the side, plus the fake-gap on the door). I could play with ZMod's MUL & do symmetric edits but that's only good for moving vertices.

    I'm not sure how to measure the track width given the models I have, but the wheelbase (based on suspension points) is within 2mm of Wikipedia's number, and the wheels look decently centered in the wheelwells so the length's probably fine. (with weight distribution, I haven't seen numbers more accurate than 50/50, so that's good enough)
  4. Just checked on the original release - wheel model .dof files were centered and car.ini shows no offset values, but they sit flush with the fender lip, so a bit too far outboard (I see that I had worked on the car.ini in 2008, maybe you had slightly different numbers originally). So, something has changed since then, regarding either body width, trackwidth (1460mm both axles, not sure about wheel ET differences from lower grade models though) or wheel mesh centering.

    Longitudinal positioning looks fine though, I agree.
  5. Hrm, I may have tweaked the wheel models since original release on my end, I've got a couple sets of dofs and I do remember them being originally too large. Once I've crossed off the major todo items I'll upload it with car.ini so we can compare notes.

    I set the track up to 0.73 on each side (from 0.72) and it's not bothering me so much anymore. Part of the effect might just be the untextured arches; the little painted lip does make a visual difference.

    The 'test' texture has the rear quarter panel reflectors, but the white body.tga doesn't - not sure what happened, none of the Raven variants do, but I don't remember making those. 'test' is from the original folder I built it in.
  6. It's tricky to pinpoint exactly where the small differences are coming from, but this last screenshot is fairly close in placement. Are the wheel models scaled for tread width or total tyre shoulder width? That is also a common issue, since you never really know the actual size of a tyre given typical variation for a given declaration (ie, rear tyres on the RZ being 255/40/17 - is the .dof file 255mm wide or more like 285mm? You have some wiggle room there if you're quite sure on the body hardpoints being correct and it still doesn't match close enough).

    The sidewall shape you currently have is a nice rounded bulge, like the 16in models have it - the 17in tyres seem to have a straighter sidewall with a much more pronounced step around the lip of the rim. Thinking about it, 255mm tread on a 8.5in rim with Z-rated compound probably requires that. I grabbed a screen from the 2002 brochure that shows the detail somewhat (low resolution unfortunately), inside and outside appear identical in shape:

  7. Could be the width, yeah. The best side-on shot I have is of an RS, which is 17" rims with the same tyres as the RZ:

    I can give the wheel models a once over to figure out what's going on later (EDIT: 236mm front, 250mm rear, which is definitely narrow for the 255s on the rear. Actual tread's only 215/235mm so there's definitely wiggleroom on both ends to make the wheels fatter. I changed the sidewall of the LF wheel and compared it to a photo, seems closer http://i.imgur.com/Egeeu.jpg). For now I'm going through things in order.
    Which means the headlights are kinda done. The front edge is curved the same way as the base car model, but doesn't exactly match the couple photos I've found, might give it another tweak later.
    • Like Like x 1
  8. Now whats crazy is I remember this thing being released. Awesome to see it come back.
  9. Yeah, it's good to get it running again.

    1st of 3 new light textures painted. Dunno why I bothered texturing the base headlight object when this is gonna be in front of it at all times. Included a small body illumination, I think it looks ok.
    • Like Like x 1
  10. Bottom one sure, top one is kinda funny shaped and not generally as visible.

    Not used to ZModeler anymore... the exporter's a lot more convenient than in Blender though.

    Brakelight textures redone too.

    EDIT: Seem to be having a shader problem - the additive lights don't let the potentially transparent glass show through them.

      ambient=0 0 0
      diffuse=0 0 0
      emission=30 30 30
      specular=0 0 0 1
    The back panel is transparent over the lights which is why it uses a blendfunc. Changing sort_offset doesn't seem to let this work - the only thing it'll do is get the brakelights showing through everything else, too.

    Weirdly, the other brakelights (which are behind the glass) work just fine.

    I wonder what shader the light flares are using, they don't have the same problem. Maybe blendfunc=add isn't good anymore?
  11. I've had trouble with add too doing the same thing but fixed it iirc, will check my car that used it later and post my shader :)

    Ps looking great :)

    Psps anyone tried using a 'headlight' for lights facing back to give a subtle red glow behind car?
    Would be nice to have one extra for brakes too?! Or more optimal change intensity values of existing ones...
    Maybe something script able?!

    • Like Like x 1
  12. I think someone did that (i was an SLS seen here) but who? don't remember
    Doing directional lights will be nice to, the problem will be the projected texture of the lights, and the orientation of the flares :(
  13. I was curious about that too, so I tried moving the start of the projection backwards, and did a test grid texture to see how it's actually projecting.

    (also drew a red dot, hoping it would work... kinda didn't)
  14. Can't you set one facing the other way? Ie a whole new light?

    Thought you could... Another feature that needs un-hard coding and then being referencable via onyx :)

    High/low beam changing texture and angle etc, and brake version changing intensity too.

    Simple as its all functionality there, its just leveraging it to more powerful end results :)

  15. At the moment, seems to be just the one light0 working, and it faces forward, yeah.
  16. Hmmm, tinkering here with the glass shader again.

    I had all manner of issues on a car I worked on, as soon as you have anything with a blend under something else with a blend, then things go awry it seems.

    I've got add working fine here under a blend blendfunc glass, but src_color one doesn't seem to work...
    What I'm missing with the add blend though is the bloom... it's present if you zoom in under the glass, but not present if you are outside of the glass... so that kinda ruins the effect too.

    There probably is a solution but it's a matter of working it out. It's a problem that needs a fix though imo as using a blend under a blend is kinda important in my view for lots of headlight related things.

    It does seem random when it works or doesn't. In my case I copied exactly the Lambo config and model layout logic and it still had trouble. Ie, the Lambo glass covers pass the emission of the opaque emission models underneath AND the bloom, but on my other car I had trouble with passing bloom with an opaque emission model.

    Arghhh... making my brain hurt now.

    As per the headlights, pitch = 180 should flip them over, yep, works here.

    Hmm, but only one light is passed.

    Try a new unique texture... nope...

    In theory if these lights are exposed to Onyx you could change the rear light intensity (if lights are on) with Onyx/brake signal, so it's still one light which is cheaper than two.

    Same with high/low beam adjusting the light properties (or even the texture!)

    So all we really need is Onyx exposure and the multi-light structure being enabled rather than hard coded to 1... probably 5 mins of work for Ruud, but loads of potential for cool night time lighting features!



  17. I don't use blendfunc=add anymore, or flares since bloom looks more realistic anyway.
    Try this...

    bias=0.15 ; Fresnel = bias + scale*exp(incident angle, power)
    scale=1.0; eg. chrome would be more like bias=1, scale=0 as it always reflects
    power=2.0 ; power just changes the curve

    ;----------------------- shaders -----------------------

    ambient=.1 .1 .1
    diffuse=1 1 1
    specular=3 3 3
    ambient=1.0 1.0 1.0
    diffuse=1.0 1.0 1.0 ;ambient and diffuse to make sure it's lit correctly
    emission=50 0 0

    Alex Forbin
  18. on lights i use blendfunc=blend, rarely causes issues for me, if it does I use shader_offset=1 and that typically fixes it. But I also never bother with flares anymore, the bloom looks better and really is easier. But just looking at your shaders might have to try the fresnel out in the vf section.
  19. Moved blendfunc=blend on the glass shader from layer0 to layer1 (envmap), and now I at least have consistency - if sort_offset is higher on the glass, then the regular brake lights work and the middle brake light doesn't. If sort_offset is lower on the glass, then the middle brake light works and the outer ones don't. When it's equal, the middle light doesn't work again.

    I think the solution is gonna be to split the third brakelight off to its own material, so it can be sorted after the glass, while the other lights are still before.

    Removing blendfunc from the brakelights shader just makes the black corners of the polygons visible.

    Bloom disappearing under glass would just be due to the 'blend' dimming it, probably.