Good points, it actually only needs a few numbers changed to get smooth jump between each impulse. Here is a new completely nonlinear example tested on Skip barber, with smoothed/pseudo-randomized needle movement between each impulse: Since GIF has low frame rate/file size limitation, here is link to original 60fps video which the smoothed/randomized impulse effect is more noticeable: https://we.tl/t-rzFjE2eaeg Each "fake" impulse is now set at a roughly(randomized) 500 to 1000 RPM range step, with upper bound override value set 100 RPM below next lower bound override value, this which smoothed out needle jumping effect (as the extra 100 RPM range) and made the impulse effect more natural looking now. Further randomization is also possible with smaller or bigger RPM step (impulse gap). Mixing smaller steps with bigger steps can also create some sort of fluctuation effect as shown from above GIF. Here is the new code with comments indicating each step range: Code: //------------ Fully nonlinear RPM Step Example: roughly 500-1000 RPM for each impulse //------------ Target RPM: 0, Override: 0 - 659, Step: 759, Smooth RPM range: 100 (659 - 760) TachometerNonlinear=(0,0) TachometerNonlinear=(659,0) //------------ Target RPM: 760, Override: 760 - 1519, Step: 859, Smooth RPM range: 100 (1519 - 1620) TachometerNonlinear=(760,760) TachometerNonlinear=(1519,760) //------------ Target RPM: 1620, Override: 1620 - 2474, Step: 954, Smooth RPM range: 100 (2474 - 2575) TachometerNonlinear=(1620,1620) TachometerNonlinear=(2474,1620) //------------ Target RPM: 2575, Override: 2575 - 3236, Step: 761 TachometerNonlinear=(2575,2575) TachometerNonlinear=(3236,2575) //------------ Target RPM: 3337, Override: 3337 - 4409, Step: 1172 TachometerNonlinear=(3337,3337) TachometerNonlinear=(4409,3337) //------------ Target RPM: 4510, Override: 4510 - 4888, Step: 478 TachometerNonlinear=(4510,4510) TachometerNonlinear=(4888,4510) //------------ Target RPM: 4989, Override: 4989 - 5664, Step: 775 TachometerNonlinear=(4989,4989) TachometerNonlinear=(5664,4989) //------------ Target RPM: 5765, Override: 5765 - 6073, Step: 408 TachometerNonlinear=(5765,5765) TachometerNonlinear=(6073,5765) //------------ Target RPM: 6174, Override: 6174 - 7304, Step: 1130 TachometerNonlinear=(6174,6174) TachometerNonlinear=(7304,6174) No worries about auto-tools. I'll probably find some time to write a script to allow auto-generating above code with a few parameters that can be adjustable by other users, such as "step range", "randomized range", "smooth range", etc.
I don't know guys, I am not that picky on these things, it was already great before. Looking forward to implement it. I guess I am giving up on my battle with sounds. Whats wrong I have no idea. At least Gods of rF2 allowed me to work on sounds for a few hours, before they decided to corrupt it for me lol The beginning was not that bad, I can myself learning to do sounds, if only rF2 decides to play them... I would probably be releasing the car right now if not that silly stuff with sounds. Looks like Iwill have to release with wrong sound. Yesterday had little bit of drivign in mai ngame for first time. I think it was not too bad, only perhaps could stiffen up suspension a little, and perhaps slightly reduce tire grip. But had fun in Monaco and Pescara. It really is not very easy car to drive, it is intense, you have to be quick and precise. And it is very sensitive to setup. Whats left of what I want to do: Fix mirrors positions and scale and fine tune driver eyes position. Related; Some mesh seems to be getting flat shading when imported to 3DsimeD; Set tacho non linearity; Bake mirror spec AO and add to texture; Front inner sidewals UV is miscentered; Get more from the shaders; Sounds (looks like have to give up on this); Skins, I hate doing skins perhaps someone would like to ? I have template; Consider upgrades for various seasons; Revise engine params again; Check rain handling and AI and adjust it; Add backfires; Hopefully increase aesthetics/immersion for onboard view; Continue studying about car from all the material I have been supplied by community and implement findings. Here is my singular attempt for makign skins, Belgian car. Doesn't look very well. Troubles started with not finding good font. And then it is very difficult to allign elements that goes through several UV islands, even while using template. In comparison to clean red look.
@svictor It is still for a linear Scale. Does it also work with smooth Needle Movements and matching the proper Values, for other Gauges with non linear Scales, like e.g. the Temp Gauge on the Bottom of the below Picture? That btw. was the Start of this Topic.
A couple quick thoughts.as I am drinking doing some mod work on the F5000s.... 1. Sounds - It should be possible to use the rF1 sounds from the 1958 mod. It should allot for something to get the car to move forward. 2. I'm about half way through watching Neils video. Outside of the tire files, almost everything should be transferable from AMS1 to rF2. I'm sure Neils files could have some insight into getting some values for some things here, even though his is based on the Maserati. I haven't tried his car yet, but it sounds amazing.
The idea is that I want to make sounds. I already had somewhat OK samples made, although definitely not perfect at all, but for most use they were working well.. and then I launch rF2 several hours later again and samples doesn't work in game anymore, they sound distorded dulled, numbed, muffled. Seemingly for no reason. That really killed my flow off making this car, would have been released by now. I could have updated 911s in this wasted time, maybe one more car... For the moment it does see mlike I'll have to borrow the sounds. It is common practice right now anyway. Whats the last mod with new scratch made sounds ? I also want to have my own physics as well, of which I am fairly confident. And which is my main motivation to mod cars. I don't think I have nearly as much knowledge, skills, experience and overall competence as Niels, lots to learn from him. His car looks like it drives great, although in my view it is too easily slidable. Of course 250F is know as a sliding car, but it still looks too progressive and too easy on sliding. That being said Niels is one of my main inspirations in simracing, probably also why I do modding, can tell for sure that his video long ago hooked me on Goodwood. I am looking forward to watch his talk and drive with Maserati very much. I hope he never stops making them. Interesting idea to apply variability to physics to simulate realistic imperfections. It probably could have slight impact in feel and immersion, but likely these imperfections were never that huge that they would be making significant and very noticeable impact, maybe only after hours of driving they would become noticeable. Still nice idea. On the other hand it would complicate physics modding unreasonably much, and only makes sense to make traditional physics, and the nperhaps add slight randomisation afterwards. Also I would guess that IRL these imperfections would also more likely cancel one another to a degree for final handling result, rather than accumulate, although anything could be.
@mantasisg great work. @redapg thanks for the advice, here is new nonlinear-scale gauge with impulse effect example: New rF2ImpulseTachGenerator tool is released in following thread for those interested in creating historic impulse tach effect for existing or new mods: https://forum.studio-397.com/index....-creating-historic-impulse-tach-effect.76049/
I was having weird issue where rain effects on body paint looked very bad and in general paint shader in rain looked very bad. Found information, an old post by @svictor about how material editor saves in wrong encoding and it needs to be changed. I have changed it, and indeed it is much better in all circumstances car looks better now. But new issues emmerged - the rain effect is gone completely, and the reflctions are sometimes choppy (lags a little). Another weird issue is after having added tach peak needle, for some weird reason needles of oilpressure and fuel got replaced with it. But I must say I love the tach now. Good that oil needle remained which is very unique. I am not stressing myself to fill the mod with all drivers yet. Of course one with blue helmet and one with brown helmet will have to be there. I just somehow can't find myself working with skins very much. And icons Not worth making icons for just a few cars yet. Besides these issues and issue that I have experienced with sounds car is almost ready for first release. If all goes well and during testing I won't find big issues, then perhaps in next few days you will be able to experience how good you are with this car. Physics are well prepared, but I suppose there could be lots of details to work on still. After this release I will probably calmly turn my head on updating previously released cars, as I have learned quite a lot, and got much more comfortable with modding lately, just need to manage time and this also means I must not spend too much time modding, which is becoming a little addiction. I have plans for years. How many players in rF2 there will be in 2027 ?
I opened wccarbody material in material editor today after I changed encoding of a json file yesterday. And apparently it is no longer an IBL material, at least not in a menu of material editor.
@mantasisg Great job. About material,make sure have a "MaterialPath" line in both of .GEN and spinner.GEN files (ex. assume you save materials in "maps" folder) such as: MaterialPath=<VEHDIR>\MOD_NAME\maps This is very important for material editor to correctly save any modified JSON file into mod folder and avoid strange issue (as I noticed earlier that "MaterialPath" line was missing from your W125 mod), this work for both showroom & on track. For needle, may be check if peak needle texture file name or extension is correctly set. I also keep a general WCCARBODY material parameter file for different base materials, which can be used as reference or fallback when tweaking WCCARBODY material parameters, something like in attachment.
Thank you, I'll check it all next time I sit down. First time I hear about material path line in gen files. Yes strange stuff with needles. It helped for me to simply open tga textures of needles and then save them again. Strange.
Yes I suppose upgrades could be utilised for that. I guess the dafault should be with inboard starter. But it is such a small element no one would ever be turning off the inboard starter anyway. How can I control how raindrops look on carbody ? Do I need to scale some particular UV channel ? I have troubles with how rain looks on car body. Not good. I have tried adding ripple maps and raindrops json. Still not working. Rain looks especially bad on the interior, I would like to remove it, or make it much less there. And whole car just looks bad with raindrops.
can you paste you WCCARBODY.json here (or share it privately)? it's kind impossible to know what's wrong without seeing the file or be able to test. as for getting rid of interior rain, you will need to make a separated interior cockpit or parts mesh to use Standard IBL shader (must avoid using skin livery shader such as the one used on WCCARBODY), this may require some extra work to setup those materials(with special aos Mr NM maps). Aos map also has big impact on specular/reflection issue, especially for formula car cockpit. For example, it's common to see on some formula car mods that cockpit inner body gets too much reflection (from road surface) or specular strength due to either using the same WCCARBODY or incorrect Aos map. This problem can be solved by using IBL standard shader with specifically tweaked Aos map (tuning down specular channel).
Here is an example of raindrop & ripple map and corresponding raindrop_desc.json in attachment (from a new WIP formula car mod that I've been working on with another modder): The important part here is to set UV in a way to minimize amount directions that raindrop would move against gravity (while car still) & forward direction (while car moving), since there is a limit to maximum amount different colors on ripple map (8 colors according to the official guide). For this WIP car, it has complex aero parts, and the UV map of each parts was initially pointing to various directions, which it exceeded maximum different colors on ripple map. In order to solve this issue, the UV was adjusted several times to unify the rain drop moving direction on UV map. So for complex car, it is important to do this right early on to avoid such issue when later working on ripple map. I ended up using all 8 set of colors for this car. Then mark down raindrop directions on UV map, paint different colors for each different raindrop moving directions, also good to write some description in raindrop_desc.json file to help identify directions later, such as "zero right, down right, up left, etc".
Thank you, there seems to be a lot of possible hooks where the problem catches on. I have tried yesterday making basic ripple map, I just slapped the default raindrop_desc yesterday, it just didn't work, no movement at all. It sounds like there would be a lot of trouble to get rain effects out of interior. I think it would be best to have it there, but just to be able to make it look good, reduce scale of effects and improve the looks. Oh well a lot to learn/discover still. I really want to release the car to steam, so I could perhaps get some feedback/help, and get unstuck. Another funny possibility is that it just looks bad in rF2, and I have never noticed it. Here is one YT video: Also is bad looking and totally static. Hey I haven't tried perhaps I need to max out my graphical settings. Here how it looks: And it doesn't change, just static look like that all the time. I wish I could at least scale it down. From outside does not look as bas, and from outside it is animated, although no movement. W125 might look bit better due being bigger and silver color. Als oit has interior setup differently: All to do/issues so far that I notce: can't solve: -Some triangulation of mesh notable on alluminum material, including steering wheel; (try fixing normals in 3Dsimed ?) -Mirror edges bad transparency due to mirror edges mask material; (probably will have to redo meshes) -AI does not respond to AInegsuspension parameter well enough; -Can't get my sounds working; -Steering arms anim, how to; -Potential IBL carpaint shader problems (rain and darkish ambient in shade); -showroom front wheels are slightly off. to do for future - better interior look; - more better detailed skins and icons for them; - riveting and similar small details of a car; - front inner sidewals UV miscentered; - improve shaders; - upgrades (still to consider what those would be); - backfires; - optimisations; - generate more physics testing.
@ThomasJohansen No, that doesn't seem to have effect. It might be that rF2 just isn't designed to ahve great body paint wet effect from cockpit view. Anyway. It was productive day. I don't know if I will be able to release tomorrow. Still have to do some more testing. Still finiding little things that I must adjust. Pace wise car is very fast, I wish it was better season for Lancia in 1955, I think they would have dominated, or at least seriously pushed Mercedes. I tested today in Belgium and in Monza. It was beating Mercedes seriously in both. I though that I might have it too fast, then checked how it performed in 1956 and it was pretty much those laptimes. Of course Castellotti did take Lancia to Belgium in 1955, but I don't think he pushed it there to the limit coming on his own, still was able to take pole position. They only did race it properly in Monaco in 1955. Lancia D50 definitely was an amazing car. Year 1955 is such a pivotal year in motor racing history. Ferrari would have never ever become such a major name if they wouldn't have gotten assets from Lancia. Three strongest rivals leaving racing (Lancia, Mercedes and Jaguar). In sportscars racing Mercedes 300 SLR with minor upgrading could have kept on winning up to 1960-1963. Lancias D program sportscars also would have been strong. Jaguar was pulling away from racing after 1955, but kept on winning by private racing teams.
