Official Real Weather Report Topic

@twistking @Shriukan @F1Aussie sorry for the tag but just reading your posts on weather and livetrack and it has really helped.

I had not realised that the livetrack is affected by the multiplier but it explains a lot. I think livetrack and time/weather multiplier need to be uncoupled.

It explains a lot of the issues we have had with running real weather replica events over at AMSUnofficial and I believe it also confuses the heck out of the AI (not that we use it).

We have been using generally low multipliers and real weather (max x5) but have found when it does rain that you can just keep on dry tires for a very long time. The multiplier explains that.

You wouldn't expect a track to instantly rubber up due to a time multiplier and neither would you expect it to instantly dry out.

If people choose very high multipliers and get very rapid weather changes then surely that is like having a short but rapid downpour. The track doesn't instantly dry out as a result. People would still be able to choose unusual combo's and if the conditions are a bit crazy because of it then that is not the games fault but the users fault. However, having the livetrack affected by time makes running real weather replica events (or real weather in general) quite difficult due to the lack of granularity with the multipliers and using any time multiplier (even 2x) seems to bring into the realms of unrealistic "livetrack".

So... for me... my request would be two things:
* Decouple livetrack from time multipliers
* Add increased granularity to the time multiplier options. x1, x2, x5, x10, etc is not granular enough. we need step 0.25 from say 1 to 3, step 0.5 from 3 to 5 and then maybe step 1 from 5 to 20 and finally step 5 from 20 to 60.

 

I don't think it necessarily needs decoupling. Livetrack is always going to be influenced by time seeing how time determines the amount of sun and temperature. What it needs is rather a way to extend the "Weather Slot transition window".

Example, if real time a track would take 10mins to get from dry to soaking, at 5x it would take 2mins. If the devs manage to find a way to code in a "Weather transition slowdown multiplier", we could then counteract the (sometimes) flash track state changes by making the transition slow down and last longer, thus giving more time for racers to change strategies.

I do agree that adding more granular steps would help. Your example fits more or less what I had in mind.

 

Yeah, I think it just needs tweaking. I use 5x multiplier all the time now and have found recently that with light rain the track wetness builds up fine. it is when there is sudden heavy rain that there seems to be an issue. Where the track can go from bone dry to almost undrivable on slicks in three corners. this is the biggest issue I have with the weather i only use real weather too so I still think it needs tweaking with how it interacts with livetrack.

 

Extending the transition would be another way to do it I guess but to me a separate livetrack multiplier is needed. That way all options are available to all.

Real weather could be sped up with normal livetrack scaling meaning transition is realistic. Of course people can match the livetrack multiplier and time multiplier to achieve what we have now if they prefer.

 

In custom championship real weather appears to act like it's in accelerated time, it'll go from blue skies to dark and wet within a few corners or a lap, and it dries up too fast as well, same with fog. This is with 20 odd lap races at 100% distance.
It seems better in single race mode though.
Haven't tried since the Monza update so I'm wondering if it's been looked at.

 

This has been an issue for a while now, before the update I had a race vat jacaperagua and it went from sunny to overcast to thunderstorm and flooded track while the same straight! There is also an issue with the weather generally. Sometimes if my next session starts off wet I can exit out of the game, go back into it and the weather will be different.

 

Yeah I get that when I restart too, that doesn't bother me too much tbh.
I hope they can fix the fast transitions though, and the ai doing multiple pit stops in showers.

 

I wonder if Wind Speeds are being correctly simulated. I was on Touristenfahrten today, Nords 24h layout. Ingame time was 25/10/2021 (today) at 13:18 and wind speed was of 4.2km/h and yet Openweathermap reported it at 11,2km/h (and Windy.com at 14km/h). I think the Wind direction (Southwest) was correct.

I added a Wind Speed and Direction widget in SimHub so I will pay closer attention to this as I believe Wind Speeds should be part of the simulation and how it affects the cars balance and braking.

 

Went ahead and checked again what OpenWeather was reporting and once again there is a discrepancy in the wind speeds today. Check the reported OWM wind speed at 18:00 (3,4m/s * 3,6 = 12,24km/h) against the ingame Wind Speed (6km/h). The wind directions also do not match. OWM says SouthWest. Ingame says North.

 

I think in general the weather system is still buggy. When I have exited a session and then gone back into it I have had different weather conditions and even different seasonal foliage. Someone made a good point recently that when you go into the first session the loading screen always defaults to a March 2019 date whether or not it is just a loading screen bug or if it is an actual date bug I don't think we know.

 

Doesn't look like the temperature matches for the time either.

 

There is no error in the loaded dates on my side, however I am doing another check today. I was reminded Practice session should be fetching the info for 2 days back so I advanced the server date by 2 days. Temps not quite matching is not abnormal for such a big track as Nords, but they are still within the expected delta. The wind speeds also vary quite a lot depending on where you are but the reported value effectively didn't match at all. Weather has been more or less the same the last few days though.

 

So with the server set 2 days ahead, practice should in theory be fetching today's weather data from OWM.

Well I checked Nurburg at 11am and was given fog until 1pm and then rain from 1pm to 3pm when there was neither today or previous days.Temps were right, Wind Speed and direction weren't. I know it's a huge track and fog will probably be present in the lower areas depending on humidity levels however I can't help but think that maybe weather values aren't being interpreted correctly.

I checked back at 5pm for live weather. Temps are right, but wind speed and direction still don't match.

 

I wonder if this is an issue with weather API that doesn't take into the account localised weather, but quite generic information. Similar to looking at current weather on the phone's lock screen which often doesn't match the reality.

I suppose the accuracy will be as good as data of weather stats supplier. Do we know who that is as it would make sense to compare it to data they provide rather than what we see out of the window. Then there would be a question whether it's feasible to switch providers who can provide more accurate data (if such exists).

 

I am comparing the ingame info directly with OpenWeatherMap which is what Real Weather is polling from. Now where I could see discrepancies is if the website gives info for a place and the API is polling from quite different GPS coordinates. But from Adenau, to Nürburg, to any GPS coords on track there are at most 8kms of difference. Enough to fluctuate temps by 2 or 3 degrees, considering the terrain, but not enough to fluctuate wind direction.

 

Ah, I didn't realised that the data in AMS2 was using OpenWeatherMap. In which case I agree the issue must be in utilising the API, whether it's down to different date (regardless to what's set in game) or just wrong geo-coordinates being used for that location.

 

The date issue happens in custom championship not sure if in other modes as well.

 

Doesn't happen on the DS but there was some further development. I was only testing on my server, so online, however trying the same settings offline yielded correct wind speeds and directions. So it might be the online implementation was missing. Crimson already forwarded it

 

While doing one of my comparison videos of the video below (timestamped 6/9/2018/ 10:00 am), I plugged that time into the Historical Weather system, and it was raining in sim, whereas in the real life video it was only overcast.

 

This happens to me regularly. I find that often if I quit the session and go back in then the weather is different