7 hours ago, Mabe456 said:

A long time ago I used to help test OPUSfsx voluntarily by stress testing it. mostly because I love weather and used the sim to fly around in weather. OPUSfsx was great for that. 80% of why i got FS2020 was for the advertised beautiful live weather.

In my testing and comparisons with MS2020's live weather vs real world weather it is consistently downloading weather once a day that is either 12 hours or 36 hours old from UTC time. While what gets loaded is "real" weather, it is not live by any stretch. Even a 24 hour lag would be acceptable, but because of the 12 hour offset in the US we are getting nighttime weather in the day and daytime weather at night. For things like convective storms, fog, etc that depend on the time of day that simply doesn't work.

What's crazy is you can't even check what actual weather conditions are loaded into the game. The weather details are greyed out. you have to look at the globe and actually load into airports , use slew mode to check cloud conditions etc. All It gives us is a generic LIVE weather picture and wind speed/direction on your chosen airport which we can't even be certain is correct. When tuning to the ATIS it has told me the same conditions of 3nm visibility and the same 3 cloud layers since launch, so that is not working either. It's stuck.

Now in the US here for a while it didn't even inject weather in to the SIM, no clouds and clear air everywhere. I discovered by connecting to European servers I DO get clouds injected even if they are 12 hours old. I saw my first clouds in Texas in the weeks since launch just today! I also found sometimes that deleting the rolling cache file, and restarting the game sometimes triggers a wx update, but not consistently. Even then it still downloads 12 hour old data at best.

In the meantime, flying on the other half of the planet, you do seem to get correct weather and some winds buffeting the plane around. I did a flight around KL Malaysia and it was amazing to see actual working weather. It seems to have good resolution, and feels totally real and immersive. I even got to use the little radars in the plane, I didn't realize they actually worked. Exactly what the game is supposed to do.

I know LONG ago FSX had issues due to time zone changes and daylight savings. My hunch is something in the game is throwing up errors or causing problems on the back-end because something simple as DST is messing it all up. I suspect this also because while flying over Arizona and hitting the little blue "live" button in the in-flight weather settings to see a sunset, it would shift the UTC time off 30 minutes from reality. It thought I was 30 minutes in the future. I could only get it to detect the real time again by rebooting the game.

TL;DR something is absolutely broken with the weather system. We're not talking about a 30 minute or hour lag, that would be totally acceptable and understandable. This is something specific to the US breaking the weather injection system, getting stuck data in the ATIS, getting nothing off the US servers, and getting 12 hour old data at best from the European servers.

If they don't fix this then I simply will not use this sim. I want as close to real world weather as we can get. Sure, HiFi uses METAR but at least it's accurate, not this 12 hour old nonsense. I don't think the patch will address this at all as they decided to go with Metoblue who do nothing more than forecast conditions. They forecast based on the various weather models like the ECMWF, ICON, GFS, UKMO, etc... those models have inaccuracies in all of them. Those models are usually run only twice a day as well. Hence the 12hr old data that is seen in the sim. Until or if they incorporate METAR data sources, we will not have accurate data with any regularity.

Edited September 9, 20205 yr by B777ER

1 hour ago, KenG said:

Saying Jeppesen weather has problems is an unfair assessment

Jeppesen weather "had problems in FSX."  I did say "had" problems in my post, not "has" problems, and it should have been obvious I was talking about FSX.  I'm sure in the real world Jeppesen is totally professional and I would trust them completely.

If you prefer, "FSX default real world weather depiction had problems."  The weather data was provided by Jeppesen, and calling it "Jeppesen weather" is a shorthand way of differentiating it from other FSX weather engines and casts no doubt on Jeppesen as a real life company.

Hook

53 minutes ago, Farlis said:

If that is really true, that I don't get how they ever made a release candidate before fixing the weather issues.

There was push back on the part of some of the Alpha\Beta testers but I think the bean counters won out.

22 minutes ago, JRBarrett said:

I am connected to the US server on every flight, and I always get live weather, as long as it is the first flight after loading into the sim. I have seen plenty of clouds (where clouds are shown on the world map), and even thunderstorms. Like everyone else flying in the US, I don’t get wind, but all other weather elements work.

I guess we all are experiencing different weather then. I'm connected and upon loading for the first time, get weather that is somewhat close to real world, but that's all you can call it. Pressure is usually several points off, sometimes double-digits... If pressure is that far off still (and they call it accurate), it makes flying using online networks difficult.

7 hours ago, ThomseN_inc said:

I wonder how much more add-ons with a monthly subscription will come up in the future.

There is not a monthly subscription coming.

How often do the developers have to repeat that?

31 minutes ago, Robert Edwards said:

I guess we all are experiencing different weather then. I'm connected and upon loading for the first time, get weather that is somewhat close to real world, but that's all you can call it. Pressure is usually several points off, sometimes double-digits... If pressure is that far off still (and they call it accurate), it makes flying using online networks difficult.

Yes. That is the problem with using a forecast model. Models are most likely to be “close” to actual conditions when the weather is changing slowly, and less  so when it is changing rapidly. I did a flight yesterday from Santa Fe to Albuquerque. That area, at the time, was on the southern edge of the low pressure system that brought early-season cold and snow to Denver. The actual KABQ METAR pressure for the time I landed (at 09:30 local time) was 29.81. MSFS live weather gave 29.60.

I checked the MeteoBlue weather app on my iPhone, which has a chart that will show forecast surface pressure, and it did indeed show a lobe of 29.55-29.60 pressure extending into New Mexico from the low pressure center north of Denver. The problem was that the forecasted position of the storm center was farther south than it actually was at the time. The surface pressure at KABQ did indeed drop down to about 29.65 yesterday, but not until later in the day.

I just looked at the MeteoBlue app forecast pressure chart for right now, and it shows a predicted pressure of 30.12 at KABQ, the actual pressure per the METAR for this hour is 30.11

Sometimes close... sometimes not.

Edited September 9, 20205 yr by JRBarrett

12 hours ago, Tuskin38 said:

I had accurate 40 knot winds (according to Windy at least) at FL380 above Australia a few minutes ago, but then it just died and went to 03.

Doesn't that happen in real life? The wind has to start and stop at some point, surely 🤔

59 minutes ago, JRBarrett said:

and even thunderstorms

I haven't seen a single 'live' weather thunderstorm in the sim yet. Rain and dark clouds, sure, but no lightning.

 

  

2 minutes ago, nbillo said:

Doesn't that happen in real life? The wind has to start and stop at some point, surely 🤔

No I mean it just instantly changed, like someone turned it off, it wasn't gradual. Plus Windy said the wind should have stayed that way for most of the trip.

Edited September 9, 20205 yr by Tuskin38

56 minutes ago, B777ER said:

Until or if they incorporate METAR data sources, we will not have accurate data with any regularity.

I think this is going to be quite of a challenge. So lets say they will let METARS override the other sources and then what happens when you leave this area ? Its possible you will see totally different weather shifting.

Edited September 9, 20205 yr by roi1862

57 minutes ago, B777ER said:

Until or if they incorporate METAR data sources, we will not have accurate data with any regularity.

Only accurate at the point the observation is made, not for any time before or after.

Anyways they said in a Pre-release Q&A they're looking into adding METAR hybrid or something.

It may just be me but I'll be totally happy with the current system as long as it gets fixed.

The benefit of the system to me seem to be that it's dynamic and ever changing. If the pressure is .20 off or the cloud cover isn't right in California, well I'm in South Yorkshire, England so unlikely to know if it's accurate or not.

As long as the weather matches the data from the Meteoblue app then I can just use that as my reference. It's still an outside application that has aviation features and the ability to search by airport identifiers.

I realise I've mentioned Meteoblue enough times in this forum today but honestly it's as easy as switching to their data for your reference. 

Since the sim came out I have flown continuous legs from Washington state, down through Oregon, into Nevada and California and now near the Mexican border in the C172 and the weather has matched the Meteoblue app for every flight. 

 

21 minutes ago, Tuskin38 said:

Only accurate at the point the observation is made, not for any time before or after.

Anyways they said in a Pre-release Q&A they're looking into adding METAR hybrid or something.

Based on the admittedly ambiguous statement Asobo made about live weather in that Q&A, I think the ultimate goal for MSFS live weather is to blend forecasted weather with r/w METAR data for airport surface winds at least.

From the standpoint of flying online on networks like VATSIM, having the same surface winds as controllers, and other online pilots using METAR-based weather injection (in sims like FSX/P3D/XP) is quite important. The other weather parameters might still come from the model in MSFS. Having the same barometric pressure as other non-MSFS clients is less important. As long as an MSFS client syncs their aircraft altimeter to the current pressure provided by live weather by hitting the “B” key, the Mode C altitude that a Vatsim controller sees for the aircraft should be correct.

4 hours ago, Colonel X said:

I think "weather themes" will become a thing, if not by Asobo, then by an add-on.

However, why wouldn't they be able to fix live weather. It's just some data not being fed appropriately, it will be after the next patch.

It's not just the problems people are having, it's also the implementation of Live Weather. Unless I've missed something, there is no way to filter or cap the injected weather like there is in add-ons like ActiveSky.

I enjoy flying in real weather, but there are times of year in the places like the Pacific Northwest in Fall and early Winter when the wind speed is just too high for light GA planes and helicopters. And with an FSEconomy flight, I have to complete every flight with a landing, so I need a minimum cloud base for breakout on final.

I want the real weather injected clouds for immersion, and at least some of the wind, but also control over things like minimum cloud base and maximum wind speeds. I can get that with ActiveSky's granular control over injected weather, and I'd like to see that level of control in MSFS.

If that isn't Asobo's "vision" for injected weather, then it's one reason why we might need a 3rd party add-on to get the kind of injected weather some of us prefer. We don't all fly airliners that can mostly just ignore everything but the absolute worst weather conditions.

 

This is what I have red in an article two weeks ago and MS thinks about their weather model:

Mathias Müller, co-founder of Meteoblue, told The Verge via email that the hurricane was “very beautiful to look at and was accurately predicted by our models even days ahead.”

https://www.techspot.com/news/86535-microsoft-flight-simulator-players-exploring-hurricane-laura-real.html

 

Predictions even day ahead, ok then. I`m just asking myself why there are predictions necessary at all. Is it so difficult to track the current weather and implement it into the sim?

 

MS claims there are 250 millions POIs, 60 atmospheric layers. Phenomenal work.

 

 

2 hours ago, Farlis said:

There is not a monthly subscription coming.

How often do the developers have to repeat that?

Dunno? I never read anything or heard anything about them...