Historical weather is absolutely a must

[Franz007 505]

17 hours ago, Abriael said:

As for the weather itself, one of its most important aspects is how dynamic it is, which Microsoft Flight Simulator reproduces better (far better) than any competitor

Til now i asked a few times but no one was able to answer that: can you explain how exactly the weather-dynamics MSFS is supposed to be better than any competitor?

[Fiorentoni 4,339]

3 minutes ago, Franz007 said:

Til now i asked a few times but no one was able to answer that: can you explain how exactly the weather-dynamics MSFS is supposed to be better than any competitor?

I wasn't asked, but I'll give you one point: MSFS weather is global, while - from what I know - all competitors have local weather only. In simple words, in MSFS transitions are much rarer (usually only when a very different airport METAR is injected into the global weather, but to be honest I can count the times I've noticed ANY sort of transition in MSFS on one hand) and you can simply look out for the weather at the horizon far away.
With ActiveSky you cannot really have an overcast on the horizon and clear weather at your position, though ActiveSky does a very good job in masking these things.

[Franz007 505]

34 minutes ago, Fiorentoni said:

I wasn't asked, but I'll give you one point: MSFS weather is global, while - from what I know - all competitors have local weather only. In simple words, in MSFS transitions are much rarer (usually only when a very different airport METAR is injected into the global weather, but to be honest I can count the times I've noticed ANY sort of transition in MSFS on one hand) and you can simply look out for the weather at the horizon far away.
With ActiveSky you cannot really have an overcast on the horizon and clear weather at your position, though ActiveSky does a very good job in masking these things.

Unfortunately this statement is wrong. XP12 has global weather with transitions, fronts moving slowly with the winds, entirely dynamic.

[Mace 1,948]

Just when I thought this thread would slooooowly fade away.

[Sky_Pilot071 581]

Time for a poll - yea or nay.

sp

[Abriael 952]

1 hour ago, Franz007 said:

Unfortunately this statement is wrong. XP12 has global weather with transitions, fronts moving slowly with the winds, entirely dynamic.

That doesn't help much when the data used is much more approximate and coarse than what is provided by Microsoft Flight Simulator's combination of Meteoblue data injected every 15 minutes and METAR. Of course, the fact that X-plane's clouds look like ricotta and cotton candy had a kid and that kid had a bad hair day, and look *nothing* like real clouds dosn't really help either.

[Franz007 505]

40 minutes ago, Abriael said:

That doesn't help much when the data used is much more approximate and coarse than what is provided by Microsoft Flight Simulator's combination of Meteoblue data injected every 15 minutes and METAR. Of course, the fact that X-plane's clouds look like ricotta and cotton candy had a kid and that kid had a bad hair day, and look *nothing* like real clouds dosn't really help either.

It's not more approximate than in MSFS (take a look at the MSFS forum). And once again the same old lame excuse; we were talking about the dynamics of weather and the only think you can provide is to talk about the visuals.  Lol. And talking about clouds, i haven't seen any post-atomic ashy-looking clouds in XP12 yet...fact is the weather-dynamics are not better simulated in MSFS, and even less "by far".

Edited June 30, 2023 by Franz007

[Abriael 952]

10 minutes ago, Franz007 said:

It's not more approximate than in MSFS (take a look at the MSFS forum). And once again the same old lame excuse; we were talking about the dynamics of weather and the only think you can provide is to talk about the visuals.  Lol. And talking about clouds, i haven't seen any post-atombic ashy-looking clouds in XP12 yet...fact is the weather-dynamics are not better simulated in MSFS, and even less "by far".

Weather data in XP12 is 2 hours late and mostly from open source providers that don't get even close to what's provided by Meteoblue in granularity, quality, and coverage. 2 hours -> 15 minutes is very far in terms of weather.

Before you talk about "fact" you may want to do some research. 

PS: if the weather doesn't look even remotely like weather in 2023, and XP12's certainly does not, it definitely doesn't testify positively on its quality and the implementation of the tech.

Edited June 30, 2023 by Abriael

[Franz007 505]

Just now, Abriael said:

Weather data in XP12 is 2 hours late and mostly from open source providers that don't get even close to what's provided by Meteoblue in granularity, quality, and coverage. 2 hours -> 15 minutes is pretty word not allowed far.

Oh Boy...You seem to mix up everything. First of all, upper weather datas are not "late" because they aren't from the past but based on forecast-datas, exactly like in MSFS with Meteoblue. . Perhaps you were talking about the local metars? But that is wrong again, because they are updated at least twice an hour. You can simply check that by verifying the weather-files.

[Abriael 952]

2 hours ago, Franz007 said:

Oh Boy...You seem to mix up everything. First of all, upper weather datas are not "late" because they aren't from the past but based on forecast-datas, exactly like in MSFS with Meteoblue. . Perhaps you were talking about the local metars? But that is wrong again, because they are updated at least twice an hour. You can simply check that by verifying the weather-files.

No. I'm not talking about METAR. Metar is the easy part that everyone can do. It's the GRIB data that is 2 hours late. Forecasts can still be late if the simulator injects them once every 2 hours. The weather is ever changing, so forecasts need to be consistently updated to be somewaht precise. If I inject predictive data now, in 1 hour from now, it's going to have lost much of its validity because in the meanwhile real-world conditions have changed and new predictive data has been created. 

Meteoblue doesn't just use (much more complex and complete) forecast data, but interpolates it with real-time satellite imagery and measurements every 15 minutes, enabling *much* more precise simulation. 

Basically, Xplane relies on numerical weather data interpolated with the bulk of its data that comes from weather forecast, and then aided by METAR, which is very sparse and offers very few data points. Incidentally, X-Plane's GRIB are from NOAA (which focuses on the US) so their global coverage is questionable. 

Microsoft Flight Simlator has all of that (and of better quality) aided by frequent real-time satellite support which corrects the inevitable lack of precision of the forecast and complements the data/coverage deficiencies of METAR. 

That's the difference between using an open source provider and having actual resources to put into having professionals do the job. 

Edited July 1, 2023 by Abriael

[Franz007 505]

@Abriael Thanks for the explanations. You are the first person who is able to explain it to me in details. I have to admit that I wasn’t aware of them adding satellite-based datas. So we can say that MSFS’-weather datas should be a bit more precise then. It would be interesting to see a difference for a given place/situation. The weather simulation is for me and by far the most interesting thing of flying.

But do you know how it can happen, that sometimes an airport reporting for example BKN060 or something similar is shown with completely blue sky? From my understsnding, this shouldn’t be possible as long as the raw-datas are being refreshed that often?

[Bob Scott 7,745]

15 hours ago, Abriael said:

No. I'm not talking about METAR. Metar is the easy part that everyone can do. It's the GRIB data that is 2 hours late. Forecasts can still be late if the simulator injects them once every 2 hours. The weather is ever changing, so forecasts need to be consistently updated to be somewaht precise. If I inject predictive data now, in 1 hour from now, it's going to have lost much of its validity because in the meanwhile real-world conditions have changed and new predictive data has been created.

The weather itself may be ever-changing, but forecasts change far less and far more slowly.  One hour from now, the predictive data will still tend to give very good results.  The state of weather modelling today is good enough that predictive data is usually valid and fairly accurate for considerably longer than the interval between forecasts.

15 hours ago, Abriael said:

Meteoblue doesn't just use (much more complex and complete) forecast data, but interpolates it with real-time satellite imagery and measurements every 15 minutes, enabling *much* more precise simulation.

The proof of the pudding lies in the tasting...this "precision" you keep touting is not at all meaningful when I look out my window or at realtime webcam video and see clear skies and then fire up the sim to see low overcast and rain or vice-versa.  Of what value is "precise but wrong"?

15 hours ago, Abriael said:

 Incidentally, X-Plane's GRIB are from NOAA (which focuses on the US) so their global coverage is questionable. 

Oh no, no, no.  NOAA's GFS (Global Forecasting System) is just that--global in scope, with data sourced worldwide, and used in a wide variety of critical real-world applications.

15 hours ago, Abriael said:

That's the difference between using an open source provider and having actual resources to put into having professionals do the job. 

When it comes to professionals doing the job of in-sim weather, the HiFiSim guys have been working minor miracles for this community for a long time now.  It royally stinks that they and their expertise have been arbitrarily locked out of this platform--I find it saddening when I contemplate what could be if they had the same kind of opportunity for collaboration that we've seen between Asobo and others--Working Title, for example.

[Abriael 952]

13 hours ago, Franz007 said:

@Abriael Thanks for the explanations. You are the first person who is able to explain it to me in details. I have to admit that I wasn’t aware of them adding satellite-based datas. So we can say that MSFS’-weather datas should be a bit more precise then. It would be interesting to see a difference for a given place/situation. The weather simulation is for me and by far the most interesting thing of flying.

But do you know how it can happen, that sometimes an airport reporting for example BKN060 or something similar is shown with completely blue sky? From my understsnding, this shouldn’t be possible as long as the raw-datas are being refreshed that often?

From what MeteoBlue has shared a few days ago, the change to hourly to 15 minutes interval is recent. 

https://www.meteoblue.com/en/blog/article/show/40088_meteoblue+live+weather+data+in+Microsoft+Flight+Simulator

As of late, I've seen much better matching, to the point of it being extremely impressive.  I was watching the Achorage Planespotting livestream by Airline Videos the other day, and it looked nearly identical, to the point that you saw exactly the same areas of the mountains East of Anchorage Airport peeking through the clouds, while the areas that were covered were also covered.

Incidentally, when you see a completely blue sky, it means you're somehow not receiving the data at all. I haven't seen it happen as of late, but I suppose it can. If there's absolutely *zero* clouds in the sky, it's very likely a data outage. 

6 hours ago, Bob Scott said:

The proof of the pudding lies in the tasting...this "precision" you keep touting is not at all meaningful when I look out my window or at realtime webcam video and see clear skies and then fire up the sim to see low overcast and rain or vice-versa.  Of what value is "precise but wrong"?

I've seen results consistent with the sim by looking out of the window, webcams, and livestreams *much* more often in Microsoft Flight Simulator than in any other sim, with or without add-ons. It's not even in the same galaxy. Incidentally, I travel quite a lot for work, and I've seen these results consistent in many locations around the globe. Discrepancies are definitely the exception, not the rule. In other sims, they're the rule, not the exception. 

"very good results" and "fairly accurate" can't compete with predictive data combined with real-time data to fix the deficencies that do exist in predictive models, and will likely always exist. Metar is not sufficient to do that on its own, because its coverage is very spotty both in terms of location and time. 

And by professionals, I mean Meteorology professionals, not add-on developers. Some developers may be locked out of the platform, and while that's a pity, it doesn't mean it's arbitrary. It's obvious that Microsoft has its reasons. 

Edited July 1, 2023 by Abriael

[ShawnG 1,823]

6 minutes ago, Abriael said:

Incidentally, when you see a completely blue sky, it means you're somehow not receiving the data at all. I haven't seen it happen as of late, but I suppose it can. If there's absolutely *zero* clouds in the sky, it's very likely a data outage. 

this

[MrBitstFlyer 3,893]

3 hours ago, Abriael said:

If there's absolutely *zero* clouds in the sky, it's very likely a data outage.

The UK has had many, many days in May/June with zero clouds in the sky.