I haven’t done a timelapse myself yet, would love to see one with the newer weather blending.
most abrupt changes are mostly winds atm tho, they fixed the aggressive changes a couple SU’s ago.
(by aggressive i mean, from clear weather to suddenly overcast to suddenly broken etc in a single second)
2 Likes
Yeah, those darn microscale changes again. And a singular METAR isn’t comprehensive or granular enough to reverse-model what’s actually generating that change, just that it’s occurred. Is it a frontal passage? A localized storm? A diurnal shift? One hour between observations (and the data that are sent/used) isn’t enough for those small-scale types of changes.
So we have to either keep them on “rails” generated over a larger forecast period and ignore them (and thus not use real-world observations in our decision making), or incorporate them and take the sudden change.
I wouldn’t mind if there was some AI blending, delayed by 15 minutes or so (I mean, observations are old by the time we get them anyway). So an observation made at T+00 would shift gradually to T+15, then the next observation at T+60 would gradually take effect from T+60 to T+75 and so on. Maybe it’s already doing that, I don’t know. Would need to collect data.
the previous method was basically using meteoblue forecast data mainly which was amazing.
1 Like
For forecasts, yes. But I cringe a bit when considering using forecasts for operational weather because forecasts, depending on how far out you are, produce diverging results over time, that at some point have to be re-integrated because of the downstream effects the actual weather has had on weather that will be included in the next forecast period and so forth. It’s chaos theory at its finest example.
I mean, observations themselves are used to create the next forecast, even in the rapid refresh modeling. So it comes back to how granular those forecasts were and how often they were being refreshed, which to me seemed way off from anything approaching reality, except in how it was rendered. That was good.
It seemed like they threw the baby out with the bathwater when they re-did it. I always wonder why it wasn’t a little from X and a little from Y. Keep what worked, toss what didn’t. But again, none of us know the deals between the parties and how the engine used whatever data were being sent.
3 Likes
THIS!
I often wondered if the person responsible for the interpretation of the forecast data in the release version left Asobo when SU7 hit , because I couldn’t imagine that the raw data delivered from Meteoblue after SU7 was so much worse than before, and yet the result was atrocious.
But maybe it’s that simple!
3 Likes
sometimes i really wonder if any of the dev team members or jorg or seb martial etc read the forums and topics like these 
3 Likes
I think this problem is quite evident, in most cases the clouds are sort of messy. It is as if it were not possible to produce layers and everything was mixed into a single mass of deformed clouds. It’s true that sometimes it looks spectacular, but most of the time it doesn’t.
Also, before the weather always changes and there were beautiful and different clouds. Now everything seems to be a copy, wherever you fly the clouds look the same.
12 Likes
I am 100% sure they don’t!
I know they keep saying it, and maybe Jorg and Seb are skimming through some topics from time to time, but otherwise they’ll for sure depend on what the community managers present to them in a condensed form.
I’m o.k. with that, as opposed to trying to pretend otherwise…
3 Likes
The predicted weather vs metar debate has been going on since August 2020. Yes, the sim used to look better. But as someone that uses real world data to plan my flights, I’m ok with the trade off. Rarely is the weather substantially off from what I am expecting, and it’s certainly better than it was even last year.
This is the compromise that was made, and whether you like it or not, it’s the direction that they decided to go in.
1 Like
Honestly they could’ve just made an option to disable the metar blending 
6 Likes
Xenviro does a much better job at representing clouds just a shame its limited by weather SDK
1 Like
I can’t speak for xEnviro but i have the full REX package as i like their trees addon etc.
I tried the weather app a couple times, it’s unfortunate it affects the entire skybox but sometimes it depicts weather more accurately (in that region).
But i prefer native asobo weather, i just want them to improve on this.
2 Likes
Hi Moach6908,
first of all, thanks that you guys care about improving the sim !
IR I loved to fly in bad weather conditions with a Piper PA 18 and a Buecker 181 and also test the limits ( as long as we had our own airfield) .
So I love to do that in the sim with “Real Weather” selected:
For me the cloud scenery (not only !) is especially great during December in Iceland, in March and April in South England and Caernarfon in North Wales + winds of 30 to 40 knots.
(I learnt a lot flying in these conditios with light airplanes in the sim !!) .
The cloud/weather scenery is so convincing that I don’t see any difference to flying IR.
We have multi layer clouds, different types of cloudiness, haze, fog, rain showers with visibility modeld extremly well, and the list could go on lengthy.
I don’t use any special settings in the sim other than volumetric clouds set to max.
The other setting is with my display: I use max brightness respectively HDR.
For me the most effective, most important setting to get this unbelievable
immersive feeling when flying in “Real Weather”.
As to the gusts in “Real Weather” :
Enjoying /training landings in bad (very bad) weather conditions,
I often use Shoreham (South England).
The modeling of gusts in the final in Shoreham is superb.
Yet the modeling of the lee effect for airfields lying on a plateau
until now is still missing, I guess. You’ll get that too,- I’m sure.
Gusts in clouds:
I have no IFR licence, but once could not withstand to try out flying with my Buecker into a strato cumulus years ago. For ten minutes the ride was without any gust, not the slightest one.
But then one did hit me so hard that I was so surprised that I could not tell at first did it go upwards or downwards. A few seconds later I fell out of the cloud with 60 degree bank in a spiral downwards with 150 knots.
What I want to say is:
You don’t encounter gusts in clouds (especially strato cumuls) all the time.
And you don’t have icing conditions all the time.
My opinion, experience, after all:
With me, the cloud scenery in “Real Weather” is more than stunning to me,
I’m thrilled with that, what I have with the sim already now. I really enjoy Real Weather most
because it is so real, diversified and dynamic.
A non simmer won’t believe that until he has tried it himself.
Bye,
Walter
Good luck!
2 Likes
Yep, i’ve always wondered why the skies have had no real improvements. Both graphically & in the lack of cloud types. I get that it’s much easier said than done but i expected to see decent improvements in MSFS 2024 which i didn’t in the trailer at least.
3 Likes
Other than unfortunate side effects from METAR blending, the only other reason I can think of for having less pleasing skies now than before is for performance reasons. That’s why I’m hoping that 2024 magically restores some of the atmospheric “fidelity” we lost with SU7, but that leaves Asobo in a tricky position.
Refusal, or inability, to backport any improvements to 2020 would, to me at least, hint that there would be negative effects from doing so.
I see myself switching entirely to 2024 when it releases, so I have no vested interest in having 2020 improve at that point as it would be old tech, just theory crafting on why companies do the things they do, and why they often fail to explain their decision making process with explanations that ring true.
1 Like
IMHO Asobo has done a fantastic job with the weather! I live in the present. Do I remember what something as subjective as weather was in the game 4 years ago? No. Can I say Asobo has improved game weather YES! Can dismiss the non constructive criticism of zero improvements? YES.
4 Likes
Perfection does not exist!!
And even more so when something is seen from different eyes…
If you are right, 4 years…
And in that time I think that no simulator has been able to stand up to MSFS 2020 in these 4 years (I think they have had enough time)
Microsoft Flight Simulator has been the one that has taken a great leap and has put everything on the table, risking everything so that we can enjoy it to the fullest…
And now we are going to judge him unfairly for everything…?
For my part, I think they have achieved something wonderful and which I am enjoying to the fullest.
That more things will improve…?
Of course, give it a try.
That MSFS 2024 will be much better, I trust that they will take another step forward.
Let’s all enjoy what we have.
Happy flights!!
3 Likes
All I miss from this fantastic simulator is the presence a cloud has flying right next to it.
I want to fly around the cumulus clouds all day long, but they are to diffuse/vague ( spelled right?) to really feel like its an object.
Flying in vr in ace combat on psvr it gives you that flying in the clouds feeling, its fantastic. But of course its a performance thing, just wish they found a way of making the clouds more 3d objects when really close to them.
2 Likes
this topic is the reason why I stopped playing mfs 2020 a year ago. I just came into terms that they were not going to restore the cloud variety we had 3 or 4 years ago! It’s so immersion breaking to fly in different parts of the world only to see most of the time the same clouds. Flying in Greenland in winter and being surrounded by cumulus like volcanic hashes? really? no thank you!! I will wait and see what mfs 2024 has to offer, thou from the trailer of nearly a year ago now, the clouds looked the same!
9 Likes
I have high hopes on that.