Make sure to test some time before 07:00Z.
Today the change happened at 06:56z.
The change were not that massive today but where i tested it were not that much clouds. I’m not at the computer now, will upload a small clip of the change. I had a live web camera beside the phootage and clouds were really accurate but much too transparent compared to the real clouds.
Hope you can confirm the strange switch as well.
can you or any user seen it happen put down the numbers of when this change is happening, from what i understand it’s happening at around 0100z and 0700z?
All of this its because people with no real life flying experience though that METAR is the correct weather depiction and wrongly suggested to Asobo to somehow integrate Metars into global weather. This created a over complicated system with a lot of problems and a lot of complaints. That is conceptually wrong. We should have it the other way around. Keep the global weather and create a way for the sim to “read” a metar for that specific airport position from the simulator weather. Yes, that won’t match the real METAR 1-1 but that’s also how it is in real life (except temperature and pressure, but even if we had 1hPa differences and a couple of degrees off the real temperature, honestly what difference would it make to your simulator experience?).
This is an amazing community but honestly some ideas should be analyzed by some real pilots before consideration by Asobo.
So true! When the sim got released i made it clear to me that the weather in the sim was the real weather in the sim. I could observe it as i can with the real weather. I also knew that simulated weather were simulated with observations as a starting point for the weather that dynamically formed around me in the sim.
I can still observe the weather but now i observe unatural things thats not occure in the real weather. That make the weather so much more unrealistic now than what i had pre su7.
But i feel our opinion doesn’t matter. METAR is the thing in simulators and will always be. We that wanted something new and different and thinking more outside of the box are now forced to use the new METAR system anyway. Pre this METAR thing we could choose what we wanted to use with those 3rd party addons around.
Alright. It’s 0600Z at LKMT (OSR, Ostrava airport) (06 UTC, the same thing) and the shift happened like 5 minutes after I’ve started the sim.
Is it possible it has nothing to do with the time, but it just starts crunching some “new” data as soon as you launch the sim?
Just look at the clouds I started with. Beautiful stratiform clouds with precip either falling down to the surface or evaporating before it does so (basically the first time I’ve seen virga in the sim):
Then it shifted to the same old puffy cumuliform nonsense 
EDIT: Now that I look at the time stamps, the shift seemed to happen after 2 minutes in the sim.
Isn’t that about the time the sim takes to process METAR? It seems the “blending” only works one way, I.e. towards diluting all clouds.
Feels like it’s something wrong with the algorithms as someone said earlier. To me it feels like it’s being forced to be in that comulus state or something. As soon as we start looking at the amazing sky it shifts into the generic state. I would call this as a bug. I hope it’s not intended to work like this.
I’ve restarted my PC, fired up the sim again and it stays the same - cumuliform clouds instead of stratiform like it was when I first started it up.
Most definitely a bug. The sim is still able to produce beautiful true-to-life looking clouds. But then something happens that makes it all go to waste. Perhaps like @ConsularBuckle4 says, once the Meteoblue is done with integrating the METARs, it then injects the blended data which look very, very bad.
Yes that may be it. Then i would say get rid of the METAR blending of clouds.
This is how it looks IRL:
Now compare it to my first firing up of the sim:
That’s VERY close if you ask me.
Not this:
Yeah really close. I hope they now can escalate this issue to the devs. Maybe more needs to confirm this?
Yup. At this point I’d really completely ditch the METARs and only leave them for winds, temps and QNH.
They could also try injecting the visibility data from Meteoblue, but I doubt they have enough quality data for it to look good. The problem with snow still persists and I can’t imagine how the sim would look like with visibility data that poor.
Well, i’ve seen meteoblue fog before that switch and it looks amazing compared to the METAR fog. I search for a METAR with fog and then check on Meteoblue if it matches. If it matches i can see Meteoblue fog in the sim but only when that switch has not happened.
They didn’t use visibility data, or did they?
Only cloud data.
I’m talking about this map right here:
Well, it’s more cloud data. I’m looking at this cross-section for low level clouds and fog. Often that matches what i see in the sim as well. But only before that switch as we’ve seen. After the switch we can only see comulus/broken clouds.
Yeah, their very low cloud depiction looked VERY good. The only problem I’ve had was that the visibility inside the clouds was never low enough to reach CATIII criteria. METAR visibility can do that, but it doesn’t look good at all.
Agree. That applies to all the clouds. They are too transparent.
Exactly. But if you make them less transparent (more dense) then they’re too noisy and look like ash clouds.
I certainly don’t envy the devs to make the choices 
Agree, but the density feels more like a different tone in colour. Not like real density if you know what i mean.
Well, i can live with those fog layers they add as long as they not interfare with Meteoblue clouds and change them into generic comulus clouds everywhere as we can see in your photos.
Can you understand the need to change the clouds to make fog layer fit?
Visibility i can see on this. This is EDDM Both METAR and Meteoblue report low visibility here.
EDDM 290620Z AUTO VRB02KT 0400 0350 R08L/1600U R08R/0450N FG BKN002 15/15 Q1020 BECMG 1500 BR BKN005
