Well guys, in the latest DEV interview Seb basically said that it would be very difficult to get rid of the volcanic clouds, so they know about it, but I cannot see an improvement coming anytime soon…
What has been mentioned is that it was dificult to change the color of those volcanic clouds. But I think that if in SU10 they succeed to improve clouds density by layers as they mentionned and if they also improve clouds depiction, those volcanics clouds should not appear again, or at least just in zones where they are justified like storms or tropical zones. For me those volcanic clouds are the result of different layers gouped together and are not true cumulonimbus.
3 Likes
It seems like this transition that often occurs some time after first starting a flight is one of the big culprits to the cumulus cloud / “volcano” cloud issue, unfortunately it’s getting buried among the other discussions going on this single post.
Is there another post already focused on this abrupt transition? If not, might be worth making one.
2 Likes
Yes, i think either the servers really need a big overhaul if the servers causing it to stop rendering as it should. If it’s METAR and they can’t make it blend as they said it should then revert back. They never said anything about a new hard transition feature in weather pre release of su7.
The guy above me posted the bug thread about that issue. I should have posted it there i think. I don’t know whats causing it and METAR was the thing that has been implemented in su7 that is the thing that brought hard transitions to occure. Thats whu i posted it here. Never seen any kind of hard transitions pre su7 without METAR. I never seen those occure IRL either. Have you?
Here is a screenshot demonstrating something I’ve been talking about for a while: How a perfectly accurate METAR report can quickly become less accurate than meteoblue due to a lack of future information. When this METAR report was taken (the bubble of clear skies), this quickly moving system was not there yet, and there was no way to know it was coming because there’s no future information in a METAR report. Since Meteoblue data includes future information, it can smoothly flow\transition into the future at all times without having to wait for anything. It doesn’t matter how fast the system is moving in RL, it can theoretically always be depicted at the proper speed in MSFS if it’s coming from Meteoblue. Unlike with METAR where there is no future information and you have to wait until the next report before you can do anything, so a fast moving system can fly right over head between reports, or just be depicted in a very choppy way.
6 Likes
Well, i really hope Asobo can understand that we don’t like that kind of weather. Some does but not everybody. I just want Meteoblue even if it’s 12 hours old. I know it’s live anyway. Everybody uses the same weather in the sim. You can see the same weather as i do when located at the same location. Thats live-weather to me.
3 Likes
This is precisely the reason why i came up with the idea of sending back sim “live weather” 24 hours into the past. The sim weather would then know what happens within the next 24 hours and can accurately anticipate. Why 24 hours and not 4 or 2? Because with 24 hours you can also accurately depict time of day typical weather (a typical morning fog in november or typical afternoon thunderstorm in the tropics).
Obviously it would cause problens with VATSIM if VATSIM wouldn’t follow this idea. You would think there would be no difference for online pilot if vatsim weather is current or 24h old, as long as all online pilots are flying within the same weather.
I mean a 4th may weather doesn’t differ much from a 3rd may weather on any given location.
But as long as people insist real world weather in the simulator has to be “time” synced with real world weather outside this idea will probably not be investigated.
I may obviously be wrong as i have no expertise in software engineering. I do strongly feel that the accuracy AND realism AND weather dynamics in the simulator can greatly profit from this.
I for one really don’t care if I get a 3rd May report when flying on 4th May.
Really the only people that would suffer from this are those that have a thing of flying a flight side by side to its real counterpart. But is it worth it sacrificing potential realism in simulation and immersion for this?
2 Likes
Agree with this. But old weather is static and nobody records old weather perfect. Better to use predictions. Those are simulated and calculated. But both opinions is as much valid. Because predicted weather would make the weather behave accurate/realistic and those that want more accurate weather will just have that (accurate) to METAR nothing else.
I preffer predicted weather for sure. But if it’s calculated 12 hours ago doesn’t matter to me as long as we see same weather in the sim. Simbrief uses predictions too to calculate fuel usage. The only thing pilots uses METAR for is to plan landings and takeoff and weather changes so fast that while you are taxing or on final the METAR maybe is not accurate anymore. It’s only really stable weather conditions that may not change but stable weather is more easy to predict. With a METAR we can only predict the weather ahead we can’t know for sure what will happen using that because weather is unpredictable. And i want the weather in the sim feel unpredictable as it is IRL.
VATSIM users only need to send their altitude correct to those controlling. And Vpilot has fixed a workaround for that thing. No need to create weather using METAR. The work they done implement METAR to me is a waste of time. They could have focused on make sure the data Meteoblue gives Asobo is injected correctly and could have implemented other data from Meteoblue instead like
dew point
humidity
or why not the Meteoblue visibility instead.
Visibility from Meteoblue would be much more realistic than these METAR circles around airports. Weather doesn’t behave like this.
1 Like
Its more than just the hard weather transition though. It seems like when the transition happens (if thats what it is) all clouds re-render as cumulus. They even maintain their general position, as shown on your images.
3 Likes
Agree, thats why Asobo needs check their data and make sure it’s rendered correctly. We have had many issues with data since release of this sim. Those issues caused them to implement METAR instead of fixing the real problem.
I really hope Asobo starts to just focus on the rendering of the data they get from Meteoblue. Now it’s even harder to fix it because they have many different sources of weather, 1 Meteoblue source and around 10000 of METAR sources around the world.
For us as users it’s impossible to know whats causing issues now. We don’t know how the system is supposed to work. Before METAR we knew the weather should have matched only Meteoblue. Now it should match all of those sources at the same time? I don’t really know. I trying to make Asobo aware. I know a guy that has tried to make them aware for a long long time. This guy @utd4life12
I posted it in the transition topic as well for a hope to get them fix it.
3 Likes
I wonder if the high clouds are really being represented. They are supposed to come from meteoblue data, but on many occasions I have realized that they are not present in the sim, but they do appear on the meteoblue maps.
For example here in Santa Cruz, Bolivia the metar is as follows:
El cual no toma en concideracion las nubes altas que aparecen en meteoblue y que estan presentes en la realidad.
In the end, the vast majority of the time in the sim, only one layer of cumulus clouds is seen in a noticeable way and if there are other layers, they can be seen with great difficulty.
4 Likes
Agree. Even seen many examples where the higher clouds are drawn on the MSFS world map but then go to that location and they’re not depicted correctly or at all in MSFS itself.
2 Likes
From SU7 until now there is almost always a predominant cloud layer (cumulus). Before there was much more variety of layers, heights and shapes.
9 Likes
Yep exactly. I’m seven seeing these rogue cumulous clouds at airports reporting CAVOK or NCD, I’m still seeing over done cumulus well below 5,000ft. I did actually think it had improved, but last couple of days It feels like back to SU7 again in areas that should be higher overcast are just a mess of cumulus
7 Likes
To illustrate for the Devs what’s still need to be improved: approaching LFRT METAR, clouds look like volcanic, kind of end of the world !, in a region that is not supposed to get this such extreme weather. METAR is reporting 2 broken and 1 overcast layers. It looks like all layers are grouped into 1. Are you checking that bottom and top of those layers are not mixed up ?. How do you proceed with METAR to evaluate how thick each layer is, since only bottom is reported ?
METAR integration with Meteoblue data still need improvements to display realistic clouds depiction and environment.
Does reported METAR time match MeteoBlue predicted time ? If not, it is highly probable to get hard trsnsitions because reported weathers are not matching at all due to time difference.
5 Likes
This is a classic example of how bad the weather can look.
3 Likes
They already explained that the weather system is in fact a SYSTEM on its own. SEB tried to tweak the shadows etc., and it made no difference.
However - i fully agree, the cloud rendering is completely off. It just looks horrible. Sometimes it renders a nice cloud scape, but mostly its just a mess of lumped cumulus clouds with horrible shading.
After the last DEV update video, I am now doubtful that they will be able to fix this. I think this is how weather will be in FS, unless they completely redo the weather engine, which they already hinted will be too much work.
I find it perplexing that they changed/regressed it so much. Not only are the clouds rendered incorrectly, they now change abruptly, which is something that never happened PRE-SU7.
I cannot understand why they simply cannot revert the changes made to weather PRE-SU7, or at least give the option to switch to that model.
8 Likes
I fear you are right, but not because they weren’t able to fix this. They were, if they wanted to.
It was a - let’s call it strange - section within the Q&A: a lead developer who seemed to be surprised that the weather within MSFS is indeed a “system” and consequently was not able to explain anything valuable in this regard, assisted by the head of the company who admitted to have failed trying to “make the clouds brighter”.
And I asked myself: who implemented the weather in the launch version, where it wasn’t 100% accurate (which it isn’t now) but at least looked stunning AND realistic? And I further asked myself, weren’t these the same guys who changed the rendering with SU5 to gain so much more performance that I could run the sim on my Iphone 6s if I only wanted a flightsim with a cartoony sky?
But rant aside and back to your point: To my understanding Seb made it very clear that they could change the appearance of the clouds, but only with a performance penalty. Remember pre-SU5 clouds? No ash look, less contrasty and hence more realistic, and gradual differences in lighting even when you were flying within a layer.
So they have to know EXACTLY how to get a better looking sky, but that would probably mean performance problems on X-Box by reverting to a pre-SU5 rendering. And that will most probably not happen. The best we can expect will be more “tweaks” of the subpar weather rendering we got with SU5 and SU7 (which added ridiculous cloud depiction to downgraded visuals).
If SU10 will not make a huge difference, we will probably have to accept that a key feature of this promising game has become a major flaw.
5 Likes