The only reason I mentioned ‘random’ is because of what Seb said in the Nov 18 2022 Q and A: Seb - The drafts are – I’m going to give a little bit of history on the turbulence and drafts: So they were there since launch. At launch, we had automatic simulation of airflow in the world that was doing hills, slopes, houses, buildings, trees, and basically, obstacles. It was simulating the airflow around obstacles, and clouds were sort of pushing upwards in the middle and downwards, so it was creating some sort of airflow on clouds. Any other turbulence, if you were outside of the cloud, behind the air was purely random. It was, depending on the wind speed, there was some random turbulence that was generated.
endquote.
I fully expect the new 20km CFD to create some enjoyable turbulence that will not be random. I also find myself dreaming about what the new CFD may also bring but I realize I over-expected when the sim was released as far as global haze and its effects such as ever changing sunrises and sunsets.
But it seems to me that if you are going to create either a 10 sq mile cube (spheroid?) made of thousands of small cubes and each cube has weather related data then maybe clouds of all kinds can b built out of these cubes and that of course includes thunderstorms. What stopped me though is wondering what happens at the border of the CFD simulation? What clouds get created there and why? Maybe it is all about the winds and nothing more and the clouds we have now are all we will get…kinda depressing that thought.
It isn’t even possible to send audio and video communications around the world without multiple minutes of delay. getting real time data and then converting it to a usable form placed in a place for all to download to their sims (which will also take time) cannot be done without some delay.
I think you need to allow for some delay.
I remember that What i think Seb mean by random is that they can’t simulate the horizontal wind velocity through cfd because it would not match those sources we use to plan a flight with. On a METAR for example it doesn’t tell the vertical wind velocities that means they can simulate that aspect in CFD because nobody can say it doesn’t match METAR or MB. For the horizontal winds they need to add what MB or METAR reports and make the air fluctuate between the average and gust value randomly. It’s not fully simulated gusts/horizontal fluctuation of air. It’s added on top of the CFD. But we can’t know for sure how it works under the hood. But i’ve noticed the horizontal airflow can’t go below the average reported winds on METAR or above the gust value on METAR.
In my opinion that randomly generated gusts is more realistic than no gusts at all. Could they improve those? I’m 100% sure they can as with all other aspects in this flight simulator.
A thing to note is that air is one thing. IRL the air moves because of convection.
What i have understand is that Asobo wants to have a full convection simulation in the future with that 20km CFD.
But that will only work when we are not flying near airports i think. Near airports we need to fly in more fixed atmosphere.
Of course I think some delay is realistic. I’d like to see it get down to 10 minutes or so and have quite a bit of fidelity of shape, specifically for precipitation. With modern weather radar, that’s not unrealistic.
In one of the dev q and a sessions, Martial mentioned using satellite imagery for cloud placement. I don’t know if anything ever became of that – whether it was dropped as an idea or if its still being pursued – but combining real time imagery and radar data should be able to produce real time weather. Its just that their current scheme of getting data from a bunch of sources and then massaging it for use by the sim…all that makes ‘real time’ anything less likely in my opinion. I wish them luck.
Depending on the mode (visual vs IR), satellite can inform several things about clouds, but it’s not good for precipitation. Radar will be best for precip, even though it also has limitations. As you say, a combination will be best. But I wouldn’t mind the whole thing stitched together with a 10-minute continuous delay, so it moves organically as the data update.
Actually, it would be nice to simply know what kind of delay and resolution/input we’re currently dealing with. That alone would help with sim decision-making.
When they added the METAR data that updates every 30 minutes or the new MB model that updates every 6 hours instead of 12 we noticed many more switches in weather. If they add updates every 10 minutes why would that be different? Would make it even more noticable with even more sudden changes of weather. Can’t have the weather move organically and at the same time have it stay accurate. Will be more like a slideshow instead of dynamic.
This need to have it match realtime weather makes the weather feel less like weather. Sattelite data will look equally static as those sattelite images they have because the sattelite images is not recorded in real time. It’s a single image every 15 minutes.
Either we want it static and accurate or we want it to behave like weather/atmosphere and adapt to the weather that occures in the sim as we need to do when the weather changes IRL too or we have the weather fixed in place like those sattelite photos.
Simple. You must to be able to use aviation weather tools to execute realistic flights. Period. Without it, what we’re doing certainly looks like aviation, but it’s missing a huge part of what really goes on. We pay so much attention other areas of realism like nav instruments and runway numbers, etc, but those are absolutely predicated on necessity due to weather. In a world where it doesn’t matter, you don’t need even those things we consider basic. Why support that and not the other?
So unless they can provide those tools, we have to hope for as close to synchronicity as we can get so we can continue to use real-world tools. I’m willing to accept a small amount of lag if it means it’s more granular and accurate because real-world tools, as we’ve discussed, also lag a bit, so flying in a world that’s 10 minutes behind is not a big deal. But when you get to 30-60 minutes or more, it is.
The computational power to just feed it starting parameters and let it smoothly run and develop on its own doesn’t currently exist and the accuracy of the zonally-injected emulation over longer periods was simply fantasy, full of easily-provable holes, especially early-on. It just looked good (and I don’t know why it still can’t - I believe that to be a sparsely-related issue), sometimes, but it didn’t behave correctly and it was absolutely unpredictable and unobservable beyond what was visual, which is antithetical to modern aviation.
Yes, you accept that. But not all of the users do. Those sudden changes and bubbles and stuff feels more like fantasy to me. That thing doesn’t happen in the real atmosphere and i want the atmosphere to feel like an atmosphere in a simulator that tries to simulate the thing that aircraft use to fly. Those planning tools you are talking about doesn’t fly. The mb model pre METAR felt much more like a real atmosphere and it uses real weather as initial conditions and then it could behave like weather until they i updated it with new initial coditions every 12 hours. Sometimes it matches the real world and sometimes it didn’t because weather IRL is unpredictable.
I don’t want to change your opinion. I understand yours and i hope you understand mine. And i hope Asobo start to understand there is more than one opinion about this and start implement more options for the live-weather. Because you wouldn’t care if i used the more fluid system and i wouldn’t care if you used the more accurate system.
I will never eschew the use of aviation weather products when using a sim in a meaningful capacity. It makes absolutely no sense. We might as well just fly on rails at that point.
Real world weather is absolutely not unpredictable on a meso and synoptic scale, and even a storm scale, once they get going and stabilize. That’s a complete fallacy or misunderstanding of weather altogether. And further, it’s observable using multiple methods. It’s not perfect, but it’s really, really good.
So we’re at an impasse because of a misunderstanding of the accuracy, availability, and importance of forecast and observation tools in a simulator that’s supposed to… simulate.
Turbulence is far away from spot on. Being thrown 30 degrees to right/left during clear skies setting in an A320 at 1000ft on final in Cologne is pretty much unrealistic. Thermics turbulence yes, but not with that high intensity.
Yesterday I tried landing 2 times at LGKY on runway 10 with the TBM 930.
But on the final, there where great fall winds or wind shear kind of turbulence that almost crashed my plane, as it pushed it down and to the side. Then I tried to land a second time and this happened again. This while the METAR said there was only around 10 knots wind or so. After that I diverted to LGKO.
If you see how LGKY runway approach 10 is located, you see the mountains next to it, I guess that must have caused this severe turbulence.
But without being a pilot, I judged the turbulence effect to be to strong. The TBM 930 is a not the lightest plane out there, and it was not like 30 knts wind conditions or so. It was around 10 kts or less.
Some turbulence could be there because of the mountains, but by plane was pushed down like meters on the final.
Are the fall winds, wind shears around mountains still to strong in MSFS? Does anybody have a educated knowledge on this? Has MSFS stated something that they are going to update this?
First of all: Thanks a lot hat you mentioned that extraordinary landing strip !
A few minutes ago I flew the Savage and made 2 traffic circuits in real weather at 15.00 local time, when I would expect thermal turbulence be strongest.
Turbulence settings was real. Sadly, I did not encounter any turbulence at all.
To your questions:
Generally speaking, compared to real, the turbulence model is already phantastic.
What you encountered, you could have encountered as well in real.
In real, I would hardly dare to land at this strip in turbulent weather conditions before talking to an experienced local pilot.
This strip is something very special and interesting to practice !!
One suggestion: Try to land there in strong headwinds and be aware of the very strong
downwinds. You’ll make interesting experiences !
Landing in strong sidewinds will be too challenging.
I only fly PA28 IRL, but in my opinion, the turbulence over land is one of the things that are actually simulated very well. On the realistic setting, I don’t find it exaggerated at all.
At any angle to the wind I regularly get knocked into a 10 degree bank, occasionally a 20 degree bank and regularly knocked up or down 5 degrees of pitch. The plane will regularly start climbing or sinking between 700fpm and sometimes even 1000fpm. Commonly the nose will get knocked in the opposite direction to the vertical speed change, e.g. the nose will get knocked down and the pane will start climbing or vice versa. This all happens half a dozen times a minute.
I see these kinds of posts in other flight sims. The truth of the matter is that the sky is a rather bumpy place. If you have real world flying experience and you only pick beautiful clear and smooth days to fly, then you may be shocked. I don’t know. But if you fly regularly, especially in summer, you will get beat up. This is particularly bad below the clouds. There it is hot, bumpy, and humid. It is not uncommon to have the plane thrown around quite a lot depending on the wing loading. Light planes have it the worst.
Contrary to what most simmers want, planes have bugs on the windscreen and there is turbulence. That’s life.
No, the real truth of the matter is that in real life the effect of turbulence on a plane is mostly vertical as the plane passes through air that is rising or falling at different rates, and a trimmed aircraft will return to straight and level flight after a deviation.
In MSFS neither of those things is correctly modelled, so while the degree of turbulence in the sim may be realistic, the effect on planes in some weather conditions can be extremely unrealistic.
Also you only have to activate the Show Thermals settings and see the difference between 0kts of wind and 3kts and it’s immediately obvious that something is not right when it comes to how thermals are modelled.
I feel like a broken record saying this. I have seen nobody in these forums arguing for an absence of turbulence or weather.
Conversations in these threads always seem so one-sided, with one group of people pointing out suspected peculiarities about turbulence or thermals in the sim, odd flight behavior (i.e. tail wagging with virtually no other turbulence), and sometimes even citing the known forms of lift and thermals that have been implemented in the sim along with their associated problems.
Meanwhile, the other group is reciting the same blanket statements like, “Turbulence exists; deal with it,” or some sardonic comment like, “What’s next, getting rid of gravity in the sim?”
What i think Seb mean by random is that they can’t simulate the horizontal wind velocity through cfd because it would not match those sources we use to plan a flight with. On a METAR for example it doesn’t tell the vertical wind velocities that means they can simulate that aspect in CFD because nobody can say it doesn’t match METAR or MB. For the horizontal winds they need to add what MB or METAR reports and make the air fluctuate between the average and gust value randomly. It’s not fully simulated gusts/horizontal fluctuation of air. It’s added on top of the CFD. But we can’t know for sure how it works under the hood. But i’ve noticed the horizontal airflow can’t go below the average reported winds on METAR or above the gust value on METAR.