REX WEATHER FORCE For MSFS

[rstough 523]

7 hours ago, Bottle said:

Don't get me wrong, I'm all for 3PDs providing new things for MFS. Not trying to rain on anyone's parade. I'm just having a hard time seeing how this could be a significant improvement and concerned that some are getting hung up on the word "model" and wrongly think ALL of the MFS weather is just modeled. In fact, I think the concept is being used disingenuously by some.

 

Meteoblue divides up the globe into small boxes on the surface and then more stacked above. The Weather Feature Discovery video stated 60 layers of cloud data and 20 layers of weather data. The ceiling is 60,000 feet. I can't remember the number of surface boxes (lost somewhere in the mists of time in one of the interviews) but I'm pretty sure it was hundreds of thousands of surface boxes, if not many more.

There are people on the ground reporting METAR data for SOME of those boxes. Yes, that data will be accurate and a lot of that data, not all, will be regularly updated. But there are nowhere near as many observations as there are boxes. For the majority of the boxes the ONLY way to get an estimate of weather is for a model to be used. There's simply no other way but to rely on a model. It's either that or literally no idea. The model is based off of the current real world observations and like all models, its inference can be wrong. The Meteoblue video explains how they try to get it as accurate as it can be. But no-one would ever argue that the predictions will be 100% accurate (though they should be in the boxes that have observations). The famous quote by George Box is appropriate,“Essentially all models are wrong, but some are useful.”  I don't know, but would expect Meteoblue to have a pretty robust source of observations and would be updating them as often as they can (for instance their cloud satellite data is refreshed every 10 minutes).

It therefore seems to me that a system that just uses METAR data would be a vastly inferior one because it would have a fraction of the data that's possible with parametric modelling. Basing an assessment of the MFS weather on what's been available to date seems to be ill advised since it appears to have not been fully implemented (or bugged). Was it ever fully functional during the alpha?

If Asobo fix the weather system and say, "Yep, that's all of it and it's working as we intend it", and then it proves to be either consistently inaccurate or significantly out of date, then I'll look for an alternative. But the only thing I can see that's likely to be an improvement would be the refresh frequency. The weather scenarios is a nice feature though but since it's a canned scenario seems to be at odds with the people complaining that the weather isn't accurate.

What Asobo is reporting is what a weather Model does.  The GFS model creates 64,800 grids across the world based upon lat/long.  It models data horizontally and vertically for each of those grid points. It calculates what the weather will be based upon several factors: metar data, balloon data, aircraft data, ship data, radar and satellite data.  But the purpose of a model is simply that.  To model a forecast.  All those other items I mentioned is what "IS", a model is what "WILL" be.  

[rstough 523]

8 hours ago, DJJose said:

SkyForce 3D is one of the best add-ons I've ever purchased. I'm looking forward any this release.

Thanks Reed!

Thank you for your very kind words.

[Cognita 1,047]

Interesting discussion. My own view is that it is too early for this kind of development.  I am going to give ASOBO some time to work out any bugs or to refine any of the features before buying addons that address perceived limitations. I have been fairly happy with the weather implementation, noting the well known issue with winds, but this ASOBO has said this will be addressed and at this point I choose to believe them. The simulator has been out for just 1 month, the SKD  is not fully available/accessible, it is not even complete  (indeed, the simulator itself is not fully complete) there are acknowledged issues that need to be addressed -- I want to give it some time before deciding what additional programs might bring real, enduring value. I am being much more thoughtful about addons for MSFS.

I also think we should be thoughtful about not adopting approaches that worked well in FSX/P3D, developed in response to the limitations those platforms had, but may not in the long run be the best for this MSFS. To assess this I need to give MSFS time to develop, to at least get out of infancy and into childhood!

 

 

[sanh 823]

I think REX coming on-board is a good thing. Doesn't matter if you are buying or not. If you become fed up with default weather at least there is now a CHOICE.

Edited September 16, 2020 by sanh

[Nyxx 6,993]

13 minutes ago, rstough said:

Your choice, but I would like to address the lighting issue with EF.  That is not an easy fix and what made it worse trying to pin down was that for some it was fine, but others not.  So finding the middle ground is almost impossible.  I think EF is an amazing product for P3D v4.

Yes it is, but really you could not add an option to turn it on or off?

[rstough 523]

8 hours ago, FAZZ3 said:

Copying my comment from the forum.

---

I don't know. The clouds look extremely similar to the ones we have in MSFS. The only differences I see are 

1. More accurate representation of real world weather (which Asobo can fix themselves)

2. Smoother transition of different weather phenomena (which I think Asobo can themselves do, but I'm not sure if they would) 

3. Their source is NOAA weather (Idk if their data sources are more accurate than Meteoblu's, I'd wager it isn't)

4. A more Comprehensive UI

We'll have to watch this closely. We shouldn't accept an addon that offers default MSFS clouds in a more prettier packaging

Let me address the items you posted here:

#1 - Clouds will look similar because we are not changing anything there.  There are no cloud textures or models.  It is built off shaders.

#2- It will be hard for Asobo to fix since they are more leaning towards the model aspect.

#3 - NOAA and WMO is THE source of world wide weather data.  Everyone uses it as their backbone for creating model data.

#4 - Our UI will be growing to give you more control over the weather both automatically and even manually in a future update.

[rstough 523]

7 hours ago, omarsmak30 said:

Back to P3D, weather addone was a really easy sell but now, is not an easy sell. But this time really, I fail to see where it fills the gap, I mean the weather system by asobo is really good to me, even Active Sky is admitting this and they are still looking for a good use case for a weather add ones. So yeah, for me is just marketing with fancy UI. But definitely, I am looking forward for some in depth unbiased comparisons, maybe we are wrong, I don't know. 

Not marketing. This wasn't just build because we felt like it. We ourselves are weather geeks.  I have studied and followed weather since I was a kid. One of the MAIN reason I like flight simulation is because of the weather. I liked the weather system in the new sim, but it was not accurate.  So that is why we build Weather Force to make it accurate.

[jabloomf1230 2,445]

This is very curious because today HiFi stated this:

https://www.thresholdx.net/news/sept20

 

[rstough 523]

6 hours ago, danhenri said:

Hello,

Just a dumb question : for simply injecting metars in the sim, why just a plugin, like NOAA plugin in X-Plane, would not be sufficient ?

Not sure I understand your question.

[MarcG 2,894]

3 hours ago, B777ER said:

Depends on how long you are willing to wait. While they may fix their current version of "live" weather, it will still not be accurate until they incorporate METAR data sources. They are strictly using forecasted weather model data which lends itself to many inconsistencies in it's depiction. 

Then I'm fine with it, as long as it's realistic enough then I'm happy but Iunderstand others may want more hardcore weather apps 🙂

[rstough 523]

6 hours ago, Weatherman said:

Actually it's quite the opposite. Leaving beside the frequent data errors (in some countries more frequent than in others) METARS just give you the information for a certain point. That might not be a big deal if you have many many stations reporting, but for most (>90%) parts of the world the stations are way too sparse to get a good 3d weather simulation from it. The next issue is: METARs give you no information about winds in atmosphere, only at the ground. That alone is a deal breaker. Cloud top level is often missing as well, and even in a lot circumstances complete cloud layers cannot be reported by METAR if a dense cloud layer is below. Cirrus is not reported at all in most cases. Conclusion: METAR is great for giving a very accurate information about current weather situation at a specific airport. You use that for planning takeoff and landing for reasons. But for drawing a complete 3d picture of the weather and atmosphere, METARs are by far not enough. Frequently updated model output is far better for that nowadays.

We do use the GFS model for our upper level winds and temperatures and a few other elements to model the atmosphere. But the metar is the core of the rendering on our part. The metar dictates how the weather will be represented, the model fills in the gap when needed.

[rstough 523]

6 hours ago, travelabroad said:

+++1

We created our own SDK. We do not believe a weather SDK is in the near or possible future with the sim.  We understand why.  That is why we didn't wait.

[rstough 523]

6 hours ago, omarsmak30 said:

LOL, so the whole thing is just marketing it seems ? 😄

No it is not a marketing ploy. We have done what we said. 

[FlyingInACessna 656]

13 minutes ago, rstough said:

We do use the GFS model for our upper level winds and temperatures and a few other elements to model the atmosphere. But the metar is the core of the rendering on our part. The metar dictates how the weather will be represented, the model fills in the gap when needed.

A lot of the airports I fly at are on the west coast of the USA, where a thick marine layer is very common (OVC at around 500-700 ft AGL, tops around 2,000-3,000 AGL). The issue is that the marine layer creeps up on the water and may be sitting 2-3 nm from the runway edge, and the airport METAR reports clear but there's a thick layer of clouds just next to the airport. Would REX catch this and use satellite forecasting etc. to display those clouds? Or would the surrounding region be clear because of the METAR?

And another question - what is weather transitioning like? Is it realistic? Can you fly from one airport to another with two entirely different weather scenes without clouds fading in / popping in randomly? Does weather "drift" with the wind and develop naturally?

Glad to hear you're coming aboard as an alternative... nice to have options at the very least. I am still hoping live weather is fixed within the core sim though and has a decent level of accuracy, because METARs aren't necessarily accurate all the time. Ideally you'd have a nice balance between METAR data for just above the airport surfaces, satellite data, and forecasting combined to try and produce the most accurate scene.

Edited September 16, 2020 by FlyingInACessna

[Lenny777 299]

I've bought most of your products over the course of fsx and P3D. The issue you need to address is that your weather programs had poor transitioning. A persistent issue of abrupt weather transitions. We all know your main competitor who had this fixed some time ago. Has this been rectified?