PMDG UPDATE

[David Mills 4,763]

I used to program for the old Radio Shack Color Computer.

[Noel 4,501]

2 hours ago, Rob_Ainscough said:

so what you're saying, there is no reason to improve simulation or graphic

Nope, what I'm saying is one man's 'improve' is another man's 'deteriorate' vis a vis making cockpits look more authentically worn, dirty, etc.  Somewhere skewed towards new but a little worn can be good.  Plus, I'm easily lulled into the suspension of disbelief already so increasing graphic detail and quality (not always realistic, ala poetic license) is always a good thing.

I have been wanting something decent to replace FSCaptain in P3D.  What's your opinion of Air Hauler 2 in MSFS if you've used it?  I used FSC for many years and really enjoyed the heck out of it, and never did any flights without it, and I do miss it.

[jrw4 624]

2 hours ago, WestAir said:

You could probably interpolate the cloud ground speed and ground winds and work your way back until the picture is complete, then let the physics engine run.

Creating a global 3D physics model of the atmosphere with sufficient resolution to account for thunderheads, etc., is a worthy goal, but I'm going to venture the guess that all of the supercomputers on earth working this problem wouldn't be adequate to the task. I would be happy to learn of such global models, however, but for now wouldn't recommend holding our collective breath. 

Best wishes for a safe and successful 2022/

[SKEWR 364]

Have you seen a real world commercial aircraft cockpit, they are dirty and smell like word not allowed!

The company I work with is still getting new 76 and 777s. We've all smelled the "new car" smell, but have you smelled the "new airplane smell"?

I do not like the extra caked on dirt look. I've even seen some folks request a dirty fuselage. From my experience, the old old (not to be confused with one old!) MDs get caked on dirt near door areas, but unless you're really up on the plane, you won't see the dirt at all.

 

Ok back on topic! 😀

[WestAir 2,522]

4 hours ago, jrw4 said:

Creating a global 3D physics model of the atmosphere with sufficient resolution to account for thunderheads, etc., is a worthy goal, but I'm going to venture the guess that all of the supercomputers on earth working this problem wouldn't be adequate to the task. I would be happy to learn of such global models, however, but for now wouldn't recommend holding our collective breath. 

Best wishes for a safe and successful 2022/

This actually gave me a thought. MSFS doesn't necessarily need the engine to simulate real-time weather. It could stream weather from a server with the computational power to do it right. The Earth is big, but not so big to make meter-sized wind resolutions impossible.

That zany idea aside, the wx in MSFS already does the job of simulating horizontal winds with ASOBO's physics models. You can feel the winds change as they pass buildings or hills, and the weather changes over time based on the starting variables. It just needs the other degree of movement added.

[jrw4 624]

7 hours ago, WestAir said:

The Earth is big, but not so big to make meter-sized wind resolutions impossible.

Check

https://windy.app/blog/what-is-a-weather-forecast-model-guide-on-forecast-models-all-around-the-world.html

for a nice review. This is exactly what MSFS did when it first came out, but the flight sim community objected vehemently to the use of models to portray weather, with the result being that a METAR-based system was deployed. Currently Asobo is wrestling with the complications that this entails.

 

[nikolajz1 90]

39 minutes ago, jrw4 said:

Check

https://windy.app/blog/what-is-a-weather-forecast-model-guide-on-forecast-models-all-around-the-world.html

for a nice review. This is exactly what MSFS did when it first came out, but the flight sim community objected vehemently to the use of models to portray weather, with the result being that a METAR-based system was deployed. Currently Asobo is wrestling with the complications that this entails.

 

This would be easily solvable if you could pick between metar and forecasted weather in the options. People playing on vatsim can use metar and be happy 

[jrw4 624]

23 minutes ago, nikolajz1 said:

if you could pick

And there's the question. To what extent does the weather depiction have to be baked into this enormously complex Earth simulator in order to achieve the spectacular atmospheric "scenery" that we all want. Given the apparent challenges that Asobo has faced in bringing to the community its METAR-based weather, do we want them to give us TWO methods that might or might not work on a given day?

BTW, one thing that also needs to be noted is that weather depiction is not just a matter of 3D physics, but rather 4D time-dependent processes. Even the best global weather models are run at lengthy intervals, and only a few times per day. So now one has the problem of somehow interpolating between the model predictions going forward and the present weather as represented in the METARs. Good luck with all that.

[JBDB-MD80 1,473]

Two different live weather options?... ok wait for that.

[JRBarrett 3,071]

47 minutes ago, jrw4 said:

BTW, one thing that also needs to be noted is that weather depiction is not just a matter of 3D physics, but rather 4D time-dependent processes. Even the best global weather models are run at lengthy intervals, and only a few times per day. So now one has the problem of somehow interpolating between the model predictions going forward and the present weather as represented in the METARs. Good luck with all that.

The model is run every 12 hours, but for each latitude/longitude grid point, there are 24 discrete forecast hours, so it’s not a case that the weather is “old”. If you are in Denver at 14:00 local time, you should get the predicted weather for that specific location and time. But, it is a prediction, not current data, so there always the risk that for any given location, the weather may be different than current observations. This is especially likely when the weather is changing quickly. 

[Noel 4,501]

1 hour ago, jrw4 said:

Given the apparent challenges that Asobo has faced in bringing to the community its METAR-based weather

LR has apparently come up with a method to integrate global and METAR weather in their upcoming XP12 release and along with it volumetric clouds.  It was be great to see another product's stab at this.   Right now I give MSFS an 8.5 out of 10 for scenery, 9 for lighting, and cloud depiction a 5.5.  If anything teases me into XP12 it will be poor cloud depiction in MSFS as I think the other areas, the SDK for example, will get there.  Not so sure about cloud depiction though it's a big challenge with this type of rendering approach over the former methods employed in ASN or other.  It appears from perusing the P3D forum that they haven't quite got their either.  I see so little cloud definition in MSFS, and often it appears clouds are rendered by stacking multiple parallel plates of low resolution 'cloud' texture together to make volume such that flying thru in certain lighting is just absolutely awful.  I think somehow they need to ramp up the density of cloud rendering considerably more.  The original cloud depiction was much better than it became I guess a couple Sim Updates back.  

Here is a shot that demonstrates this layering of essentially flat plates to try to generate volume, and while in the distance it can work, up close it doesn't work well:

And here's a photograph of cumulous--it appears we need much more density to get it much more right:

 

Edited December 29, 2021 by Noel

[abrams_tank 7,419]

36 minutes ago, Noel said:

LR has apparently come up with a method to integrate global and METAR weather in their upcoming XP12 release and along with it volumetric clouds.  It was be great to see another product's stab at this.   Right now I give MSFS an 8.5 out of 10 for scenery, 9 for lighting, and cloud depiction a 5.5.  If anything teases me into XP12 it will be poor cloud depiction in MSFS as I think the other areas, the SDK for example, will get there. 

I was just reading a Reddit thread about clouds in MSFS vs XP 12.  Here is the actual comparison picture in that Reddit thread with MSFS clouds vs XP 12 clouds (MSFS on the top, XP 12 on the bottom):

Suffice to say, almost everyone in that thread was commenting how much better MSFS's clouds are to XP 12's clouds. Well, this is not the final version from XP 12 but LR has their work cut out for them.

Edited December 29, 2021 by abrams_tank

[mtaxp 666]

13 minutes ago, abrams_tank said:

I was just reading a Reddit thread about clouds in MSFS vs XP 12.  Here is the actual comparison picture in that Reddit thread with MSFS clouds vs XP 12 clouds (MSFS on the top, XP 12 on the bottom):

Suffice to say, almost everyone in that thread was commenting how much better MSFS's clouds are to XP 12's clouds. Well, this is not the final version from XP 12 but LR has their work cut out for them.

And a few weeks later they looked:

 

 

 

 

 

So for now we only can wait and see, however the x-plane 12 weather engine is based on xavion which is tested and used in real flights

[laserit 93]

It's hard to accurately predict the weather for a weather man with the very latest up to date, weather radar and simulation software. What I personally look for and hope I don't see are unrealistic transitions. I haven't experienced enough long distance flights to judge the transitions, but what I have seen, I have been impressed with. I do have a very beefy rig, I don't notice any kind of transition.

[Kilo60 660]

Is PMDG releasing a cloud add-on?

 

I'm confused to see all this talk of weather in a PMDG Update thread?