[SOLVED] METAR and SMP don't mach

[Tompolth]

Hello, 

I've upgraded from V4 to V5 and I have just loaded Xplane in my home airport to test if everything is as expected. 

I am using Xplane 11.51 vulkan, SMP v5, RWC (last version) and FSGRW. As you can see in the pictures according the METAR (and what I am seeing from my window) it is raining and cloudy. However SMP is not depicting what it should. 

The rain is not visible at all and  the plane is proyecting shadows as it was in summer. 

Best

[Cameron]

Hi there!

It's possible that an old METAR was present and loaded prior to the new one being loaded. You can:

1. Go to the SkyMaxx Pro menu and force a weather re-load

2. Adjust the Real Weather setting to not change visible weather (I personally like keeping this active though)

 

You also need to make sure you have FSGRW selected in the Real Weather settings of SkyMaxx Pro.

[Tompolth]

All of those options were tested :-(

I've even deleted the rwx files to be sure the new files were loaded. 

If I disconnect the SMP and RWC and then connecting only SMP the weather is more realistic but with too much fog (there is no fog neither the metar nor here). However as soon as RWC is on (and I refresh weather from FSGRW) the sunny day appears again. :-( 

[Cameron]

1 minute ago, Tompolth said:

All of those options were tested :-(

I've even deleted the rwx files to be sure the new files were loaded. 

If I disconnect the SMP and RWC and then connecting only SMP the weather is more realistic but with too much fog (there is no fog neither the metar nor here). However as soon as RWC is on (and I refresh weather from FSGRW) the sunny day appears again. :-( 

Can you please attach the FSGRW SMP file in your X-Plane folder, and to confirm, LEXJ is the airport you're at?

[Tompolth]

Here it is (two rwx files are generated). Anyway, I've just disable volumetric clouds and now it is working properly. Even though the plane is casting shadows it was a sunny day

fsgrwsmp.rwx METAR.rwx

[Cameron]

So you're saying with SkyMaxx Pro enabled, but volumetric disabled, the expected scene is present?

As for shadows from aircraft, we actually don't control that in any way. That's simply X-Plane casting the shadow on its own.

[Tompolth]

Yes, with SMP enabled and disabling the volumetric clouds the scene is right, but rain is not visible 

 

 

[FlyAgi]

Also, with the volumetric clouds the cloud height seems to be off, mostly the clouds are to high and the lower cloud layers are not present at all.

[sundog]

16 minutes ago, FlyAgi said:

Also, with the volumetric clouds the cloud height seems to be off, mostly the clouds are to high and the lower cloud layers are not present at all.

At present there's no good way to blend volumetric clouds with terrain, so we enforce a minimum height of 500 meters (AGL) on volumetric cloud layers for now. If it's below 500 we push it up to 500. That's probably what you're seeing. We do hope to refine this in subsequent releases.

 

Edited January 23, 2021 by sundog

[FlyAgi]

Well... actually not, the clouds are placed much higher than I would expect even if the metar does not report them extremely close to the ground. So far the clouds where always higher positioned than in the reports for me (with a Robinson R44 helicopter I could not get above any volumetric cloud layer so far).

Also, even if I tried to force the clouds come down a bit using a manual weather setting the tops still were hardly reachable for me (using the R44 helicopter again which is limited to about 13.000 ft). Maybe this is because the clouds are larger/thicker/higher than the old ones but it feels to me so far as if the cloud base is always to high.

 

 

[sundog]

5 minutes ago, FlyAgi said:

Well... actually not, the clouds are placed much higher than I would expect even if the metar does not report them extremely close to the ground. So far the clouds where always higher positioned than in the reports for me (with a Robinson R44 helicopter I could not get above any volumetric cloud layer so far).

Also, even if I tried to force the clouds come down a bit using a manual weather setting the tops still were hardly reachable for me (using the R44 helicopter again which is limited to about 13.000 ft). Maybe this is because the clouds are larger/thicker/higher than the old ones but it feels to me so far as if the cloud base is always to high.

 

 

Hm... not sure why that would be, but I'll look into it. Are you using Real Weather Connector?

[FlyAgi]

Yes, I'm using RWC 1.2.

[sundog]

9 minutes ago, FlyAgi said:

Yes, I'm using RWC 1.2.

So there is also some logic that will move volumetric layers around to prevent them from intersecting with each other, which might also be a factor. If there's a specific METAR report you're seeing this with it would be helpful to see it so I can reproduce what you're seeing.

[FlyAgi]

I will do some special tests in a very flat location and report my results later on. :-)

[FlyAgi]

I did another test and this seems to confirm my observation. Weather comes from FSGRW (historical, this afternoon, files attached to this post).

 

Reported cloud base by FSGRW (confirmed in XP weather menu after injection):

Few, 2200 ft
Broken, 4600 ft
Broken, 28675 ft

 

 

 

 

Field elevation (EDXE Rheine-Eschendorf, flat area): 130 ft

 

No cloud base at 2200 ft, no cloud base at 4600 ft but at about 5800 ft I start touching clouds which makes not much sense to me (500 m = 1640 ft so 2200 ft for the cloud base should be outside the safety margin).

 

 

 

 

 

 

wxdata.zip

Edited January 23, 2021 by FlyAgi

[sundog]

Thanks, that's helpful. I'll look into it.

[sundog]

OK, so part of what is happening here is that the cloud layers at 2200 and 4600 feet are intersecting. Volumetric cloud layers cannot intersect, so we discarded one of them. However I think discarding the lower one is a poor choice, so I'm going to change that for the next build I do (that may not be the next release though, as I've already handed off a couple of bug-fix releases to X-Aviation)

So the remaining issue is the layer at 4600 feet appearing to start closer to 5600 feet for you. That's at least partially a bug where we were applying a correction intended for non-volumetric clouds to the volumetric layer, and I've fixed that too.

[FlyAgi]

I have checked manual weather and there it seems to be fine. Also I had another try with XP real weather and there again, the lower cloud layer was not present but only the higher one, in general it seems as if I always get only one cloud layer using real weather (with RWC all the time) and the lower cloud layers are always missing.

 

[FlyAgi]

1 minute ago, sundog said:

so we discarded one of them.

In case of intersecting cloud layers: Instead of discarding one of them could you try to just move the upper one upwards and sowhow 'staple' the layers so they don't get lost? Discarding cloud layers seems to create very similar weather all the time and reduces the vertical diversity which always was one of the reasons I loved SMP.

 

[sundog]

3 hours ago, FlyAgi said:

In case of intersecting cloud layers: Instead of discarding one of them could you try to just move the upper one upwards and sowhow 'staple' the layers so they don't get lost? Discarding cloud layers seems to create very similar weather all the time and reduces the vertical diversity which always was one of the reasons I loved SMP.

 

Yes, that's not a bad idea.

[FlyAgi]

I finally got some very impressive weather situations and it seems there were not so many cloud layers discarded (or at least it did not feel like that) and in those more complex situations SMP 5 really shines, the new thunderstorms are looking amazing and the integration of ray casted clouds and billboard clouds is about perfect. I think getting the cloud layers right in position would greatly enhance the current experience as FSGRW usually injects lots of layers and if they don't get lost the average weather situation should look more interesting and impressive.

 

[sundog]

11 hours ago, FlyAgi said:

I finally got some very impressive weather situations and it seems there were not so many cloud layers discarded (or at least it did not feel like that) and in those more complex situations SMP 5 really shines, the new thunderstorms are looking amazing and the integration of ray casted clouds and billboard clouds is about perfect. I think getting the cloud layers right in position would greatly enhance the current experience as FSGRW usually injects lots of layers and if they don't get lost the average weather situation should look more interesting and impressive.

 

Excellent! I've made some improvements here that should remove the need to delete or move cloud layers at all, so watch for that in a future update.

[FlyAgi]

Of course... waiting patiently. :-)

[ddunwoody]

Glad to see this thread; thanks for bringing it up and the clear comments from @sundog.  I think this explains why my flight into EBLG last night was easier than expected this this METAR:

 

EBLG 240120Z 29010KT 2300 -SN -PL BKN003 M00/M01 Q0994 REFZRA R22L/590294

(I'm using ASXP and RWC)

I think it's a great idea not to remove the lowest cloud layer, and my suggestion for working around the difficulty of cloud/terrain intersection would be to sharply restrict visibility while in OVC (and possibly in BKN) down to the lowest cloud base.  In the case above limit visibility to (a guess) 100m until EBLG+300ft and then up the visibility rapidly to 2300m.  This would give the impression of breaking out/close to at minimums, which is what I had been hoping for above.

To finish on a positive note - it's a great product and the clouds do feel more lifelike in motion vs 4.  The "problem" I think you have is that v4 is so good at everything that everyone is going to be pushing for improvements to 5 until it gets to the same level!

David

[Tompolth]

I've been doing some additionals tests based what has been said in this thread, and definetily the issues is with volumetric clouds. The lowest layers dissapear and the feeling is as a sunny day. I hope finally you will be able to manage some solution to use the volumetric clouds with a realistic aspect according to the metar. In any case, the non volumetric clouds are amazing in visual and performance way

Edit: Oops, I've just realized there was a new version correcting this issues. Let´s test it. Thanks guys. Very good job.

Edited January 24, 2021 by Tompolth