X-Plane 12 CB clouds

[mSparks 2,086]

8 minutes ago, ryanbatc said:

XP clouds are really blurry

hmm, I have absolutely no complaints regarding level of detail with the clouds set to max quality, tbh they came in well beyond my expectations, - some weird overall shape and distributions, but more often than not I look out the window and see clouds that look like XP12...

They only get blurry when the quality is turned down, but in 2D I can (aside from that EHAM issue) run everything at max quality with no problems at all, never going below 60fps, generally in the 100s, Phoronix reporting the same:

 

[Daube 353]

1 hour ago, mSparks said:

much closer than full mesh shaded snow. really just needs to get the cirrus fixed which is on the near term todo list.

I'm having issues with VR, main reason I've not flown much the last month or two, not sure if its XP, valve or nvidia to blame, but steamvr keeps crashing 15 minutes or so into a flight (which is insanely frustrating, because those 15 minutes are absolutely glorious) - seems possibly related to clouds/rw weather....

Just a shot in the dark here, but for SteamVR, don't you have a choice of versions in the "betas" section of the SteamVR app properties in Steam ? Maybe one of the other available versions would produce different results ?

 

[coastaldriver 919]

I still find that Laminar is way way better than LM and P3D in the cloud display and look.  As MSparks and I would agree the worst area is of course cirrus which is just, well, not cirrus looking at all, it looks like Cumulus Then again the weather engine is right out of whack at flight levels, temps are incorrect. Just do the math using the standard adiabatic lapse rate of 2 degrees C per thousand feet of temperature decline - based on ISA standard ground temp of +15C at ground level - at F370 you should see -59C - I have yet to see realistic OAT's anywhere at high altitudes in the sim using real weather!

Goran - XP is producing the same effect of black cloud bottoms as well (Had them on a flight over London just a few hours back) which means the way light is being treated, managed or depicted is wrong with respect to clouds and it gets worse or like that a low ambient light levels (on dusk or dawn in the sim) all real clouds are reflective of light and because they are just water vapour and they also let light through as evident by the real thunderstorm picture I posted where of course lightning isdischarging from within the CB, which it always is as it is generated by the mix up of positive and negative ions within a turbulent cloud such as a CB. To get the real lighting effect at night would mean having an effect or script that would generated a light source within a depicted cloud masss (Maybe if they did that the FPS would probably be 1) 

The other general issue is rain at FL's , at minus temperatures water freezes (discarding the more complex issue of supercooled water droplets that make clear ice form) so Cirrus cloud is always visible water vapour in the real sky but it is all frozen in other words they are flat sheets of ice particles. The spread out of the top of any thunderstom into a flat sheet of cloud is caused by the top of a CB hitting the tropopause in real life so no more vertical development and the high wind speed spreads it out in that characteristic anvil look. My comment here streaming rain I see in the sim at FL's (I think what the heck?) so the weather computation is not getting the temps right, not getting the cloud depiction right and hence the weather is not right at all. 

This is all really hard by the way because any simulation is trying to depict a 3D world and I am sure the modelling algorithims they use cannot handle changes in the lapse rate from + to - to positive to negative again, which is what the real world data would be giving it! 

Have a look at this it is called an aerological diagram it is a data tracing of temps and winds - translating this into the sim would be a nightmare, so be grateful they have managed what they have. 

 

Now have a look at this - this is a standard grid wind-temp forecast that RW pilots get (now if you know your met well you can work out where the CAT temp is likely to be and you can see the changes in the levels of the tropopause from southern to northern latitudes (critical stuff for turbine pilots) And if you know what your looking for (hint wind speeds in a consistent block of greater than 60 knots) you can also see where the jetstreams are.

Here I have blown up a section for you to see, three columns and 5 pressure levels per block, first value is wind direction, second value is wind speed, third value is temperature, nothing very linear here you can see. 

This is the data that the weather engine would be trying to do something with is also the point of this if your wondering!

 

Edited February 2, 2023 by coastaldriver
additional information

[coastaldriver 919]

FWIW here I have highlighted two areas based on the data that indicate the existence of a jet stream and hence possible CAT. Interesting! Now lets see them model severe turbulence at these flight levels!

So anybody flying from Adelaide, Brisbane, Sydney or Canberra to Melbourne today or the other way is going to hit some speed bumps in those areas!

Edited February 2, 2023 by coastaldriver

[mSparks 2,086]

50 minutes ago, Daube said:

Just a shot in the dark here, but for SteamVR, don't you have a choice of versions in the "betas" section of the SteamVR app properties in Steam ? Maybe one of the other available versions would produce different results ?

 

Valve blames nvidia, nvidia is several driver versions in now without a fix, but if I download real weather to get the weather then set it back to manual I "generally" can complete a flight, which makes me suspect XP12 is at least partly to blame. just did a complete half hour R44 VFR flight 100% by the numbers with not a single stutter that way. Although the desktop hard locked trying to exit VR..

Previously I couldn't complete a flight even in 2D with download weather enabled without a hard desktop lock. As of recent XP releases (e.g. the latest fixing the mem leak bug) I can complete a flight in 2D without much problem... just that flying in 2D is sad face and Im up to VR flight testing the 744 as next in the to do list...

All that stops me doing impulsive flights, because I would generally be otherwise working on the machine and having to restart is just to much of a pain.

Im not quite at the point of caps lock screaming at everyone to just please fix their #$%@#, more bouncing in my chair with excitement for when they finally do.

Plus pretty much waiting on openxr support from laminar, which I'm really hoping sorts it all out, especially since monado is now so much more advanced than openhmd.

Edited February 2, 2023 by mSparks

[MrBitstFlyer 3,873]

2 hours ago, ryanbatc said:

XP clouds are really blurry and there's almost no detail.  As I understand adding detail hurts fps but some detail needs to be added!

I would like to see some progress with the clouds too - the CB photos look very nice with extra detail.  I still find it odd that I prefer XP12 clouds to MSFS.  Clouds in MSFS are impressive, no doubt at all, and based on a screenshot I would say they look better.  However, the way XP12 portrays the atmoshere makes the clouds more part of the scene, rather than having a 'CGI' look like they often do in MSFS.

I liken MSFS clouds to what we had with previous generation clouds in products like Active Sky - details off the charts, but obviously not part of the atmosphere itself, but pasted on afterwards.  I'm not suggesting ALL clouds are like this in MSFS,  but I wouldn't swap the way XP12 clouds look 100% part of the atmosphere, just to gain some extra detail. Of course, If LR pull off the current look with extra detail, I'm in 🙂

Edited February 3, 2023 by MrBitstFlyer

[Bjoern 1,157]

7 hours ago, GoranM said:

If I saw that first cloud in ANY sim, I'd be asking what in high heaven is causing that internal 4 million nits bright light.

Lightning and the curve editing utility of any image editing software.

 

Quote

The second image...I find has very unusual shading going on.  I don't think I've ever seen any kind of "black" in any cloud lining.

You've never observed water saturated clouds at dusk?

[GoranM 3,764]

1 hour ago, coastaldriver said:

XP is producing the same effect of black cloud bottoms as well

Please post a screenshot.  TBH, I haven't seen it.  I've seen dark edges.  But not (absence of light) pitch black.

6 minutes ago, Bjoern said:

Lightning and the curve editing utility of any image editing software.

Well, sure.  That would be a modified photo, then.  I'm very well versed in curve editing, and post processing in Photoshop/Affinity Photo.  I have several processed photos on Airliners.net.

7 minutes ago, Bjoern said:

You've never observed water saturated clouds at dusk?

I have indeed.  I saw them as recently as less than a week ago, when we had a huge thunderstorm in my area.  Very dark, yes.  Black?  No.

I have never seen "black", as in RGB 0, 0, 0 clouds or cloud edges.  

Besides that, the screenshot he posted doesn't show a thunderstorm.  

It's just general cloud cover at sunset.

 

[coastaldriver 919]

GoranM here are some XP12 screenshots of flight not so long ago in Wales (I think I posted it elsewhere as well) this was mid morning, The realism was amazing actually, again low level, but note how dark the clouds have become.

Looks like the real world I used to fly in, 

But still a lot to be done on TS and CBs and high level Cirrus clouds.

[GoranM 3,764]

At the risk of coming across as repetitive, yes, they're dark, as I expect.  But not "black".

It's fine.  We don't have to keep going on this particular detail.  Everyone will have an opinion.

 

[mSparks 2,086]

  

1 hour ago, MrBitstFlyer said:

I still find it odd that I prefer XP12 clouds to MSFS.

I think its fundamentally, that XP12 clouds look like clouds except for when they dont.

I've not seen a single MS video or screenshot that has a "believable" clouds system - their CB with sprawl is well done visually, but aiui not physically.

42 minutes ago, coastaldriver said:

but note how dark the clouds have become.

What do you think it should look like?

They look fairly close to 3.50 in this to me

 

Edited February 3, 2023 by mSparks

[MrBitstFlyer 3,873]

2 hours ago, coastaldriver said:

GoranM here are some XP12 screenshots of flight not so long ago in Wales (I think I posted it elsewhere as well) this was mid morning, The realism was amazing actually, again low level, but note how dark the clouds have become.

In that weather I would think the clouds in you screen shots are maybe too light.  Many, many times cloud bottoms can be VERY dark - who hasn't remarked to someone IRL how DARK the clouds are?

[coastaldriver 919]

Well given how light effects how a cloud looks in real life (reflection, refraction and diffusion) and the darker the cloud the more dense (in terms of water vapour and droplets) hence nimbostratus and cumulonimbus (nimbus meaning heaped or fat) are particularly prone to that heavy dark or black look the light available (early, midday or late evening) can also change the colours reflected or seen, I am sure we have all seen CB's with grey, black, blue, purple and green hues in real life.

So I think they are managing the light issue in the sim reasonably well given the variables that the sky presents!

I can recall a trip up the coast of North Queensland many moons ago at night and we were well out to sea and there was literally a wall of thunderstorms along the coast line, the colours were spectacular - ir resembled a chinese lantern show. 

But it all depends on density so finally here is a real world picture of an absolutely massive thunderstorm of the coast of Australia taken by me at 500ft agl, the left portion which unfortunately is obscured was pitch black and it was raining like hell right over the airstrip we wanted to refuel at, this thunderstorm was in excess of 30nm wide. 

Edited February 3, 2023 by coastaldriver

[jcomm 5,918]

8 hours ago, MrBitstFlyer said:

I would like to see some progress with the clouds too - the CB photos look very nice with extra detail.  I still find it odd that I prefer XP12 clouds to MSFS.  Clouds in MSFS are impressive, no doubt at all, and based on a screenshot I would say they look better.  However, the way XP12 portrays the atmoshere makes the clouds more part of the scene, rather than having a 'CGI' look like they often do in MSFS.

I liken MSFS clouds to what we had with previous generation clouds in products like Active Sky - details off the charts, but obviously not part of the atmosphere itself, but pasted on afterwards.  I'm not suggesting ALL clouds are like this in MSFS,  but I wouldn't swap the way XP12 clouds look 100% part of the atmosphere, just to gain some extra detail. Of course, If LR pull off the current look with extra detail, I'm in 🙂

Less at night where there are no clouds at all.... At least in MFS we have clouds all day long, as IRL.

But I do agree they looked a bit "cartoonish" n previous versions although presently they got really better (in MFS).

Daylight, and excluding those weird extensions of  mid/high level clouds that look like walls, XP12 clouds do sometimes look really very good indeed.

In this area I hope LR keeps developing XP12 and fixing it's problems with weather depiction and effects. MFS is doing it too as was just announced yesterday on their latest twitch with lots of good news for upcoming updates regarding weather too, so, good reasons for both to now lower their arms...

Edited February 3, 2023 by cagarini

[turbomax 3,284]

On 2/3/2023 at 2:08 AM, GoranM said:

yes, they're dark, as I expect.  But not "black".

 

 

Edited February 5, 2023 by turbomax