sudden sky changes

[doug wornell]

As Mr. Dowson writes in the FSUIPC User guide towards the end under global weather in 2004 being problematic I continuously have a sudden loss of clouds and precipiation when using real weather on approach to airports after climbing to altitude and being in the air for a while. Visibility and winds are preserved. This is nopt a normal weather variation but something happening either with my machinery or the program. Any suggestions or thoughts.

[Pete Dowson]

I continuously have a sudden loss of clouds and precipiation when using real weather on approach to airports after climbing to altitude and being in the air for a while. Visibility and winds are preserved. This is nopt a normal weather variation but something happening either with my machinery or the program. Any suggestions or thoughts.

Hmmm. Not sure what's going on there. If this is with FS's own dowloaded weather it sounds like you are moving into areas with either no specific weather (in which case the defaults take over) or just clearer weather. However, you say the winds and visibility stay okay -- this is really the exact opposite of most folks complaints about FS2004's weather, that the winds change unpredictably and suddenly, and the visibility also does so instantaneously (though you should be able to fix the latter with FSUIPC facilities).

Sorry, I've never seen the clouds changing incorrectly. Check the weather menu (ALT W W), see if you have weather dynamics set to change things (the slider low on the display). Set that to minimum, see if that helps.

Regards,

Pete

[afterburner]

Hello Peter!

In the FSUIPC user guide from December 2003 you have written this:

"FS2004 global weather control is problematic

Following lengthy investigations I have come to the conclusion that, with my current state of knowledge of FS2004's weather system, there is no way to make a workable weather system using GLOBAL weather only. In other words, any external weather program which uses only global weather, and does not also set local stations, is pretty well doomed from the start.

The reason is this: I cannot find any method of setting global weather that does not eventually get 'localised' and modified in FS2004. It seems to take anything from about 20 minutes to 50+ minutes here-maybe it is also dependent on distance flown or something-but eventually the weather becomes localised and thereafter no amount of Global fiddling will change it. The only way to 'de-localise' it and get global weather back again is to clear the weather and start again.

This is true with dynamic weather setting turned off, so it is not that. I've proven that FS's weather system is always 'live and changing' by doing this:

1. Remove FSUIPC so there is no possibility of it interfering.

2. Load FS2004 and go to the Weather menu (ALT W W).

3. Set dynamics off, select user defined weather, and 'customised'.

4. Select 'all weather stations' (to make your settings 'Global').

5. Set up some weather that you can remember-some clouds, rain maybe, winds, and so on.

6. Now take off or slew and start moving.

7. Check that the weather looks more or less right.

8. Switch on the A/P or something, go away for 30-50 minutes.

9. Come back and see what's changed.

In my case the winds seem to stay as set, as does the visibility, temperature and pressure. But the clouds disappear or change and the rain stops. Checking the weather at various places in the FS dialogues I find that some places still have that 'global' weather I set originally, and others don't. The chances are all the very distance ones are still the same -- FS transforms and localises the weather around you, not on the far side of the world.

All this is despite weather dynamics being set to "Rate of Change = None", which evidently doesn't quite mean what it says, after all.

The problem is that, once the aircraft is flying through a cluster of WX stations all equipped with their own local weather, no amount of changing of the GLOBal weather by an external program will accomplish anythingActually, that isn't quite true. It does seem to accomplish something -- it makes the local weather change too, but not in any way which seems predictable. The changing of all those distance non-localised stations seems merely to accelerate the local changes along their own indeterminate course."

Several days before I read this document, I´ve recognized that a cloud type which was set in "global weather" menu has been changed or cleared during flights. This seems to be exactly how you have described: I went away for 30-50 minutes and came then back to see what´s changed.

The result is that "overcast" clouds become less density and "few", "broken" and "scattered" clouds become sometimes clear.

You´re right that it seems not to be predictable when and where weather change (cloud change) will happen. Yes, winds, temperature, pressure and visibility keep their values. Only clouds are changing during flights.

I´ve found out that setting "Cloud draw distance" does have an influence to this unwanted clouds change rate (the dynamic weather change is set to "none"). Setting low cloud draw distance (30 miles) will clear setting clouds soon, a high distance value (80 miles) will let clouds changing later and rarer.

Setting a high cloud draw distance will reduce the discribed problem, but not solve it.

So my question is: Is there any way to keep the clouds set in the global weather menu at all places and all times?

I don´t use updated real weather because I don´t have a fast internet conection (only 56K modem). Moreover, I often create weather by myself.

I don´t think that the cloud problem can be solved without any external programs.

I´ve some ideas how it could be theoretically solved (but I´m not sure):

1. There might be an external program where I can set my own "global weather" and that´s connected with FS2004.

One could set "real weather (updated)" in FS2004, and the flight simulator will get periodically weather from this external weather program where I´ve set my own weather before which is for all weather stations worldwide.

2. There might be a tool that refreshes the global weather in FS2004 periodically.

I experienced that checking the weather menu in areas where clouds have become changed or clear, the nearest weather station is also changed or "clear". Switching back to global weather and clicking at "OK", will not change the circuumstances. But changing any value in "advanced weather" and then clicking OK will bring clouds back. (i.e., change the pressure around +- 0.01 mm Hg). So, it´s not necessary to clear all weather and then to re-define the previous weather.

So, the tool would do exactly the opposite the tool "weatherset2.exe" does: Not continually reading a weather from FS, but read it one time when the weather is set in the menu, and then to WRITE it periodically into the FS weather stations.

It´s like one would go to the weather menu and define a weather for several weather stations around you.

So, is there any solution to keep the "lovely clouds"?

And is there a freeware or a payware tool that let them keep?

Thank you for any suggestions!

Konstantin

[Pete Dowson]

In my case the winds seem to stay as set, as does the visibility, temperature and pressure. But the clouds disappear or change and the rain stops.

Is this with still air? If there are any winds at the cloud levels, I think they will move -- as they do in the real world. I don't think this is really the 'problem' I was talking about so much as simply the advanced atmosphere simulation (that made its debut in FS2004) taking effect.

Checking the weather at various places in the FS dialogues I find that some places still have that 'global' weather I set originally, and others don't.

Yes, this is more to do with the 'problem' -- if you deliberately clear all local weather settings and set global weather, you'd think the weather would be the same everywhere. Then, even if the wind did blow, the same universal weather would simple move but still be everywhere. Why it changes with the changes minimised, I don't know. Evidently changes minimised don't actually stop them.

The chances are all the very distance ones are still the same -- FS transforms and localises the weather around you, not on the far side of the world.

Yes, the weather engine only works on a subset of waether zones, based on the weather stations, in a matrix around you.

All this is despite weather dynamics being set to "Rate of Change = None", which evidently doesn't quite mean what it says, after all.

Exactly. I tried automatically setting this right at the start, so that older FS2002-compatible weather programs could be used okay still. After we determined that we couldn't actually stop weather changing I advised all the weather programmers to change over to setting local weather.

The problem is that, once the aircraft is flying through a cluster of WX stations all equipped with their own local weather, no amount of changing of the GLOBal weather by an external program will accomplish anythingActually, that isn't quite true. It does seem to accomplish something -- it makes the local weather change too, but not in any way which seems predictable. The changing of all those distance non-localised stations seems merely to accelerate the local changes along their own indeterminate course."

Yes, I couldn't have said it better myself.

So my question is: Is there any way to keep the clouds set in the global weather menu at all places and all times?

Not that I know of, no, for all the reasons just discussed.

I don´t think that the cloud problem can be solved without any external programs.

Even then it isn't easy and you may get unwanted cloud flicker effects if an external program tries to change (or keep) local weather. The best results occur when the weather program changes the weather ahead of you, not in direct view.

1. There might be an external program where I can set my own "global weather" and that´s connected with FS2004.

One could set "real weather (updated)" in FS2004, and the flight simulator will get periodically weather from this external weather program where I´ve set my own weather before which is for all weather stations worldwide.

I don't know how that could be accomplished sensibly, unless you are going to maintain a server which can provide the Jeppesen-encoded weather data replacing the Microsoft/Jeppesen site where it downloads the current "real" weather from. I suspect that getting it to go to your site (which of course may be on your own local Network anyway) is relatively trivial compared to cracking the encoding/compression Jeppesen have applied.

2. There might be a tool that refreshes the global weather in FS2004 periodically.

The original FS2002 weather programs (FSMeteo and ActiveSky amongst many) were using Global weather only, and they lost control in FS2004. The problem is, as I state in the chat at the top of the forum, that in FS2004 the global weather is only applied to local stations when they have no weather. Once the local stations start "changing" (developing) their weather, they are all equipped with "local weather", so any re-imposition of Global weather doesn't have any effect locally (except possibly adverse effects from the changes this makes further afield).

The only way to re-assert global weather authority is to clear the weather completely every time. Unfortunately it is really impossible to do with without any visible and possibly physical results (physical in the sense of aircraft buffeting or whatever).

So, it´s not necessary to clear all weather and then to re-define the previous weather.

Well, using the menus you are effectively doing that, or more usually replacing/re-asserting ther local weatherthe visual disruption is hidden simply because you are not actually running the simulator whilst in the menus. It is easy to do that. If you don't mind periodic disruption, automatic from a program instead of manual by use of menus, then, yes, it could be done by program. But it won't be nice to watch, even infrequently.

The program is better off changing the local weather stations rather than clearing all weather and reassting global settings, but even then you'd probably notice some flicker, and maybe FS would stutter a bit whilst it re-calculates the N nearest stations.

It´s like one would go to the weather menu and define a weather for several weather stations around you.

With visible stutters and flickers, yes -- these you don't see in the Menu because you have stopped the action.

So, is there any solution to keep the "lovely clouds"?

I seem to experience "lovely clouds" most of the time. Nearly all "real weather" is cloudy a lot of the time in the UK where I fly. I get "lovely" clouds by ensuring that the 3D cloud slider is at 100% and the quality/density slider below is also well over to the right. The other two sliders can be adjusted then to give back some frame rates, but if possible both of those are best full right.

Certainly the cloud effects in FS2004 seem far superior to all previous versions, and so I'm not really sure what you are trying to improve. I thought what you were after was keeping the exact clouds in the same position "fixed" in the sky, no matter what weather changes were afoot -- and I don't think this is possible (or realistic). But why are you so unhappy about the cloud graphics?

Regards,

Pete

[afterburner]

Hello Pete!

Although it´s more than 2 months ago since you´ve replied to my post, I want to thank you for your detailed answer and observations to my question about possibilities to keep clouds on their place.

In the meantime, my interest in FS2004 became less because I was crushed in somebody :oops: and I had other passions than for flight simulators.

But now, I can enjoy flying again and so I have chosen to thank you for your answer. Sorry that I didn´t it earlier. I hope, you´re not very dissapointed. I´ve thought, better later than never...

It seems that you have misunderstood the word "lovely clouds" I´ve used in my post...

I have set it in "", because I´ve read your FSUIPC readme and you have called the clouds "lovely clouds".

In my post, I´ve only quoted your expression...

That doesn´t mean that I´m unlucky about the cloud graphics in FS2004.

The opposite is the case - I really think that this cloud graphics is the best I´ve seen, compared to previous FS versions.

However, I´m not really satisfied with the new weather engine of FS2004.

One reason is the clouds dissapearing after flying some rounds (this was the topic of my post), the second reason is the fact that you can´t see blue sky and the sun at a visibility of 10miles, which is often a typically visibility in Europe.

In previous versions, this was solved better, but I can´t understand why one will not see sun and sky any more at a visibility of 10 miles in FS2004.

In reality, one can see skies at a visibility of even 5 miles!

But the sky colour is really great, especially at dawn/dusk!

Well - in my opinion, there are no more bugs (except of weather) that dissapoint me dramatically.

But it´s another topic

In any case - thank you for your answer and best wishes for your (eye-)health!

Konstantin

[Pete Dowson]

However, I´m not really satisfied with the new weather engine of FS2004.

One reason is the clouds dissapearing after flying some rounds (this was the topic of my post)

"Rounds"? What are they?

How are your clouds 'disappearing? Do you mean the dynamic weather? You can reduce that to almost zero by setting the slider all the way to the left. However, from what I've seen, sometimes clouds develop rather than 'disappear'. Possibly you are selecting one of the themes? I think they have dynamics built in.

If you are setting your own weather then possibly you are only setting your local station(s). If there is a wind at the cloud level, they will move with it. If there are no clouds following, the sky will clear. If this is the problem then you are better off downloading the real weather, or using one of the weather programs such as ActiveSky or FS_Meteo.

the second reason is the fact that you can´t see blue sky and the sun at a visibility of 10miles, which is often a typically visibility in Europe.

In previous versions, this was solved better

I'm not sure that problem was ever any different, though I don't think the limit is 10 miles? In FS2002 it was about 4 or 5 miles at which that 'cut-off' happened. Maybe the limit value has changed, but I've not particularly noticed it. Best if you use graduated visibility so you can climb out of it.

The real problem with the way visibility is implemented (and this probably has as much to do with the way video cards work as anything) is that the visibility limits apply in all directions -- i.e. up and down as well as horizontally, whereas really it should be the horizontal limit and it should taper off quite quickly with viewing angle.

I don't know if that's possible with video card 'fogging', but you should send your suggestions NOW (in time for the next version of FS) to Tell_FS@microsoft.com.

Regards,

Pete

[afterburner]

Hello Pete!

I have learned to live with the dissapearing clouds. 8)

They dissapear i.e. if one set a cloudy weather and flies around an area of weather stations (I´ve meant that with the expression "flying some rounds").

It´s exactly how you have described in your FSUIPC-guide:

"I've proven that FS's weather system is always 'live and changing' by doing this:

......

5. Set up some weather that you can remember-some clouds, rain maybe, winds, and so on.

6. Now take off or slew and start moving.

7. Check that the weather looks more or less right.

8. Switch on the A/P or something, go away for 30-50 minutes.

9. Come back and see what's changed.

In my case the winds seem to stay as set, as does the visibility, temperature and pressure. But the clouds disappear or change and the rain stops."

Well - in my case the same is happening with the clouds and rain.

But I´ve found out that when I´ve created a flight plan and I follow the planned route (or when I fly simply forward, no matter if a plan was created or not), the clouds stay on their right place.

But when - how you have described - one flies back to the departure weather station, there are generally no more clouds (or less), also when you have saved your flight on your route and select it after ending the flight. (In my experience winds don´t influence this).

So - I can live with this.

Of course it would be better if this wouldn´t be a problem - but one can´t get anything...

Or maybe I have to experiment with the dynamic weather, i.e. trying to set different dew point and temperature values?

Now to the visibility:

It´s right that in earlier versions they could notice this "cutoff" (the grey mist) at a visibility of below 5 or 3 miles, in FS2002, FS2000 and FS98.

That´s the reason why I can´t understand Microsoft why they have implemented this unrealistic all-directions-grey-visibility limit at 10miles, where there weren´t any problems in earlier versions with it?

If a suggestion to Microsoft will help?

They should be intelligent enough to know that it isn´t grey in a clear day at a visibility of 10 miles...

But what would me interest:

Is there theoretically, from a programmers point of view, any possibility to create haze/visibility limit as a special effect in FS in dependent of visibility value in weather menu? (The FS´own visibility could be always "unlimited", but only the created special effect could be visible?)

When the answer would be yes, maybe I will think about studying programming by myself... :idea:

Konstantin

[Pete Dowson]

I have learned to live with the dissapearing clouds. 8)

They dissapear i.e. if one set a cloudy weather and flies around an area of weather stations (I´ve meant that with the expression "flying some rounds").

Yes, because when you manually set weather it is well nigh impossible to set all the weather stations in the world. The closest you can get is to set so-called global weather, but I think that also only sets stations within a certain range. So they change. They become localised and change just as any other local weather does.

I think all this is part of the new rather too clever weather simulation we have in FS2004. It is very different from the unrealistic one in previous versions. And I do quite like it now I am over the shock. :wink:

It´s exactly how you have described in your FSUIPC-guide:

"I've proven that FS's weather system is always 'live and changing'

Yes, of course. I emhasised that there because from using previous versions folks were used to being able to set such-and-such a weather and have it stay that way. FS2004's weather is not really like that.

I still don't believe clouds only disappear. They also appear too. But your best bet, as I advised, is certainly to download real weather for the whole world, so that all stations are populated, or to use one of the weather programs. Some of them have evolved to deal with FS2004's weather very well -- I have heard great things about the next version of AS2004.

But I´ve found out that when I´ve created a flight plan and I follow the planned route (or when I fly simply forward, no matter if a plan was created or not), the clouds stay on their right place.

What is the "right place" for a cloud? Surely they are moving? Or are you ensuring there are no winds at any level?

It´s right that in earlier versions they could notice this "cutoff" (the grey mist) at a visibility of below 5 or 3 miles, in FS2002, FS2000 and FS98.

I'm pretty sure it was more than 3, maybe more than 5, but I'm not so sure about that.

That´s the reason why I can´t understand Microsoft why they have implemented this unrealistic all-directions-grey-visibility limit at 10miles, where there weren´t any problems in earlier versions with it?

To be honest I'd not even noticed that the 'limit' had changed at all. Maybe I have been lucky. Using external weather programs and with "Extend METAR max vis" enabled in FSUIPC, the "10SM" which in US METARs simply means "10 miles or more" is almost always extended beyond the 10 miles. In Europe, where the value is "9999" meaning "10 km or more" it is a little more likely that this will give me 6.2-10 miles sometimes, but as I say I've not noted it as a problem before. Maybe I should look out for it on the next chance I get to fly (not so often at present).

If a suggestion to Microsoft will help?

It cannot hurt. It will help them to prioritize things. However, as I pointed out, it isn't this 'limit' which is so much a problem as the imposition of the same visibility in all directions, up and down included. If that is an inherent function of video card design or DirectX then they, the FS team, may be rather limited. I'm not even sure that the apparent increase in the 'limit' from 5 or 6 to 10 miles was of their doing, it may have been a side effect of changes in DirectX.

Is there theoretically, from a programmers point of view, any possibility to create haze/visibility limit as a special effect in FS in dependent of visibility value in weather menu? (The FS´own visibility could be always "unlimited", but only the created special effect could be visible?)

This is a graphics and DirectX question. I am sorry to say that it is not my area of expertise in any way whatsoever. Sorry. Maybe you will find the answers in the reference works for DirectX 9?

Regards,

Pete