Good that you mentioned this! This has to be reported to FSGRW developers. We had to fight tooth and nail to get this "fixed" in Opus.

 

Long story short, FSX will render thunder storm, when there is CB mentioned in the METAR. Especially in the Northern Europe, airports will append CB in the METAR if there is convective weather and/or severe icing (but not lightning). In the US, they report CB when lightning is present.

 

This is the reason why couple of the weather engines have the "suppress thunderstorm when CB reported" option.

 

Without this option, you will get thunderstorms all winter long in Finland

Actually I already reported this very issue and also a common solution is to add an option where the user can choose to have thunder and lightning or not with CB in the METAR and I got a positive reply from the developer so let's hope we'll see this implemented in the near future!

 

Skickat från min GT-I9300 med Tapatalk 4

 

 

 

 

I have no idea where 'brutal' comes from in this regard...!

 

Just read thru the instructions from the developers and its not like just click the installer....I must be getting old!

How accurate can your weather possibly be in changing conditions if half the stations are reporting what was there a half hour or more previously?

 

Exactly, that's why I'm not interested in 100% accuracy ... a smooth weather experience is by far more important to me.  Haze ON, 2 minutes later Haze OFF.  80% clouds 1 minute, then clear, then 30% clouds the next minute, then clear again ... to me, what defines a good weather engine for the confines of FSX is how well it can take all the data and still present a realistic and believe weather experience even if not 100% accurate.  But for me, the last thing I want to see is a low visibility approach suddenly go clear for a few minutes then fog, then clear, then low visibility ... all because a weather station or two isn't reporting accurately or isn't reporting at all.  A complex weather engine should be able to determine drastic variances in reported weather data from "set to set" (each weather set is basically created at a specified end user frequency).

 

But like Scott says, he's into "more accurate" and I'm into "more smooth".  But that's why I think it should be open to user control rather than some simplified sliders that go "more smooth" or "less smooth".  A complex weather engine will most like NEED to run on a separate computer (WideFS, SimConnect, etc.) because it will consume precious CPU cycles to do multi-layered compares of existing "used" weather sets vs. what's currently being reported.  It can be local to PC also, if one has CPU cores to spare.

Exactly, that's why I'm not interested in 100% accuracy ... a smooth weather experience is by far more important to me.  Haze ON, 2 minutes later Haze OFF.  80% clouds 1 minute, then clear, then 30% clouds the next minute, then clear again ... to me, what defines a good weather engine for the confines of FSX is how well it can take all the data and still present a realistic and believe weather experience even if not 100% accurate.  But for me, the last thing I want to see is a low visibility approach suddenly go clear for a few minutes then fog, then clear, then low visibility ... all because a weather station or two isn't reporting accurately or isn't reporting at all.  A complex weather engine should be able to determine drastic variances in reported weather data from "set to set" (each weather set is basically created at a specified end user frequency).

 

But like Scott says, he's into "more accurate" and I'm into "more smooth".  But that's why I think it should be open to user control rather than some simplified sliders that go "more smooth" or "less smooth".  A complex weather engine will most like NEED to run on a separate computer (WideFS, SimConnect, etc.) because it will consume precious CPU cycles to do multi-layered compares of existing "used" weather sets vs. what's currently being reported.  It can be local to PC also, if one has CPU cores to spare.

 

That sounds good to me.

 

Which one do you like?

 

 

Which one do you like?

 

I think he is thinking out loud about the perfect weather addon and not about one of the currently available...

We're getting some good replies here.

 

What got me to looking at other weather engines, and ultimately flying my tush off on the same route/same weather trying to find a good combination in AS2012 is a short fight near my home. The clouds were the same at the destination as at departure, but upon checking the weather map they should have gone away. This probably had something to do with Suppress Cloud Redraws. I know I've had differing weather in the distance with DWC many times, so I don't know what went wrong here. Doesn't really matter as I found settings that worked.

 

I don't like shifting visibility either. Does anyone have enough experience with FSGRW to tell how these situations are handled? Another thing I don't like is the barometer changing fast enough to see it. I've got that covered with some extreme smoothing in FSUIPC.

 

Something I like about AS2012 that I suspect is impossible in Opus is finding isolated ground fog 30 miles from the nearest METAR station. Also, seeing varying cloud formations out over the ocean when I'm flying along a coast and there are no METAR stations out there. FSGRW should handle both of these quite well.

 

One thing I cannot deal with is cloud popping. This doesn't mean the occasional redraw with the same clouds in the same place, it means the entire sky changes instantly. You can either have totally absolutely accurate METARS to the 6th decimal place with cloud popping, or you can have smooth change depiction and be sightly off from the METAR for a few minutes. In AS2012 you can use Suppress Cloud Redraws to eliminate this at the expense of getting some cloud formations that don't make much sense.

 

I live just north of Bonham Texas (F00) in case you want to look this up. I'm between several weather stations, but oddly enough AS2012 usually displays reasonably correct clouds for my location, pretty much what I see out the window, even to having the correct clouds in the distance. Of course, this took considerable experimentation and tweaking of AS2012 option settings.

 

Hook

 

Edited to fix stupid, extremely annoying formatting error.

 

 

Which one do you like?

 

None.

 

Not sure I'd say perfect weather add-on, just one that has more sophisticated smoothing and interpolation.  For example, I live in the SF/Bay Area and I've never seen the weather correct or consistent for this location (in any product, eval or purchased).  Typically there is a marine layer over SF/Bay Area that draws back to Pacific Ocean during mid day hours then rolls back into the area late afternoon/evening.  What all the weather engines appear to do is clouds for the entire SF/Bay Area or clear for the entire SF/Bay Area - switch back and forth every few minutes.  It's so hopelessly unrealistic for this area that it's better to just turn OFF dynamic weather.

 

I'd be happy to pay $100-$150 for a weather engine that worked to my expectations ... but it's gotta work not sorta works

 

I'm going to stay with REX for now because it does more than just weather and no other product has "solved" the dynamic weather problems.

 

Here is what I think a good weather engine will do:

 

1.  Gathers multi-source weather data constantly and interpolates (not tied to updating FSX)

2.  Update frequency to FSX is user defined

3.  Smoothing process is also user defined with many parameters

4.  Able to run on a networked computer

5.  Has a weather pattern system (micro climates) specific to locations which is used to define how the data is smoothed

 

Heck, regional weather engines could even be "sold" as add-ons to the base product ... for example West Coast US weather engine, and Central US weather engine, and East Coast US weather engine, Island Weather engine, etc. etc. (you get the picture).  But the key, these need to be more than just revenue makers, they need to actually work and work well with those regions.

 

Weather is such an important part of flight and realism, I think it's very "sellable" if done right.

 

 

I live in the SF/Bay Area

 

Oh Man, that explains everything.

 

I think you would need a dedicated add-on just for that.  :lol:

None.

 

Not sure I'd say perfect weather add-on, just one that has more sophisticated smoothing and interpolation.  For example, I live in the SF/Bay Area and I've never seen the weather correct or consistent for this location (in any product, eval or purchased).  Typically there is a marine layer over SF/Bay Area that draws back to Pacific Ocean during mid day hours then rolls back into the area late afternoon/evening.  What all the weather engines appear to do is clouds for the entire SF/Bay Area or clear for the entire SF/Bay Area - switch back and forth every few minutes.  It's so hopelessly unrealistic for this area that it's better to just turn OFF dynamic weather.

 

I'd be happy to pay $100-$150 for a weather engine that worked to my expectations ... but it's gotta work not sorta works

 

I'm going to stay with REX for now because it does more than just weather and no other product has "solved" the dynamic weather problems.

 

Here is what I think a good weather engine will do:

 

1.  Gathers multi-source weather data constantly and interpolates (not tied to updating FSX)

2.  Update frequency to FSX is user defined

3.  Smoothing process is also user defined with many parameters

4.  Able to run on a networked computer

5.  Has a weather pattern system (micro climates) specific to locations which is used to define how the data is smoothed

 

Heck, regional weather engines could even be "sold" as add-ons to the base product ... for example West Coast US weather engine, and Central US weather engine, and East Coast US weather engine, Island Weather engine, etc. etc. (you get the picture).  But the key, these need to be more than just revenue makers, they need to actually work and work well with those regions.

 

Weather is such an important part of flight and realism, I think it's very "sellable" if done right.

 

That's simply impossible with FSX. Actually, the real world weather source data that our (FSGRW's) weather servers have access to will have the information required to simulate the effect you mentioned - but FSX is simply not capable of doing it, because weather stations in FSX can either have clouds or they don't - there's nothing in between. The only way to get around this problem is by having more weather stations (a more detailed grid of weather stations). However, FSX limits the number of weather stations, so that's no a solution either.

 

If Version 2 of Prepar3d has a better weather engine, I'm sure we'll be able to get the real data to be better simulated!

 

Bernd

Most of these effects are described in approach charts of the specific airports

 

Actually all of them are from Jeppesen Charts, because we want users of FSGRW to be able to consider the weather effects when planning their approach (and for this to happen they simply have to know that they exist).

 

Btw, there's a lot of local weather effects in development right now (our internal list is at ~ 100 airports) and we will start to release them once Build 12 (the next update) is out.

 

Bernd

But I do question the coding "depth" of the various dynamic weather engine because if they were indeed complex weather engines, it seems they would have also provided much more end user customization options.

 

FSGRW doesn't create the weather on the client, our weather servers do that (because the process of doing that is very complex, takes ~1 GB of source weather data per weather file and requires a lot of resources). That's why the options on the client are limited.

 

Bernd

Good that you mentioned this! This has to be reported to FSGRW developers. We had to fight tooth and nail to get this "fixed" in Opus.

 

Long story short, FSX will render thunder storm, when there is CB mentioned in the METAR. Especially in the Northern Europe, airports will append CB in the METAR if there is convective weather and/or severe icing (but not lightning). In the US, they report CB when lightning is present.

 

This is the reason why couple of the weather engines have the "suppress thunderstorm when CB reported" option.

 

Without this option, you will get thunderstorms all winter long in Finland

Noticed how Bernd already commented on some stuff in this thread but here's some other great news regarding the thunder and lightning issue in Scandinavia for example.

 

http://fsgrw.freeforums.org/post681.html#p681

Hi Bernd,

 

I am greatly enjoying FSGRW, keep up the great work

Hi Bernd,

 

I am greatly enjoying FSGRW, keep up the great work

+1

 

IMO weather simulation in FSX doesn't get better than this for the time being :-)

Exactly, that's why I'm not interested in 100% accuracy ... a smooth weather experience is by far more important to me.  Haze ON, 2 minutes later Haze OFF.  80% clouds 1 minute, then clear, then 30% clouds the next minute, then clear again ... to me, what defines a good weather engine for the confines of FSX is how well it can take all the data and still present a realistic and believe weather experience even if not 100% accurate. 

 

 

I know that you were making a general statement about weather engines, but the weather conditions in OPUS are so spatially accurate that occasionally it seems unrealistic. For example, you can be flying along and there's constant rain, but there are patches of blue sky visible almost right above you. This is probably is more an artifact of the weather database than anything else. A station can be reporting rain and a sky condition that is not overcast. If you consider that the observer is describing the entire panorama at a site, the data do not translate down to a narrow viewpoint like what you sometimes can see out of the aircraft's windows. For some reason, I have never seen this condition with FSGRW.

Noticed how Bernd already commented on some stuff in this thread but here's some other great news regarding the thunder and lightning issue in Scandinavia for example.

 

http://fsgrw.freeforums.org/post681.html#p681

Great to hear a response so quickly! and nice to see I'm not only one with this problem, i thought twice before posting thinking it would just be me and i'd come off sounding negative

No need to think you're the only one, this is an old issue not only with FSGRW and some other wx engines solved it as mentioned by letting the user decide if a TS descriptor needs to be present but what Bernd mentions in the thread I linked about FSGRW sounds even better :-D