Weather settings in Condor2

Moderators: Uros, Tom, OXO

lordauriel
Posts: 58
Joined: Thu Jan 30, 2014 10:12 am

Re: Weather settings in Condor2

Post by lordauriel » Tue Mar 26, 2019 12:18 pm

can I ask, which are the sceneries known to have proper albedo maps?

janjansen
Posts: 1090
Joined: Wed Feb 24, 2016 9:26 pm

Re: Weather settings in Condor2

Post by janjansen » Tue Mar 26, 2019 12:30 pm

FWIW, I did a little digging. Condor uses DirectPlay for its network stack. I found this in MS documentation about the directplay ping timer:
The period for this timer is 35 seconds.
https://docs.microsoft.com/en-us/opensp ... 79187f9e43

This value doesnt seem to be configurable, so as I feared, this may not be an easy fix of just adjusting some time out value somewhere. The weather calculation code would either have to be sped up significantly to avoid the issue all together, or have to be threaded/interrupted so the client can respond periodically to pings, even if it takes >35s to calculate.

User avatar
EDB
Posts: 1575
Joined: Mon Oct 31, 2011 1:17 am
Location: The Netherlands, Europe Continent, Earth Planet, Milky Way Galaxy, Virgo Super Cluster

Re: Weather settings in Condor2

Post by EDB » Tue Mar 26, 2019 12:41 pm

Maybe something in the way it is calculated can be changed. Difficult to say right now... But as Jan Jansen says... It is a known problem and it is on the very long list...

But most sceneries need a normal (Slovenia2), high (or the non-existing very high) setting which work fine. So for the time being there is a work-around. It's not a BSOD issue...

It's only a problem if you want the extreme situation where there are no thermals in the valleys.

There is the option for a thermalmap update if a scenery doesn't work as expected... Maybe make two updates. One with a corrected (extreme) thermalmap... And one to restore it. And don't forget to have different ini with landscape in both versions...

Version 1.0 vs 1.0x for example...
Last edited by EDB on Tue Mar 26, 2019 1:03 pm, edited 1 time in total.

janjansen
Posts: 1090
Joined: Wed Feb 24, 2016 9:26 pm

Re: Weather settings in Condor2

Post by janjansen » Tue Mar 26, 2019 1:02 pm

EDB wrote:
Tue Mar 26, 2019 12:41 pm
Maybe something in the way it is calculated can be changed. Difficult to say right now...
Indeed, its somewhat strange it takes so much longer to calculate fewer thermals ;) so there may be a performance bug that could be simple to solve and as a side effect, solve the timeout problem, but it probably makes more sense to review this code as part of a larger reworking of the weather model (like better Cu streeting and probably a few dozen other suggestions on the list).

TL;DR, I wouldnt count on this being fixed soon.

janjansen
Posts: 1090
Joined: Wed Feb 24, 2016 9:26 pm

Re: Weather settings in Condor2

Post by janjansen » Tue Mar 26, 2019 1:41 pm

lordauriel wrote:
Tue Mar 26, 2019 12:18 pm
can I ask, which are the sceneries known to have proper albedo maps?
Not an exclusive list, but at least these reportedly use ground albedo thermal maps:

- Slovenia
- AA2
- Cascade Range
- Southern Norway
- Matamata
- Nephi
- CW A/B/..
- Netherlands HD
- Lake Keepit

There could be more though.

User avatar
EDB
Posts: 1575
Joined: Mon Oct 31, 2011 1:17 am
Location: The Netherlands, Europe Continent, Earth Planet, Milky Way Galaxy, Virgo Super Cluster

Re: Weather settings in Condor2

Post by EDB » Tue Mar 26, 2019 2:56 pm

janjansen wrote:
Tue Mar 26, 2019 1:41 pm
lordauriel wrote:
Tue Mar 26, 2019 12:18 pm
can I ask, which are the sceneries known to have proper albedo maps?
Not an exclusive list, but at least these reportedly use ground albedo thermal maps:

- Slovenia
- AA2
- Cascade Range
- Southern Norway
- Matamata
- Nephi
- CW A/B/..
- Netherlands HD
- Lake Keepit

There could be more though.

Did you check those thermal-maps..?

I personally wouldn't have placed a list here, because many landscapes are still a work in progress I guess and I think it is the responsibility of the landscape builder to inform his user. And in some cases the list can be a subject for a debate that will be endless. And the list isn't complete/current. But now this list is here I have to react on it, because you used some of my research I think. In a research done on 27 of the 48 or so current landscapes (most used in many current competitions I flew in) it appeared that many didn't had a thermal-map or only a shadow-map thermal-map. Both are not the way to do it in Condor 2. The result is confused Condor users... And the Condor Team is not happy about it..! The above list should be divided in (at least) two IMHO.

Real 1/Albedo (or at least tried to), In alphabetic order :
- AA2
- Matamata
- Nephi
- Netherlands HD

1/Albedo (or at least tried to), In alphabetic order :
- CW A/B/..
- Cascade Range
- Lake Keepit
- Slovenia2
- Southern Norway

In the first group there are the landscapes with thermal-maps that use real life data with many surface variations and relatively high resolutions. So they might come really close to flying there in real life. Although I know that the Netherlands HD landscape needs some tweaking with other data. So the other ones might need some tweaking too (as do the thermal-maps in the second group)

The second group is a special one in that some landscapes have important surfaces missing, or only some basic ones or use a trick (like Lake Keepit). They can come really close to real life, but when doing a real 1/Albedo thermal-map from real data the map can be totally different. For Hybrid landscapes this can be expected because they use repeating textures. But they at least tried and they at least tried to put what you see in the textures in the thermal-map. So yes, I consider them as 1/Albedo thermal-maps, but the first group is a different kind of beasts if surfaces are assigned a correct Albedo... The first group can lift Condor to a higher level of realism. And they might need some tweaking with altitude info in the future or because of changes in the Condor thermal Engine... But this last point is a point for the second group too... The Southern Norway landscape in the second group has some added tweaking by someone who flies in that area very often, so I consider it the top in that group. It maybe should be added to the first group because of it...

I made some (flatland) landscapes that used real data and that were tweaked with more added data for even more realism for Condor V1 back in 2014 that were really fun and challenging to fly and they would come really close to real life and real life thermal-maps. I could see all the specific differences in areas I knew and flew in in real life. We then also used real life airspace to make it even more realistic and challenging. The current Matamata and Nephi ones are less tweaked then those thermal-maps I made back in 2014 though. First see how the current versions work out with the current Condor 2 engine. Tweaks can come later if needed...

I hope (and I know the Condor Team also hopes) that in the near future all published landscapes will have 1/Albedo Thermal-maps. Tweaked or to be tweaked in the future... You have to start somewhere... I personally think a landscape builder should make a thermal-map first, and his textures second... In a Hybrid scenery you can use the thermal-map to make the textures... So I support the AA2 0.5 version with a proper Thermal-map, but with some tweaking to do in the eastern part visible textures and maybe dynamic water...

- EDIT number ## -
One extra warning regarding (realistic) Weather settings and Nephi.
Nephi was created as a test landscape during Beta testing for it's relatively high altitude and because some well known YouTube glider pilot maybe would do some beta testing. Nephi can have cloud-base as high as 6000m in real life. For this the cloud-base (up to 5000m) and inversion height (up to 8000m) can be set to really high values now. The current wave model however is designed for landscapes like Slovenia2 which is the base scenery of the sim. The wave tops are at around 5000m. So in Nephi (and an other even higher situated landscape I was building, but currently is on hold) you can create unrealistic weather were thermals are higher then wave lenti's. It is up to the task-setter to avoid this at the moment. Keep thermals below say 3000m/4000m if you also have wave. It's even better to avoid both thermals and wave in Nephi at the same time completely though. Use thermals or wave and not both, or do a good check that the situation doesn't get unrealistic. Or inform people that you like to try something that doesn't happen in real life...
This issue is on the very long list. Maybe we can set the altitude of the stable layer needed for wave manually in the future... (and that layer should be in the range close to the mountain tops)

Happy flying...
EDB

janjansen
Posts: 1090
Joined: Wed Feb 24, 2016 9:26 pm

Re: Weather settings in Condor2

Post by janjansen » Tue Mar 26, 2019 7:12 pm

EDB, you know where I got that list from; I deliberately didnt mention you as source because I didnt want to drag you in to any "controversy", I just wanted to provide a simple, if incomplete, answer to lordauriel's question without opening cans of worms.

arneh
Posts: 554
Joined: Mon Nov 16, 2009 10:34 am

Re: Weather settings in Condor2

Post by arneh » Tue Mar 26, 2019 10:53 pm

EDB wrote:
Tue Mar 26, 2019 2:56 pm
The second group is a special one in that some landscapes have important surfaces missing, or only some basic ones or use a trick (like Lake Keepit). They can come really close to real life, but when doing a real 1/Albedo thermal-map from real data the map can be totally different.
For Southern Norway the thermal map is based on real surface types, but it also matches the textures so that it should be fairly easy to see which areas should be good. For those who flew the last Southern Norway task in Condor World Cup, the task started from an airport which lies in an area with a lot of bogs, which are not good for thermals, so there were poor thermals in the start area. Not helped by the the task setter having a max start altitude only 500 m above the terrain, it made for a tricky start. But once you got into the mountains the thermals improved a lot, while the forest areas later in the task were more average.

I won't post the entire thermal map, but here is a small part of it, showing how different terrain types look in the thermal map:
ThermalMap.jpg
The forest has both some very poor areas, but also clearings which are particularly good, hence the high variability there. At least in Norway the forest clearings are some of the best thermal generators in the low lands. The clearings should be easily seen when flying.
Not shown in that part of the map are some poor terrain types like snow and bogs.

In addition there have been some manual tweaking to areas which are known in real life to be particularly good or bad.
You do not have the required permissions to view the files attached to this post.

User avatar
jcomm
Posts: 252
Joined: Mon May 16, 2005 12:51 am

Re: Weather settings in Condor2

Post by jcomm » Thu Mar 28, 2019 3:50 pm

So... from the listings above, not even the default "Slovenia2" scenery, proprietary, belongs to the rankong of the most customized ?
José Oliveira
Hey you Chicks flying bricks....
Come Fly Gliders ! Be Outsiders!

User avatar
EDB
Posts: 1575
Joined: Mon Oct 31, 2011 1:17 am
Location: The Netherlands, Europe Continent, Earth Planet, Milky Way Galaxy, Virgo Super Cluster

Re: Weather settings in Condor2

Post by EDB » Thu Mar 28, 2019 4:38 pm

In my eyes he doesn't have to be on that list. For beginners it is the landscape to start with... You don't drop someone in the Atlantic Ocean if he is starting to learn how to swim...

I see way more forrest in real data... Is this important for a partly fictive landscape? I guess not...

It's a hybrid landscape with repeating textures... So the thermalmap will always be different to the real map. And you can not put a real life thermalmap underneath a hybrid textured landscape with forest in different locations. At least the landscape is 100 legal...

The competitions at virtual soaring that almost only use Cadfaels hybrid landscapes are the best in the business. So it all is very rellative... Is that the right word..?
You also need a good tasksetter that knows how to use the landscape to his advantage...

Personally I prefer thermal maps that are made of real life data/info, but then with a little bit more contrast. A bit more extreme. But these might be a bit too extreme for the general public. One example was the C1 Hahnweide SW Germany thermal map... Real data can also be quite boring... Some real life soaring areas are boring...

In Nephi I only used albedo data. In it's current state that can be quite hard to fly in already (with no 'very high' Flatlands Activity available)... Landscapes with different elevated plateau's may need some tweeking in Condor. Matamata maybe needs some tweeking like I did with my other landscapes...

I don't mind differences in the above two groups. What I think that is important is that all other landscapes get 1/albedo maps too. With or without real life data or tweaks. All is better then no thermalmap at all... Or wrong shadowmaps... Shadow maps are done by the Condor engine.

Rotareneg
Posts: 241
Joined: Mon Dec 04, 2017 1:59 am

Re: Weather settings in Condor2

Post by Rotareneg » Thu Mar 28, 2019 8:15 pm

The thermal strength setting effects anabatic winds. On the task I was testing them with (AA2, Sollieres area, July 21 at noon) moderate strength was enough to just break even and occasionally climb with a JS1-21, but at very strong greater 4 kt lift was pretty common.
Image

lordauriel
Posts: 58
Joined: Thu Jan 30, 2014 10:12 am

Re: Weather settings in Condor2

Post by lordauriel » Fri Mar 29, 2019 9:54 pm

Many thanks for all the replies!!

PS:
Nephi is awesome.

witor
Posts: 200
Joined: Wed Dec 15, 2010 11:47 pm

Re: Weather settings in Condor2

Post by witor » Thu Jul 08, 2021 6:34 pm

I know this is a bit old topic but I stumbled on it googling for something else...

Reading all the above I have an impression that knowledge of how weather settings affect weather algorithm and knowledge of what the thermal map actually looks like gives an unfair advantage in competitions as it is not a common knowledge. As an example I could use a post about Norway thermal map and description of how it was evident from the thermal map that thermals will be weak in the start area, but much better in the mountains. I also disagree that how weather algorithm works and how it is affected by the settings should be some type of big secret. This only leads to uneven playing field. I think that details of how weather and settings in condor work should be published along with the thermal maps for all sceneries. At least everyone would have equal information and no unfair advantage. I am pretty sure knowing what thermal maps look like would affect my decisions in many, tasks I flew.

Post Reply