Board index FlightGear Development Effects and shaders

Sunrises (version 1.3 available)

An exciting "new" option in FlightGear, that includes reflections, lightmaps, the particle system etc.. A lot is yet to be discovered/implemented!

Re: Sunrises (version 1.1 available)

Postby Thorsten » Wed Feb 22, 2012 1:01 pm

Cannot you add a line of code to one of the Nasal's to switch the skydome on (whenever it is disabled)?


Well, I suppose I might have to do that, because Local Weather also checks if the skydome shader is on and takes control of its parameters.

Which is why it annoys me a bit that the information that the control property changed didn't make it to me - I am the only person who has worked with the scattering shader in recent months, and I have used the property in a number of cases. Somehow, there must be a better way of breaking things and hoping that the maintainers of dependent code realize it in time. This actually broke the LW interface to the skydome shader in 2.6 without me realizing.
Thorsten
 
Posts: 11106
Joined: Mon Nov 02, 2009 8:33 am

Re: Sunrises (version 1.1 available)

Postby Hooray » Sat Feb 25, 2012 11:56 am

Thorsten wrote in Wed Feb 22, 2012 1:01 pm:Which is why it annoys me a bit that the information that the control property changed didn't make it to me - I am the only person who has worked with the scattering shader in recent months, and I have used the property in a number of cases. Somehow, there must be a better way of breaking things and hoping that the maintainers of dependent code realize it in time. This actually broke the LW interface to the skydome shader in 2.6 without me realizing.


We could start maintaining a list of properties (or fgcommands) whose names and or default behavior somehow changed. This could be done using the wiki.
On the other hand, for core developers, it would probably be easier to maintain this info in $FG_ROOT/Docs
Please don't send support requests by PM, instead post your questions on the forum so that all users can contribute and benefit
Thanks & all the best,
Hooray
Help write next month's newsletter !
pui2canvas | MapStructure | Canvas Development | Programming resources
Hooray
 
Posts: 11354
Joined: Tue Mar 25, 2008 8:40 am

Re: Sunrises (version 1.1 available)

Postby Gijs » Sat Feb 25, 2012 12:30 pm

@Torsten, it was discussed on the mailing list (you even replied), but I must admit that I didn't explicitly mentioned that stuff linking to the old props should be updated. Sorry for that.
@Hooray, like http://wiki.flightgear.org/Aircraft_maintenance ;)
Airports: EHAM, EHLE, KSFO
Aircraft: 747-400
User avatar
Gijs
Moderator
 
Posts: 9380
Joined: Tue Jul 03, 2007 2:55 pm
Location: Amsterdam/Delft, the Netherlands
Callsign: PH-GYS
Version: Git
OS: Windows 10

Re: Sunrises (version 1.1 available)

Postby Hooray » Sat Feb 25, 2012 12:53 pm

Maybe more generic? I mean, Thorsten probably wouldn't be looking through aircraft-related regressions when he's looking for stuff that might affect local weather as a subsystem?

This is actually something that was discussed previously, funiliy exactly 2 years ago, when Thorsten first brought up such an issue here: viewtopic.php?f=23&t=7064&p=66581&hilit=#p65812

We need to come up with a policy to prevent such issues, these are unnecessary regressions.

It would be better for properties or commands whose names/behavior are changed to be explicit about it, i.e. by printing a statement to the console whenever the "old" property/command is still used, reminding people that the feature will be phased out eventually.

EDIT: http://wiki.flightgear.org/Property_Tree_Reorganization
This looks more like it should be adopted and updated. We should add another section for fgcommands and Nasal extensions functions, too. Also, a table for the FlightGear versions affected.

Regarding http://wiki.flightgear.org/Aircraft_maintenance - is this actually linked to anywhere from?
How do I find that out? It would make sense to add links to it...
Please don't send support requests by PM, instead post your questions on the forum so that all users can contribute and benefit
Thanks & all the best,
Hooray
Help write next month's newsletter !
pui2canvas | MapStructure | Canvas Development | Programming resources
Hooray
 
Posts: 11354
Joined: Tue Mar 25, 2008 8:40 am

Re: Sunrises (version 1.1 available)

Postby Ernest1984 » Sun Feb 26, 2012 12:06 pm

One question.

I have FG 2.6. I've just downloaded v.1.0, when i set the preferences i checked Advanced Weather to read the METARS, then clicked OK. When i log at EPWA airport, i see only single clouds, when in reality there are thick overcast clouds here (i live ~ 5nm from the EPWA airport so the weather should be the same).
What's the problem?
EPWA-hub: ERJ-195 clsgn: ERNIE
-----------------------------------
Around The World with B1900D!
callsign: AROUNDw - currently suspended till the FG will have a proper look on different locations
User avatar
Ernest1984
 
Posts: 438
Joined: Wed Dec 01, 2010 3:47 pm
Location: Poland / Canada
Callsign: ERNIE
Version: 2.10
OS: Mac OS X 10.8.2

Re: Sunrises (version 1.1 available)

Postby Thorsten » Mon Feb 27, 2012 7:58 am

Wrong thread, lightfield shaders don't do anything with METAR but: What does the METAR string say?
Thorsten
 
Posts: 11106
Joined: Mon Nov 02, 2009 8:33 am

Re: Sunrises (version 1.1 available)

Postby abassign » Mon Feb 27, 2012 5:13 pm

I'm thinking if it is possible to interpolate the data METAR between airports so that there is too drastic a change in weather. I thought I was a kind of interpolation of METAR data between airports, wondering if you know where are the procedures that read data METAR.
abassign
 
Posts: 809
Joined: Mon Feb 27, 2012 5:09 pm
Location: Italy (living 5 Km from airport LIME)
Callsign: I-BASSY
Version: 2018.3
OS: Linux Mint 19. x

Re: Sunrises (version 1.1 available)

Postby Thorsten » Tue Feb 28, 2012 8:13 am

I'm thinking if it is possible to interpolate the data METAR between airports so that there is too drastic a change in weather. I thought I was a kind of interpolation of METAR data between airports, wondering if you know where are the procedures that read data METAR.


Also wrong thread, but anyway: That's what Advanced Weather has been doing all along (it uses a 1/r weighted interpolation between all stations which have made a report within 250 km).
Thorsten
 
Posts: 11106
Joined: Mon Nov 02, 2009 8:33 am

Re: Sunrises (version 1.1 available)

Postby planetacancun » Wed Feb 29, 2012 4:55 pm

Anyways all wrong threads, one problem I see (and I don't want to start a new one) is the transitions in all weather states, affecting the shader effects like clouds and water most noticeably, when the METAR system updates the weater in FG the thansition should be smooth and not some drastic weather change, it makes it unreal (clouds popping up/dissapearing and sea waves moving fast). Is either prolonging the updates, disabling it once logged in (like one update at startup) or making up a new transition solution from one state to the other.

I have been testing the sunrises code 1.1 with my latest GIT Thorsten, pretty cool, I always enable all shaders to see whatsup and yeah my computer fans spin like nuts cooling down all those shaders hehe. Im like waiting for you guys to unite Rembrandt + Sunrises into GIT for some serious messing around with new videos :D
Callsign:Fallen-
Videos: http://youtube.com/planetacancun2
planetacancun
 
Posts: 323
Joined: Thu Jul 30, 2009 4:52 pm
Callsign: Fallen-
IRC name: Fallen-
Version: GIT
OS: Ubuntu

Re: Sunrises (version 1.1 available)

Postby Thorsten » Thu Mar 01, 2012 7:51 am

Anyways all wrong threads, one problem I see (and I don't want to start a new one) is the transitions in all weather states, affecting the shader effects like clouds and water most noticeably, when the METAR system updates the weater in FG the thansition should be smooth and not some drastic weather change, it makes it unreal (clouds popping up/dissapearing and sea waves moving fast).


Are we talking Advanced Weather here? Because I haven't ever observed METAR updates there, there are about 100 lines of code written to the express purpose of hiding them from the user.

A METAR update doesn't do anything to advance weather until the next tile is built, then it creates a weather parameter vector for interpolation and builds a tile according to cloud cover info. There are two issues here:

* a you can see a tile being built, i.e. clouds 'popping up'
-> that's currently unavoidable for large cloud visibility distances, because in order to do cloud-terrain interaction you need to know the terrain, in order to know the terrain you need to have it loaded and the range out to which it is loaded is given by the visibility, so unless someone separates the distance at which terrain is loaded from visibility, that is how it has to be. But it hasn't got anything to do with METAR. If you select the visibility distance of clouds small enough, you can hide the appearance of clouds.

* wind (pressure, temperature...) change after a METAR update
-> new interpolation vectors are faded in with a weight of 0.01, increasing each loop iteration by 0.01. The timer for loop iteration is 0.2 sec, so effectively it takes 20 seconds for the interpolation point to be fully recognized. But even then, you're not seeing 100% of its effect, you're seeing the 1/d weighted average with all other interpolation points present. I've actually taken the pains when I designed the system to write out curves of all weather parameters during a 10 minute flight with the ufo probing a few weather stations and plot them later on to verify there are no discontinuities being created, all parameters change very smoothly. If you think otherwise, repeat the exercise and show me the plot... In other words, as far as I am concerned, there is no issue of sudden changes of any weather parameter.

The problem with the water shader is that it can't deal with smooth transitions too well. But in reality, there are really fast wind changes, like in gusts the wind can change by 10 kt within a few seconds easily, so one cannot assume that wind changes slowly. But that's unrelated to the weather system.
Thorsten
 
Posts: 11106
Joined: Mon Nov 02, 2009 8:33 am

Re: Sunrises (version 1.1 available)

Postby planetacancun » Thu Mar 01, 2012 1:31 pm

Hey Thorsten, man I love your work okay? :) AAnd sorry for my typing errors (not native language) alright.

Is just what happens when im just flying and I have enabled METAR for weather updates. When this updates occur, sometimes suddenly all clouds dissapear, or on a clear sky all clouds appear, etc. and I think this "transition" from some weather state to another (like clear from cloudy) could be like slower, or like some effects do in some games like GTA that the cars and pedestrians pop up behind your sight :) and what I mean by not starting a new thread is just that relates to some of your work here, and I just wanted to place my comment. I'll go read more on that "Advanced Weather" and see where I can say moar there.
Callsign:Fallen-
Videos: http://youtube.com/planetacancun2
planetacancun
 
Posts: 323
Joined: Thu Jul 30, 2009 4:52 pm
Callsign: Fallen-
IRC name: Fallen-
Version: GIT
OS: Ubuntu

Re: Sunrises (version 1.1 available)

Postby Thorsten » Thu Mar 01, 2012 1:41 pm

Hey Thorsten, man I love your work okay?


I'm not offended - I'm just trying to understand the problem, because I can't seem to reproduce it. But lets change thread.
Thorsten
 
Posts: 11106
Joined: Mon Nov 02, 2009 8:33 am

Re: Sunrises (version 1.1 available)

Postby Thorsten » Thu Mar 08, 2012 8:09 am

Thanks to help from FredB, I have now been able to write effect files which allow to change between the default and the lightfield rendering scheme at runtime. I'm at the moment updating to most recent GIT, then I'll test everything and make it available afterwards.

This allows for the first time to make direct comparison of the same scene seen in the two different rendering schemes.


Dawn takeoff from PHNL in default scheme:

Image

And rendered in lightfield scheme:

Image

(note the much harsher shadows cast by the hills and buildings against the bright sky - I'm rather proud of that effect :-)

Two minutes later, passing UPP, the sky rendered in default:

Image

and in the lightfield scheme

Image

Note that the lightfield shaders place the horizon atthe top of the haze layer rather than at the true horizon position (unless the view is very good). Also, the lightfield tend to result in darker scenes given the same sun position. At least part of the reason is that while the light might be very bright at aircraft position, it is much reduced below. Another part of the reason is that based on photographing terrestrial sunsets in Finland over the winter I've come to the conclusion that for such low sun as shown in the picture, the light should actually be red.

Lightfield may underestimate the actual amount of light at high altitudes though... here much less atmosphere is traversed, and that's not really taken into account.
Thorsten
 
Posts: 11106
Joined: Mon Nov 02, 2009 8:33 am

Re: Sunrises (version 1.1 available)

Postby Sealbhach » Thu Mar 08, 2012 9:22 am

I know which one I prefer. :wink:

.
Sealbhach
 
Posts: 942
Joined: Wed Jun 30, 2010 9:17 am

Re: Sunrises (version 1.2 available)

Postby Thorsten » Thu Mar 08, 2012 12:25 pm

I just updated to recent GIT and it seems to be running just fine - so new version linked in first post, this makes everything runtime switchable and doesn't destroy default rendering.
Thorsten
 
Posts: 11106
Joined: Mon Nov 02, 2009 8:33 am

PreviousNext

Return to Effects and shaders

Who is online

Users browsing this forum: No registered users and 1 guest