Re: [Flightgear-devel] *** SPAM *** Re: Scene ambient and specularcolor changes
Melchior FRANZ wrote: * Vivian Meazza -- Saturday 04 April 2009: I can fix it here, to my entire satisfaction, so if everyone else is happy, don't worry. I can't really say much about this, as I'm using a rather old and not so great monitor that I don't seem to be able to calibrate correctly. But I agree that the ambient has now become significantly darker. I just can't say if that's right or wrong. :-/ To be honest both screenshots don't look too bad to me :-/ But if it's going to be adjusted it needs to be done in Data/Lighting/ambient Erik -- ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] *** SPAM *** Re: Scene ambient and specularcolor changes
Erik Melchior FRANZ wrote: * Vivian Meazza -- Saturday 04 April 2009: I can fix it here, to my entire satisfaction, so if everyone else is happy, don't worry. I can't really say much about this, as I'm using a rather old and not so great monitor that I don't seem to be able to calibrate correctly. But I agree that the ambient has now become significantly darker. I just can't say if that's right or wrong. :-/ To be honest both screenshots don't look too bad to me :-/ But if it's going to be adjusted it needs to be done in Data/Lighting/ambient This is how I think it should look, and arguably it's still a little dark: ftp://ftp.abbeytheatre2.org.uk/fgfs/Light/Seahawk%20-%203.jpg ftp://ftp.abbeytheatre2.org.uk/fgfs/Light/Seahawk%20-%204.jpg I'm doing a small adjustment in light.cxx - seems to work: float ambient = _ambient_tbl-interpolate( deg ) + (0.25 + 0.75 * visibility_inv/10); Not sure that I fancy tinkering around in Data/Lighting/ambient - someone has obviously taken a lot of care to craft that one. Vivian -- ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] *** SPAM *** Re: Scene ambient and specularcolor changes
* Vivian Meazza -- Saturday 04 April 2009: This is how I think it should look, Does indeed look much better here (on my *still* quite bad monitor ;-). m. -- ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] *** SPAM *** Re: Scene ambient and specularcolor changes
Erik Hofman wrote: After reading the comments I agree with it. I'll take some time to adjust the ambient accordingly. This has been committed to CVS now. Let me know what you think. (keep in mind that the darkest ambient color is defined in data/Lighting/ambient which has not been touched for ages). Erik -- ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] *** SPAM *** Re: Scene ambient and specularcolor changes
Erik Erik Hofman wrote: After reading the comments I agree with it. I'll take some time to adjust the ambient accordingly. This has been committed to CVS now. Let me know what you think. (keep in mind that the darkest ambient color is defined in data/Lighting/ambient which has not been touched for ages). IMO, without any real data to judge it by, the specular adjustment is good enough. Ambient is still significantly too dark - on the side of an ac lit only by ambient light, it is just about black. In my experience, on the brightest days when the difference between ambient and diffuse is greatest, there is always enough ambient light to penetrate even the most intense shadows. I'll do a couple of screen shots later Vivian -- ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] *** SPAM *** Re: Scene ambient and specularcolor changes
On Fri, Apr 3, 2009 at 7:26 AM, Vivian Meazza wrote: Erik Hofman wrote: After reading the comments I agree with it. I'll take some time to adjust the ambient accordingly. This has been committed to CVS now. Let me know what you think. (keep in mind that the darkest ambient color is defined in data/Lighting/ambient which has not been touched for ages). IMO, without any real data to judge it by, the specular adjustment is good enough. Ambient is still significantly too dark - on the side of an ac lit only by ambient light, it is just about black. In my experience, on the brightest days when the difference between ambient and diffuse is greatest, there is always enough ambient light to penetrate even the most intense shadows. I'll do a couple of screen shots later One thing to possibly consider is that when we (someday) get back to having shadows cast by the aircraft, we may need to recalibrate some of these parameters. I recall long ago when I was trying to play with ambient/diffuse parameters for terrain, I quickly discovered a big part of the problem is the lack of dynamic range of a computer monitor. The diffuse lighting is based on the angle of the surface relative to the angle of the light source, and near dawn/dusk this gets really low. And because of the way opengl does the lighting math you can't really do anything about that (nor would you want to.) However, in order to get the scene brightness back up to something usable, you have to boost the ambient component of the lighting artificially high, and then all shadows go away. Contributing to the problem (I think) is that most of us view the scene in a normally lit room. If we forced everyone to only view FlightGear inside a perfectly dark room, I think we could do a little better. But this is a really hard problem. Very easy to get wrong, impossible to get right, very hard to get mostly right (or not distractingly wrong.) Regards, Curt. -- Curtis Olson: http://baron.flightgear.org/~curt/ -- ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] *** SPAM *** Re: Scene ambient and specularcolor changes
LeeE On Wednesday 01 April 2009, Erik Hofman wrote: Heiko Schulz wrote: I'm against this code- I can't really understand what's the intention of this code is- destroying glossiness? Only when it's foggy. On bright clear days it will shine like before. The real intent is to make it a bit more realistic. I would like to see a change of the colors depending of the clouds setting- I think I can remember that the color changed with the weather settings in plib-versions. Example: less glossiness and more greyish colors on thunderstorm/ rainy day While not this may not be exactly true all the time but using METAR will update both visibility and cloud settings accordingly (heavy clouds will only be possible in situations with reduced visibility anyhow). Erik As I understand it, and from my 3D experience, the specular light component should define the brightness of highlights, on shiny/glossy surfaces, of reflected light sources. In real life, things are illuminated by a combination of direct light, scattered light and caustic light (however, we can't do caustics). The light source dictates the total amount of light entering the atmosphere, which scatters some of the light, mostly in the blue part of the spectrum, producing the ambient light. What is left then makes up the direct light component. So, under a clear sky, the direct light should be a bit yellow and the ambient light should be a bit blue. In fog or clouds though, where the fog or cloud is thick enough to scatter _all_ the light, there should be no direct light but much more ambient light, which should also be whiter as it now includes all of the colours from the light source and not just some of the blue component. Where this becomes relevant to specular light is that with no direct light source, there's nothing to produce a specular highlight. Although when near the ground though, the sky will still be brighter than the ground, so then there will be a slight degree of direct light. The amount of specular then, should be dependent on the glossiness/shininess of the surface (to provide a reflective highlight in the first place) and the amount of direct light. If you're in a bank of fog, where the horizontal visibility is low, but the fog is only say, forty feet thick, you should still get specular affects when the sun is high in the sky. Under an overcast sky though, where you can't make out the sun through the fog/clouds, you should get very little specular even if the horizontal visibility is high because there's no direct light source to create the reflective highlight. Using values up to 45km seem way too high to me - perhaps ~2km high value might be more realistic. I mentioned 'caustics' earlier; this refers to secondary reflected light. You can see the effects of caustic illumination when you stick something non-reflective, next to a coloured surface - the side of the non-reflective object will be illuminated by the coloured light reflected from the coloured surface. Like I said though, this isn't really viable in RT rendering so perhaps the ambient level should just be increased a little to account for it, forgetting about the colour cast because over a large area of different coloured background and objects the caustic colours will average out. Good summary. Vivian -- ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] *** SPAM *** Re: Scene ambient and specularcolor changes
Hi there, let me join this discussion: Very good summary indeed. Pretty much what I was just writing. If you want to get a feel of specular highlights, look at car-paint, and observe what is reflected on the metal that actually causes the highlights. You will find that it is the sun, street-lights or lit billboards. Basically anything emitting a lot of light. As for causastics, I'd say that you add the it to the ambience. You could try and sum up and average the diffuse colours of the scenery tiles every once in a while to get the causastic part for the ambient colour, if you like. You may also want to look at your fog implementation. It may yield a clue on how to determine the amount of light passing through depending on the altitude of the object and density of the fog. You can also try to modify the size and brightness of the highlight, depending on the cloud cover _above_ the object (overcast - clear). Kind regards, Josef -- ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
