[James Ross]

Genma Saotome, on 17 January 2014 - 04:31 PM, said:

Should weather conditions and/or time of day alter the fog density or the view distance? Beats me.

Weather should set the fog density, but not viewing distance IMHO. (Fog density has no effect on performance by viewing distance has a massive effect.)

[Genma Saotome]

James Ross, on 17 January 2014 - 04:54 PM, said:

Weather should set the fog density, but not viewing distance IMHO. (Fog density has no effect on performance by viewing distance has a massive effect.)

That was my first thought too James but what if fog density, due to weather conditions, is opaque at 1/3 of max view distance? No real need to render anything further away.... Or is all of that already in place?

[nicober]

HelLo!

I make a little video to show the difference of the frog transition from original OR and the third test make by BB25187. The transition is very smooth without this effect of a wall becoming in front of the camera.

In the first part, I used the version X-1931 for the original frog in OR - In the second part, I used the latest Open Rails sources X-1944, plus the third modification suggested by BB25187 and rebuilt the program. In both case I click on the keys CRTL + ( + ) for each step of frog, never in continuously.



That would be a good addition for the next version of Open Rails. I think so.

Cheers!

Nicober

[BB25187]

Hi James,

Thanks for the explanation on the reasons of the choice made when first implementing this fog fading. It is clear now.
As for the idea of mixing the two approaches, it makes sense. It will probably require some tests to find a good tradeoff and have a smooth transition between the two models. Anyway, I attach the patch of the modifications illustrated in my former message. I preserved the four variants (including the original code). The first three ones are commented out. Feel free to use any part of this patch if you think it is useful.
Thanks Nick for the video. It is very demonstrating. Do you confirm that even with the current OR code, the fog is present in the foreground? Or maybe this is a consequence of the point of view / framing of the video?
By the way, the smoke should also be impacted by fog. This is not the case today. The change is probably not straightforward, since fog and smokes are implemented in two different shader files. If I have time, I will check how this could be done.

Regards

Attached File(s)

•  SceneryShader_Fog_SVNPatch.zip (610bytes)
  Number of downloads: 211

[Csantucci]

What I still miss in fog display is that the sky is not subject to fog. Also in Nicober's video in both versions you see the clouds in the sky and see well the horizon (skyline). In reality when you have fog you have fog all around you, even towards the sky, and you don't see a neat skyline or don't see it at all.

[James Ross]

Smoke should be fog-able reasonably easily, but the sky's trickier since it is at "infinite distance" so I'm not really sure how it should be fogged.

[RTP]

I fully agree with Csantucci.

In my experience, standing into a thick fog is like stand into a bottle of milk.
No visibility in any direccion.

Of course, no horizon at all.

Please excuse my english.

Regards.

[roeter]

Walter Conklin, on 18 January 2014 - 06:13 AM, said:

Hello,

In reality , I have seen instances of thick fog and mist hovering around ponds, lakes, and rivers, but the rest of the landscape does not have the banks of thick fog and mist.

Quite true. I have even seen situations where the sun actually shone through the mist - the fog was really just a 'blanket' only a few meters high. That's known here as 'ground frog' (EDIT : :) ground frog??? eehmm - let's call it ground fog :)) and occurs on cold winter days if there is no wind. Sometimes you just see the head of a cow or horse sticking out of a 'wooly blanket'.

Another interesting phenomena would be fog banks - but that would be hard to simulate : one moment you're in the open, the next you just run into a thick wall of fog.

As for the sky : why not just use a sky with only random variations of light gray (or white in winter snow) - that is pretty close to what it looks like most of the time in reality.

Regards,
Rob Roeterdink

[Csantucci]

roeter, on 18 January 2014 - 07:27 AM, said:

As for the sky : why not just use a sky with only random variations of light gray (or white in winter snow) - that is pretty close to what it looks like most of the time in reality.

Regards,
Rob Roeterdink

This could be tested, however the transition at the horizont must be much less sharp as now.

[BB25187]

Hi all,

I agree with the point raised on sky and horizon. This is especially (mainly) true for a very dense fog (with a light mist, the sky, the horizon line and the sun often remain visible). The idea proposed by roeter is Worth being tested, because as emphasized by James, the sky dome is a very special object which cannot be handled as regular scenery ones.
As for the ground fog, this is a good idea too. I found some websites which describe some method which make use of an elevation map to mimic such effects.
In the meantime, I made some tests on a route with distant mountains (the black forest route). I choosed a medium fog density, something like a light mist. The main purpose of this test was to verify the aspect of the mountains in the background.

http://imagizer.imageshack.us/v2/800x600q90/194/byvv.jpg

http://imagizer.imageshack.us/v2/800x600q90/10/ya3g.jpg

http://imagizer.imageshack.us/v2/800x600q90/690/e7ba.jpg

http://imagizer.imageshack.us/v2/800x600q90/199/bh8x.jpg

http://imagizer.imageshack.us/v2/800x600q90/839/f4ce.jpg

http://imagizer.imageshack.us/v2/800x600q90/14/0a8v.jpg

http://imagizer.imageshack.us/v2/800x600q90/197/e7zm.jpg

http://imagizer.imageshack.us/v2/800x600q90/208/yyvh.jpg

http://imagizer.imageshack.us/v2/800x600q90/541/5573.jpg

Cheers