In-game VS Satellite

[Shack95 2,681]

Colta, Peru

[Shack95 2,681]

Cotahuasi Canyon, Peru

Edited April 3, 2020 by Shack95

[Shack95 2,681]

Southeast of Islamabad, Pakistan

[Shack95 2,681]

Sheikh Zayed Mosque, Abu Dhabi, UAE

[Shack95 2,681]

Miami, USA

[Farlis 2,350]

The Pueblo Viejo mine picture shows once more, that they really need to work on the implemantation of waterclasses.

In the source image one can see that the lakes on the upper left next to it are also polluted by the chemicals from the mine, in the flight-sim that is missing.

Edited April 3, 2020 by Farlis

[Claviateur 789]

42 minutes ago, Farlis said:

The Pueblo Viejo mine picture shows once more, that they really need to work on the implemantation of waterclasses.

In the source image one can see that the lakes on the upper left next to it are also polluted by the chemicals from the mine, in the flight-sim that is missing.

Although water class data could be available to use or the different major classes could be established and applied procedurally for global water types, such particular cases for the so many inland water surfaces on this planet might not be obvious to implement unless someone goes manually through all the water masks and checks what water surface type is this or that. A huge task.

Then again, we might think the AI scanning could bring such information if asked to... Well yes it could but I am not sure it's high priority.. I could be mistaken.

But then again, let's think about it: unless the person flying there is a local and knows that it's a polluted water or unless people have Bing opened and synced with MSFS so they verify every piece of land and water, I doubt other users would notice that the water type is not accurate...

 

[mausmaster 2]

What I'm still very curious about is whether there's anything they can do about the baked shadows. 
In all of the mountain shots, when you compare the shadows in the satellite image to the simulator, they are at the same place. You never really see where the sun is but it seems to me that these screenshots are taken at the right time of day, so the shadows of the sim are not overlapping the satellite ones (In the Madrid screenshot, you can see the overlapping/double shadow)

I think in some scenarios it can look really weird if there are for example 2 dark sides of a mountain. The same counts for the buildings with baked shadows of course. 

What I'm interested about mainly is if it would be possible for an AI to detect shadows? Maybe calculate them according to the time of day and position when the image was captured.. and compare that to a height data or something? (maybe this discussion is not for here, if so, feel free to move it somewhere else) 
And if that is possible, could it also remove the shadows? It's not just a matter of color correction...
 

[Dominique_K 10,156]

1 hour ago, mausmaster said:

What I'm still very curious about is whether there's anything they can do about the baked shadows
 

Curiously, the  baked shadows/double shadows issue never appears in the « top alpha issues » of the « feedback snapshot ».  

[GlideBy 324]

6 hours ago, Claviateur said:

Although water class data could be available to use or the different major classes could be established and applied procedurally for global water types,

As you said, they could use AI to at the very least pull the water color.  They don't need to know if it is polluted, they just need the look.  It would look better if they just copied that part from Bing maps.  And as far as a non-local noticing, well some of these places are famous.  But even if they aren't, part of the reason people want to fly this sim is to see the beauty of the world.  So getting the color of local water is important.  Not good if the whole world has the same blue water, and not even different shades.

The other thing they need to do is get different waves for different sizes of water bodies.  An ocean shouldn't have the same waves as a pond.

[adino 263]

13 hours ago, Shack95 said:

Pueblo Viejo mine, Dominican Republic

Nice to see the Caribbean being represented 🙂

[adino 263]

Agree the water type classifications should be WIP for Asobo, I believe that AI is capable of learning that too it shouldn't be that difficult isn't?

[767lover 285]

On 4/4/2020 at 12:22 AM, adino said:

Agree the water type classifications should be WIP for Asobo, I believe that AI is capable of learning that too it shouldn't be that difficult isn't?

Yes, agree water color is major. Some of the mountain lakes, like lake Peyto in Banff are very intense and opque blue, which is very characteristic of the area. I do hope they get this right, as it is a standout feature of such landscapes.

It would also be incredible if they could simulate mountain rapids to some degree. Even if it is some years after release, it would be fine for me, as long as I know they are working on such things. 

Otherwise, I think the water looks great so far in the sim. 

 

 

 

[Victoroos 205]

On 4/3/2020 at 1:31 PM, Shack95 said:

Sheikh Zayed Mosque, Abu Dhabi, UAE

it is so white 😮

[Farlis 2,350]

10 hours ago, Victoroos said:

it is so white 😮

Just google it. It is insane. 

I'm a bit torn on it though. On the one hand it is so breathtaking that I could see it still admired as an architectural marvel in 400 years.

On the other hand it looks not even remotely like its architecture belongs into the year 2007, or whatever 15th Century year the muslim calendar that is.

Imagine someone building a new cathedral and choosing to build it completely in gothic style. I would consider that bad taste and a lack of imagination.

Architecture of sacral buildings has evolved, so why would someone build a mosque in a style that was modern 1300 years ago?

Here is what I mean: https://en.wikipedia.org/wiki/Faisal_Mosque#/media/File:The_Shah_Faisal_Mosque,_Islamabad.jpg

That is a mosque that looks like it belongs into the time it was build in.

 

Edited April 6, 2020 by Farlis