2 hours ago, Claviateur said:

I think we are again relying on Alpha screenshots that could be produced with so many graphic settings. So in comparison to the official videos, I personally can see many differences in terms of graphic quality. It could be an illusion and I could be mistaken though...  

The "Water Class" seems not implemented or not part of the Alpha who knows... The official video (linked in a previous post) shows what looks like Water Class. But maybe that was part of the Ortho/Aerial image faded through a mask with the dynamic water. Or "Water Class" implemented but not globally... Who knows.

So I think graphic details from Alpha Screenshots must be taken with a grain of salt considering the variety of graphic settings out there while anything related to the worldwide data will be more visible in the beta for sure.

Waterclass ? This is not a landclass/waterclass sim, is it ? Sat/aerial imagery drapes the terrain, not textures chosen in a table. And we have already seen coastal water imagery in past updates. Thats why the ywo Caribbean pics are a little strange IMHO. 
 

24 minutes ago, domkle said:

Waterclass ? This is not a landclass/waterclass sim, is it ? Sat/aerial imagery drapes the terrain, not textures chosen in a table. And we have already seen coastal water imagery in past updates. Thats why the ywo Caribbean pics are a little strange IMHO. 
 

Waterclass is just an expression just like other expressions I sometimes use and that come from other simulators or ESP... 

Yet Waterclass is not related to textures. It's rather a data to classify water and its specs based on xy location and factors.

So Water classification, just like vegetation or architecture classification, is not related to a specific type of asset or engine rendering method or technology. 

In modern dynamic engines, a water class (aka available water data) could be rendered dynamically with the water types the engine is capable of rendering.

I believe what we saw as coastal water could be either the aerial imagery faded with a mask in the dynamic water as I mentioned and perhaps done so either for the sake of the demo, or as a first draft like many other things we observed. Then it could be, who knows, that this transition is removed with tighter coastal masks, to give room to dynamic water rendering. Just guessing of course.

Or they did not deploy this dynamic classification all of the world, just like the trees or footprints or other data driven rendering.

But as for capability, the engine is certainly capable of that. It's the matter of injecting the data that is the classification of these type of waters.

On the other hand, with the absence of data, again, procedural methods could be used based on location and other factors.

Edited March 27, 20206 yr by Claviateur

Villa O’Higgins, Chile

 Waterclass refers to a classification nomenclature as landclass does. Waterclass frame the application of textures through a coordinates table. I do not see how the concept relates to imagery except of course when you have to blend imagery and waterclass driven textures at the border of each. Which is often a PITA in FSX and P3D in my modest experience. That explains the poor representation of tropical waters in these two sims.

I do not understand what you mean by dynamic water. 

 

Edited March 27, 20206 yr by domkle

6 hours ago, Shack95 said:

I agree. I‘m sure it can be done. One way would be to keep the ortho underneath in rivers and shallow waters. In the latest Grand Canyon screenshot you can see this. That‘s also how you do it in Ortho4XP. The only problem might be baked in waves. 

@Shack95 Yes I suppose this is the technique behind this video screenshot. Now if we know where this is on this planet, we can compare it to the Bing imagery.

21 minutes ago, domkle said:

 Waterclass refers to a classification nomenclature as landclass does. Waterclass frame the application of textures through a coordinates table. I do not see how the concept relates to imagery except of course when you have to blend imagery and waterclass driven textures at the border of each. Which is often a PITA in FSX and P3D in my modest experience. That explains the poor representation of tropical waters in these two sims.

I do not understand what you mean by dynamic water. 

 

Sorry we are confusing each other here.

I was talking about Water Classification, like Land Classification etc. These are expressions referring to real world census / data and not the ESP/Sim engine lingo waterclass or landclass.

I know that we used commonly in the context of FSX Waterclass, landclass, just like autogen and they became very specific to the ESP engine world.

But it is a matter of having data about the classification of water same as the classification of land... And then depending on your engine technology, you render this classification with textures or, in the case of the new engines, with a dynamic water (like a dynamic atmosphere or lighting).

Dynamic rendering is usually not related or based on textures but on dynamic values and ranges rendered through the GPU capability vs Opengl / DirectX / Vulkan / etc to simulate the desired effect.

For example, FSX relies on textures to simulate the projected light from the aircraft whereas dynamic light uses parameters to dynamically illuminates the ground. 

Dynamic water could perhaps use base textures, I am not sure in the case of Asobo's engine, but usually adjusting the dynamic ranges of colors vs masks can produce a variety of transitions in water.

Now when it comes to more complex renderings (than just water tones), for example, like the screenshot from that demo video (previous post), I suppose more complex procedural methods must be involved. Of course, unless it is in this case, the transition between the Ortho and the water through a mask...

 

Edited March 27, 20206 yr by Claviateur

I went quickly through Unigine Documentation, (the engine that is more or less similar to Asobo's engine technology).

Here is how Unigine's water is managed and customized (for those who license the engine of course). Well, they still used base material (Aka textures) with of course the dynamic capability of the engine.

https://developer.unigine.com/en/docs/2.2.1/content/materials/water_global_base/?rlang=cpp#textures

 

27 minutes ago, Claviateur said:

Now if we know where this is on this planet, we can compare it to the Bing imagery.

It's Arcachon Bay in France

33 minutes ago, Claviateur said:

 

8 minutes ago, Shack95 said:

It's Arcachon Bay in France

 

Thanks! Ummmm at first glance, it is as you said, a transition between Ortho and Water à la O4XP.

Now the camera angle of MSFS makes it tricky to compare both imagery. Moreover, the imagery is certainly processed and not implemented as is for sure.

This makes it seems at places, not the same image but I believe it is (processed) and the technique you described is most likely, the one used here.

Edited March 27, 20206 yr by Claviateur

@Claviateur I guess you're right.

Skardu, Pakistan

@Shack95I did a quick Photoshop work to simulate in a primitive way how the effect of water vs land transition could have been done in the engine. (Similar to O4XP).

Of course, the Engine is way more sophisticated than this but the overall method could be the same.

46 minutes ago, Claviateur said:

did a quick Photoshop work to simulate in a primitive way how the effect of water vs land transition could have been done in the engine. (Similar to O4XP).

Great work! Looks like in the screenshot. If I remember correctly there‘s also a video sequence showing this spot. 

3 minutes ago, Shack95 said:

Great work! Looks like in the screenshot. If I remember correctly there‘s also a video sequence showing this spot. 

Yes the screenshot comes from the sequence shared in a previous post here 🙂

 

@Shack95 And I think we can see the land Ortho/water Mask effect on this coast as well (a screenshot from the multiplayer video)