11 hours ago, WarpD said:

Here's Prepar3D v5...

 

Not sure how you can be disappointed by the MSFS version over everything else that's available.  Honestly.

<sigh> My point was to report to Asobo if they read this forum that the Bangkok scenery is wrong . Since early on, they have asked for the community's feedback and I think it s a good thing for the final release. If you are content with what you see in the FS20 Bangkok screenshot just because P3D Bangkok looks like vomit, to each his own but you don't put the bar very high. Asobo said many times that they strive to do their best. They have to try harder for Bangkok. 

2 hours ago, WarpD said:

ESP doesn't exist any longer.  H 

1 hour ago, RALF9636 said:

 To me that is a purely academic discussion.

The trees shouldn't hide the forest.  There are a lot of new trees, great improvements made by the LM team.  But, at the end of the day, the forest is still the same : has the basic flight architecture changed  ? Has the terrain layering (mesh, LC/imagery, autogen) and resolution changed ? Air dynamics ?  The basics are still here. Or so we see.

Don't make me wrong, I  am no P3D basher,   the good ole ESP has still some life in it but if  Asobo delivers and that remains an open question,, it is still to be shown that the ESP can be technically brought up to the same level. 

As to say that 1/ P3D is "not for you"  ie for the training industry and not the consumer market, I will believe it when I see their accounting books 2/ MS couldn't sell FS20  to the said training industry, you really believe that they will forsake a high margin market as such ?

35 minutes ago, Dominique_K said:

But, at the end of the day, the forest is still the same : has the basic flight architecture changed  ?

True. That's basically why I chose XP over P3D a year ago. Until MSFS is released at least. With the little time I can invest into simming at the moment, I guess I don't even care much about initially missing addons.

12 hours ago, WarpD said:

Here's Prepar3D v5...

Not sure how you can be disappointed by the MSFS version over everything else that's available.  Honestly.

Omg that looks terrible I don't recall my FSX looking that bad.

I suppose one important factor is as bings 3d photogrammetry gets better our sims and visuals can get better when Asobo updates the world serverside.

Where as P3D is still running autogen from 12 years ago. 

2 hours ago, Kpeters said:

Omg that looks terrible I don't recall my FSX looking that bad.

Same. Just checked with my FSX. So much better. Considerably too. I don't get it. 

2 hours ago, Doug47 said:

Same. Just checked with my FSX. So much better. Considerably too. I don't get it. 

Post a screenshot

On 5/1/2020 at 11:04 PM, Claviateur said:

It could be that the procedural algorithms during the implementation of the infrastructure make this footprint grouping happen.

I think I originally identified the location 😁

The scenery definitely looks plausible, actually amazing in comparison to P3D.

Asobo is using OpenStreetMap and the MSFS scenery seems to matches the OSM footprints.

https://www.openstreetmap.org/#map=18/13.74321/100.49668

 

 

 

27 minutes ago, bonchie said:

Post a screenshot

When I get the chance, I will. 

18 minutes ago, MatthewS said:

I think I originally identified the location 😁

The scenery definitely looks plausible, actually amazing in comparison to P3D.

Asobo is using OpenStreetMap and the MSFS scenery seems to matches the OSM footprints.

https://www.openstreetmap.org/#map=18/13.74321/100.49668

 

 

 

Oh you are absolutely right. Thanks for sharing the OSM area. I was trying to analyze the grouping process heh 🙂

This is another hint that OSM is being used where local data is not available or not available yet.

And purely procedural formulas are also applied like in the neighborhood of Tokyo. 

Knowing that MS most probably shared millions of automatically scanned footprints for North America with OSM a couple of years ago, I suppose using OSM data even for the final version is something very possible.

In fact the AI scanning Bing imagery to fetch footprints for the planet was probably planned or undergoing, I doubt... Or it could be something they put on ice considering many factor including the world crisis we are living...

Let's see now if the other OSM data types will be used as well (roads, power lines, railways etc.)

Edited May 2, 20206 yr by Claviateur

@MatthewS Here is another evidence that you are right about the OSM data. You might already know this site. Here you can see how the OSM data is rendered as 3D infrastructure on the map.

We can see that this area has no tags especially not for building heights, hence the same height applied by default in the preview

I set the view to be similar to the MSFS screenshot:

https://osmbuildings.org/?lat=13.74688&lon=100.49565&zoom=16.8&tilt=45&rotation=46

So I presume Asobo applied procedural rules most probably to make the buildings have the variations we see.

Edit: Also empty areas were filled probably procedurally with infrastructure...

Edit2: Is Bing using OSM footprints? Now did Asobo used Bing's footprint that are similar to OSM 🤓

https://www.bing.com/maps?osid=2a080070-606d-46c3-a3b4-0382a9b1b2b1&cp=13.743239~100.496507&lvl=18&v=2&sV=2&form=S00027

Now we all agree that the Poor AI was not responsible of this process or any global scenery generation 🙂

 

Edited May 2, 20206 yr by Claviateur

29 minutes ago, Claviateur said:

We can see that this area has no tags especially not for building heights, hence the same height applied by default in the preview

Some of those apartments blocks in MSFS have massive footprints, I don't think you would see such "mega-structure" apartment blocks even in the soviet era USSR.

IMO Asobo should err on the side of smaller buildings and presume that footprints over a certain size are "data-errors", and fill the footprint with smaller 1 storey (75%) and  2 storey (25%) buildings.

I believe they said somewhere that they use the AI to try and match roof colours and If you look at the roof colours of those auto gen buildings, most of them do match or are at least close to how they look in Bing's satellite imagery.

Not sure what they use to determine building height.

Edited May 3, 20206 yr by Tuskin38

1 hour ago, MatthewS said:

Some of those apartments blocks in MSFS have massive footprints, I don't think you would see such "mega-structure" apartment blocks even in the soviet era USSR.

IMO Asobo should err on the side of smaller buildings and presume that footprints over a certain size are "data-errors", and fill the footprint with smaller 1 storey (75%) and  2 storey (25%) buildings.

If they used OSM, then they processed .osm files as they are at least this is what seems to be the case.

And the footprints in this area are traced by people who decided to draw wide infrastructure for some reason, instead of the smaller ones we see in aerial imagery. Probably because they wanted to go fast who knows... 🙂  Contributors to the OSM database sometimes trace footprints in a more precise way than others and some take the time to add tags while others just trace the shape... 

So Asobo took the country or region's data and processed it. Same process with the OSM data of other regions and countries...

And here, we are focusing on one tiny spot in this globally processed OSM adata... This sector of the city is one little drop in the ocean of the processed OSM data files (again f they really processed OSM as it seems to us). 

I did the same process for the other simulator and I did it for 4 geo tiles. I used procedural rules and then when I flew over it, I noticed all sort of things resulting from the data quality... This is how it goes with OSM procedural scenery...  This automated process is done with rules and filters...

All in all, Asobo won't be visiting every city on this planet to inspect the results or analyze the quality of the OSM footprint shapes based on reality or compare it to the Bing imagery... 

 

Edited May 3, 20206 yr by Claviateur

I was originally under the impression that orthophotos of the entire earth were scanned by AI to create the building footprints in MSFS (the same method they used to place trees) however there are many large areas in Bing Maps (for instance, large portions of Bangkok, located just outside the area covered in the screenshot shown in this thread) that do not have building footprints for existing buildings. This leaves me wondering if those locations will not have procedural buildings placed on top of the terrain in MSFS. That would certainly take away from the realism in those locations.

11 hours ago, Claviateur said:

 

Now we all agree that the Poor AI was not responsible of this process or any global scenery generation 🙂

 

I am afraid I don't belong to the 'all'. The AI is of course responsible or rather how it has been programmed. It needs to be trained to interpret imagery, it is a long process of trials and errors and what the picture show is its young age and lack of supervision. Training could mean, for instance, having now a constraint telling no building higher than two storeys in a well defined Bangkok area. Your second arrow from the left is, in real Bangkok, a perimeter wall. Other arrows are low buildings. Same could be said of the buildings along the river.

As a generic  user I am only interested by the end result not the technicalities explaining me that I should expect it to be  wrong.  Asobo said they were serious about delivering the best, and I think they are, they'd better have a systematic review of high visibility places like Bangkok.