An Idea For P3Dv6

[mpo910 2,112]

5 minutes ago, G-RFRY said:

Yes and P3D is not a Scenery simulator it`s an aircraft flight simulator, i don`t use spot view when flying PMDG aircraft at 30,000+ ft only on take off and landing do you see the airport, you have other sims that do VFR better that`s why MS placed so much on camera views for the XBOX controller in MSFS ready for the game console. 

I would love to see the airports surroundings greatly enhanced. Not these empty spots and trees here and there. But every building and "every" tree should be there. Rivers should look like rivers and oceans should look like oceans.

Here I think is really some room for improvements.

And I think the textures are not up to date anymore. Some regions are really FPS demanding, these regions are sometimes less dense as some others.

EGLL is a good example.....more taxing compared to san francisco....

But I agree....I don´t need the whole world at highest level of details......but what about 30NM around each airport? That would be a real enhancement.

Marcus

[G-RFRY 1,708]

That`s why i don`t buy ORBX TE i don`t need it and no seasons, and as ORBX said TE with seasons takes to much work and they would not be able to charge the price it would need to justify it, P3D will be the only sim with season going forward for some time, if you notice season has dropped of the radar in MSFS not even in the 2022 to do list and possible performance impact with DX12 coming in 2021.  

Edited December 7, 2020 by G-RFRY

[micstatic 4,316]

I agree on the lack of seasons.  However with True Earth Florida it's largely a moot point.  that's one reason why I really enjoy it

[shivers9 1,174]

4 minutes ago, micstatic said:

I agree on the lack of seasons.  However with True Earth Florida it's largely a moot point.  that's one reason why I really enjoy it

I tried TE Florida and to be honest, I just did not like the look of it. I just don't see the point of it at all. I guess it is just a matter of taste. LOL my wife says I don't have a creative gene in my body and she is most likely correct!!

[micstatic 4,316]

16 minutes ago, shivers9 said:

I tried TE Florida and to be honest, I just did not like the look of it. I just don't see the point of it at all. I guess it is just a matter of taste. LOL my wife says I don't have a creative gene in my body and she is most likely correct!!

For airliner flying I've found it to be nice.  Though up close I don't like it as much.  But it blends well with all my payware airports.  And also I like the look of busy urban areas 

[B777ER 2,484]

16 hours ago, dave2013 said:

1. Develop a scenery engine that uses Google Maps terrain instead of landclass and texture tiles.

2. Develop a technique to generate appropriate 3D objects and place them based on imagery analysis

3. Don't waste time and resources trying to dazzle users with super-duper photogrammetry, just focus on accurate terrain and 3D objects.

Now, I know what you're thinking: this is what MSFS does.  Yes, and it's exactly what LM should do, only better.

The core P3D platform is already pretty darn good.  It's a tried and true platform that has been around a long time.  If they can use photo-real Google Maps terrain and render the buildings, vegetation, etc. we'd have a winner.

Just an idea.

Dave

Here is the problem. Read FSElite articles and you can probably see where I am going. New development scenery for P3D is dead. Sure some may still come out and if so, it's because it's already been in the P3D pipeline. Article after article on FSElite for weeks now is just one new airport after another for MSFS. I seriously doubt any big name 3PD's will be starting any new projects for this sim. Once those scenery projects that have been in the pipeline are released, nothing new will be started. You'll see some XPL stuff coming out but P3D is a financial carcass (per their own words) for most scenery developers. This is very unfortunate for those of us that are staying with P3D but this is the reality we must face. Some may disagree but in time they will realize this. It's like building a house. You get the block walls and foundation done but all the interior vendors have disappeared. Only MS pulling the plug on MSFS will change this but with the Xbox part of it, don't think that will happen. Even if it takes to 2022 to get top level planes, scenery dev's don't care, they are making money in MSFS and they will die on that MSF hill before coming back to P3D.

Edited December 7, 2020 by B777ER

[Benjamin J 751]

24 minutes ago, micstatic said:

[...] And also I like the look of busy urban areas 

This exactly. TE provides some wonderful city sceneries, I think. The blend between custom models and the 3DM buildings is pretty darn good, especially in LA and SF, but I thought the Florida east coast looked very nice as well. To be perfectly honest, the main reason I got the TE series (on sale!) was for the city sceneries, because they cover  some cities that nobody else has covered in a feature complete package. For example, Aerosoft's city sceneries were made for LA and SF, but they don't have night lighting and they lack neighboring cityscapes such as Oakland, Century City and Westwood. Without them I don't think the Bay Area or LA metro area are 'finished'. The rest of the TE package is a little hit or miss. E.g. in California I immediately swapped the photoscenery with that of MSE.

But on the whole, personally I'm pivoting to photoscenery wherever possible, because I do feel it enhances the world greatly for me. However, I do look for a balance and get the photoscenery for areas where it matters... For example, the American Southwest benefits immensely from MSE, because of the arid and mountainous terrain that landclass  cannot capture nearly as well. I have also covered the American South and East Coast, through Texas to Florida and from Florida up to Massachusetts. The terrain here is less predictable and has some beautiful mountain areas, such as the Smoke Mountains and Apalachians, which MSE enhances, even if it's not as good as in the Southwest. However, I have left the midwest mostly untouched because OrbX landclass does an amazing job depicting the rural areas, in my opinion.

[F737MAX 5,555]

3 hours ago, dave2013 said:

Can't agree with some comments here about the landclass/texture scenery being good enough.  ORBX has done a phenomenal job with the landclass and texture tiles, but this scenery system will never be as good as photo-real terrain with accurate 3D objects.

I agree that the 3D buildings and vegetation need to be improved, but that is easily achievable.

I'm a big supporter of P3D and have advocated strongly for openLC Asia, which ORBX finally agreed to make as there is indeed a market for it, but I'm under no illusions that this type of scenery is the future of flight simulation.

Photo scenery is currently the most advanced method of replicating what you can see in the real world.
However, it still has some significant limitations, namely outdated aerial imagery, poor quality imagery, colour variations between tiles and shadows from all objects (manufactured or natural) embedded in the ground images.

The future (who knows when), will be synthetic landscapes similar to landclass, but using much, much higher quality of textures and OSM data (let's call it Landclass+).
Landclass+ will replicate the real world as good as any photo can, which would far surpass what we've seen from any sim platform or any 3PD.

Also, Landclass+ would negate the problems of depcting seasons too. There'll be no need to correct summer imagery to the other 4 flight sim seaons.

1 hour ago, B777ER said:

I seriously doubt any big name 3PD's will be starting any new projects for this sim. Once those scenery projects that have been in the pipeline are released, nothing new will be started. You'll see some XPL stuff coming out but P3D is a financial carcass (per their own words) for most scenery developers. This is very unfortunate for those of us that are staying with P3D but this is the reality we must face.

We'll see. It's difficult to disagree with your view.

That said, if MSFS doesn't improve into a competent flight sim platform, then new developers will come to P3D to fill the gap in the market.

 

Edited December 7, 2020 by F737NG

[dave2013 1,123]

2 hours ago, F737NG said:

Photo scenery is currently the most advanced method of replicating what you can see in the real world.
However, it still has some significant limitations, namely outdated aerial imagery, poor quality imagery, colour variations between tiles and shadows from all objects (manufactured or natural) embedded in the ground images.

The future (who knows when), will be synthetic landscapes similar to landclass, but using much, much higher quality of textures and OSM data (let's call it Landclass+).
Landclass+ will replicate the real world as good as any photo can, which would far surpass what we've seen from any sim platform or any 3PD.

Also, Landclass+ would negate the problems of depcting seasons too. There'll be no need to correct summer imagery to the other 4 flight sim seaons.

I used MSFS for about a month and visited many location all over the Earth.  The imagery was very good everywhere I went.  The only issues I found were related to poor quality photogrammetry, which one can do without.  The scenery will get even better as the 3D object placement process is improved and more varieties of buildings, vegetation, and other objects are created.

OSM data is not good enough to accurately depict every square kilometer, or even square mile, as it is in real life.  It does define areas such as parks, residential and industrial areas, schools, hospitals, and other specific businesses and POIs, but not land classification outside of urban areas.  Even with land classification from other sources, lets say an area is defined as cropland or forest for example, you wouldn't necessarily know the specific type of crop or type of forest, so what shows up in the sim would be a guess.  There are also many varieties of desert with varying colors, so if an area is defined as desert how would the sim know what colors to use?  Then there is the issue of the level of detail of the classification data, its resolution.  I doubt that it's better than one square kilometer, but even if it's half of that it is not nearly as accurate as actual imagery of the area.  This is a major limitation of the 1 square kilometer landclass tiles that we currently have.

Therefore, no matter how high you make the landclass texture resolution, the depiction of the terrain using those textures will not be accurate and true-to-life.  It may be a lot better than what we currently have, but I don't believe this method can compete with satellite and aerial imagery.

I am confident that seasonal changes can be achieved with satellite imagery by using shading techniques and changing the vegetation.

Dave

Edited December 7, 2020 by dave2013

[dave2013 1,123]

4 hours ago, B777ER said:

Here is the problem. Read FSElite articles and you can probably see where I am going. New development scenery for P3D is dead. Sure some may still come out and if so, it's because it's already been in the P3D pipeline. Article after article on FSElite for weeks now is just one new airport after another for MSFS. I seriously doubt any big name 3PD's will be starting any new projects for this sim. Once those scenery projects that have been in the pipeline are released, nothing new will be started. You'll see some XPL stuff coming out but P3D is a financial carcass (per their own words) for most scenery developers. This is very unfortunate for those of us that are staying with P3D but this is the reality we must face. Some may disagree but in time they will realize this. It's like building a house. You get the block walls and foundation done but all the interior vendors have disappeared. Only MS pulling the plug on MSFS will change this but with the Xbox part of it, don't think that will happen. Even if it takes to 2022 to get top level planes, scenery dev's don't care, they are making money in MSFS and they will die on that MSF hill before coming back to P3D.

Let's give it time.  MSFS is the shiny new thing right now and yes, a lot of developers have jumped on board that train as they see money-making opportunities.  Only time will tell just how successful it will be.  The only thing I liked about it was the scenery.  It has numerous flaws that are still being "investigated" and features that are "planned" by the devs, and they haven't even yet begun to investigate many important issues.  Just take a look at the development update to see what I'm talking about.  They still show a bunch of pretty screenshots, but the sim has a lot of problems.  Granted, it was released recently and we should give them time to fix things, but I'm not as sure as I was a couple months ago that it is the "be all end all" of flight simulation.

If LM developed a scenery rendering method for P3D similar to what MSFS uses, then it would be a success IMO.

Dave

Edited December 7, 2020 by dave2013

[F737MAX 5,555]

32 minutes ago, dave2013 said:

I used MSFS for about a month and visited many location all over the Earth.  The imagery was very good everywhere I went.  The only issues I found were related to poor quality photogrammetry, which one can do without.  The scenery will get even better as the 3D object placement process is improved and more varieties of buildings, vegetation, and other objects are created.

OSM data is not good enough to accurately depict every square kilometer, or even square mile, as it is in real life.  It does define areas such as parks, residential and industrial areas, schools, hospitals, and other specific businesses and POIs, but not land classification outside of urban areas.  Even with land classification from other sources, lets say an area is defined as cropland or forest for example, you wouldn't necessarily know the specific type of crop or type of forest, so what shows up in the sim would be a guess.  There are also many varieties of desert with varying colors, so if an area is defined as desert how would the sim know what colors to use?  Then there is the issue of the level of detail of the classification data, its resolution.  I doubt that it's better than one square kilometer, but even if it's half of that it is not nearly as accurate as actual imagery of the area.  This is a major limitation of the 1 square kilometer landclass tiles that we currently have.

Therefore, no matter how high you make the landclass texture resolution, the depiction of the terrain using those textures will not be accurate and true-to-life.  It may be a lot better than what we currently have, but I don't believe this method can compete with satellite and aerial imagery.

I am confident that seasonal changes can be achieved with satellite imagery by using shading techniques and changing the vegetation.

When I said "poor quality imagery", I should have followed it with '...in places'. Usually those at further northern or southern latitudes.
Photo scenery in MSFS is very good in the vast majority of locations. Though less fantastic in places such as Alaska, northern Norway and southern Chile, for instance.

Fields, forests, deserts, bush, everglades, savannah and steppe are relatively easy areas to paste in believable ground textures covering significant areas of the globe. So long as the tiles don't repeat as often as they do in vanilla P3D or the older Orbx LC products, you can save large amounts of disk space usage for flight simmers by avoiding using aerial imagery for those areas.

However, I've never found mountains, glaciers and most importantly cities to ever be credible enough to prevent immersion from being broken. It is important to use aerial imagery for those locations (and probably a few others that I can't recall off the top of my head).

OSM does a job, but there are plenty of other types of data sources coming on-line too. We have the ability to track where every tree in a country is located, almost down to the tree type. We also have data showing where leaf fall is occuring each Autumn and farming data that identifies crop type in every field. That is a lot of information to put into good use for scenery recreation.

Don't get me wrong, I love using photo scenery. Right now it's so much better than landclass.
Yet give it a few years, plus some proper financial backing, I think that 'Landclass+' will probably be the next step until we get Petabyte-sized disks and near real-time aerial imagery.

Until then, the positives for using synthetic Landclass+ for the World remain: avoidance of baked-in shadows, much newer data to replicate the World, and the biggest of all (literally), the avoidance of needing a huge disk space required to hold the World's aerial imagery on a PC or on hundreds of data centres for streaming.
 

[G-RFRY 1,708]

MSFS is not going to see off P3D for some users maybe if that`s what they want, you should use the product that suits your needs, LM will continue development even if all avsim users stop using it.

But for XBOX game pass MS would not have returned to flight sim, they offered it for sale years ago to any game developer that wanted to take up the commercial development of the sim.

[Benjamin J 751]

2 hours ago, dave2013 said:

I used MSFS for about a month and visited many location all over the Earth.  The imagery was very good everywhere I went.  The only issues I found were related to poor quality photogrammetry, which one can do without.  The scenery will get even better as the 3D object placement process is improved and more varieties of buildings, vegetation, and other objects are created.

OSM data is not good enough to accurately depict every square kilometer, or even square mile, as it is in real life.  It does define areas such as parks, residential and industrial areas, schools, hospitals, and other specific businesses and POIs, but not land classification outside of urban areas.  Even with land classification from other sources, lets say an area is defined as cropland or forest for example, you wouldn't necessarily know the specific type of crop or type of forest, so what shows up in the sim would be a guess.  There are also many varieties of desert with varying colors, so if an area is defined as desert how would the sim know what colors to use?  Then there is the issue of the level of detail of the classification data, its resolution.  I doubt that it's better than one square kilometer, but even if it's half of that it is not nearly as accurate as actual imagery of the area.  This is a major limitation of the 1 square kilometer landclass tiles that we currently have.

Therefore, no matter how high you make the landclass texture resolution, the depiction of the terrain using those textures will not be accurate and true-to-life.  It may be a lot better than what we currently have, but I don't believe this method can compete with satellite and aerial imagery.

I am confident that seasonal changes can be achieved with satellite imagery by using shading techniques and changing the vegetation.

Dave

I would echo the observation that I found the quality of the imagery generally quite good in MSFS, but I was rather disappointed to frequently find it several years outdated. For example, NYC lacked a couple of high profile skyscrapers, and in SF the salesforce tower was missing as well. This is currently probably the tallest tower in downtown SF, so not having it is kind of like missing the Empire State Buildijg in a rendition of Manhattan. But, It’s possible this got fixed with the recent world update though.

[Mace 1,902]

7 hours ago, dave2013 said:

Can't agree with some comments here about the landclass/texture scenery being good enough. 

It's not that LC is "good enough".  It's not.  But then neither is photoscenery.

Photoscenery has inherent limitations with photo boundaries, contrast/color issues, limitations of seasons/vegetation/lighting.  I've never seen on-the-fly processing that handled it well enough in any flight sim.

Photoscenery is like the nice bad rabbit (of Peter Rabbit fame) -- when it is good, it is very very good.  But when it is bad, it is very, very bad. 

[dave2013 1,123]

1 hour ago, F737NG said:

When I said "poor quality imagery", I should have followed it with '...in places'. Usually those at further northern or southern latitudes.
Photo scenery in MSFS is very good in the vast majority of locations. Though less fantastic in places such as Alaska, northern Norway and southern Chile, for instance.

Fields, forests, deserts, bush, everglades, savannah and steppe are relatively easy areas to paste in believable ground textures covering significant areas of the globe. So long as the tiles don't repeat as often as they do in vanilla P3D or the older Orbx LC products, you can save large amounts of disk space usage for flight simmers by avoiding using aerial imagery for those areas.

However, I've never found mountains, glaciers and most importantly cities to ever be credible enough to prevent immersion from being broken. It is important to use aerial imagery for those locations (and probably a few others that I can't recall off the top of my head).

OSM does a job, but there are plenty of other types of data sources coming on-line too. We have the ability to track where every tree in a country is located, almost down to the tree type. We also have data showing where leaf fall is occuring each Autumn and farming data that identifies crop type in every field. That is a lot of information to put into good use for scenery recreation.

Don't get me wrong, I love using photo scenery. Right now it's so much better than landclass.
Yet give it a few years, plus some proper financial backing, I think that 'Landclass+' will probably be the next step until we get Petabyte-sized disks and near real-time aerial imagery.

Until then, the positives for using synthetic Landclass+ for the World remain: avoidance of baked-in shadows, much newer data to replicate the World, and the biggest of all (literally), the avoidance of needing a huge disk space required to hold the World's aerial imagery on a PC or on hundreds of data centres for streaming.
 

Wow, I didn't realize that such detailed data existed, like the farmland crop type and tree locations.  In that case Landclass+ may well be a viable scenery generation method rivaling what we have with current satellite imagery.

The only shortcoming I can see is that this type of highly detailed geographic data likely only exists for perhaps 50% of the world, and maybe less than that.  Still, it would be a huge improvement over what we have now.

Dave