Feeling bad for Laminar and others

[Alpine Scenery 1,609]

I don't see Google or Apple having any interest in this market, unless XP 11 approaches them as an integration. Another major player entering into it would cause the market become diluted too much for any of the companies to really make enough. I think Microsoft's interest is because of past success and because they can dominate the market, and MS sees money in the third-party add-on market MORE than the actual SIM itself i think. Otherwise, they wouldn't be selling it for $1.

There are a lot of complex licensing issues when it comes to Ortho, but there are some things XP 11 could do to work around that to some degree.

 

[ShawnG 1,823]

Just now, airlinejets said:

I tend to agree. Laminar would be wise to approach Google.

Other way around.  Austin hates ortho. And probably google.  Google will make the move

[Alpine Scenery 1,609]

I just don't see Google having any interest, there is no money in it for them, as they wouldn't be selling the simulator itself or the add-ons. There is also not a financial incentive for Xplane to give Google any control just to gain some higher-res ortho, when the ortho would have to be downscaled anyhow.

The Bing imagery in MSFS doesn't look that much better than the 1.2m NAIP to be frank, it's just that MSFS overlayed a very sophisticated ground texture to take away some of the lack of detail that the ortho has, and the trees hide much of the rest of the issue. The Bing imagery does have higher res natively though, but it's being downscaled to almost the same resolution as the 1.2m NAIP.

XP 11 ortho is much higher res looking than MSFS if you download 60cm or greater ortho from Ortho4XP.

Edited August 22, 2020 by SceneryFX

[Dominique_K 10,152]

1 minute ago, abrams_tank said:

,Cool, nice to know this.  I mean, if I were Lockheed, the most sensible business decision is to try to make a deal with Microsoft to license MSFS for the commercial market and give Microsoft a cut for every commercial version sold.  For Microsoft, it may be in their best interest to push Lockheed aside and enter the commercial market themselves with MSFS - that's a nice additional revenue stream for Microsoft.

We all see this issue through our own relatively narrow window, the consumer product. And extrapolate from here. But besides the direct stake of how to make profit from one’s simulator, there are larger stakes around the technologies who helped to build the sim, specially coupling powerful networks and AI on GIS Big Data (not only imagery). Big markets ahead, civil and military, and LM is no slouch in AI and imagery interpretation.

[Alpine Scenery 1,609]

6 minutes ago, Dominique_K said:

We all see this issue through our own relatively narrow window, the consumer product. And extrapolate from here. But besides the direct stake of how to make profit from one’s simulator, there are larger stakes around the technologies who helped to build the sim, specially coupling powerful networks and AI on GIS Big Data (not only imagery). Big markets ahead, civil and military, and LM is no slouch in AI and imagery interpretation.

I don't think the commercial simulator market is that big because it is heavily diluted by proprietary systems, but I guess if MIcrosoft and some of the major aircraft manufacturers brokered a commercial deal to use private physics that would never appear in the consumer version. The companies keep their physics models very secret, Boeing could give us the exact simulation physics served on a silver platter if they wanted to. From my understanding, it's all considered highly regarded IP, because if they disclose the physics it allows their competitors to more easily beat their own designs by having a benchmark of what they need to do. I know they are willing to disclose some of it to certain parties at certain times, but they never served it on a silver platter AFIK, like in a whitepaper all at once.

Edited August 22, 2020 by SceneryFX

[ShawnG 1,823]

2 minutes ago, SceneryFX said:

I just don't see  Google having any interest, there is no money in it for them, as they wouldn't be selling the simulator itself or the add-ons. There is also not a financial incentive for Xplane to give Google any control just to gain some higher-res ortho, when the ortho would have to be downscaled anyhow.

None of this is about flight simulators...  and I agree that Austin probably wouldn’t go for it, unless it was Apple doing the bidding, which would be bad news for us. (Say hello to your new Mac-only X-plane!). 

[norman_99 209]

I’d bet the commercial interest in MSFS is more as an image generator  incorporated into a larger system. It wouldn’t be used as the core system itself, just the fancy visual.

[Alpine Scenery 1,609]

Laminar has made some bad missteps here. They should have hired some graphics gurus from one of the gaming companies at least a year ago to start getting the jump on this. The problem is Laminar doesn't pay the same salaries those guys are used to, but they need to hire at LEAST one graphics guru. They have some guys on their team that are pretty good at graphics, but I don't think they have anyone that is at the bleeding edge of texturing like Asobo has (and Rockstar games). I would have attempted to hire one of the guys from Rockstar games that worked on RDR 2 textures to fix the water and clouds and weather. They could have easily beat the weather system by using something closer to RDR 2's weather.

It's not too late for them, they can still fix their missteps, but the weather effects in these sims aren't even close to RDR 2 levels, even MSFS. It doesn't have to be "THAT GUY" specifically, but there are several guys that are known as leading experts in texturing that they could have tried to hire. You get one of those guys on the team, the entire team gets instantly better at graphics because he will teach them the tricks.

 

Edited August 22, 2020 by SceneryFX

[Bob Scott 7,745]

Well, it probably wouldn't be a bad idea for the others to hold off and see how the limitations of this streaming scenery pan out before we insist all the other sims incorporate it.  I'm seeing a *lot* of aberrant scenery, especially near the seams where the sat photos are stitched together.  As an example, fly northwest up B37 from LVD to SSR (Ketchikan to Juneau AK) and there's an ugly seam off to the left where the colors are all mismatched and the snow and grass textures appear blended to give a blotchy yellow-green snow on the the ground.  Northwest of FZNA (Goma, DR Congo) there's an area that looks like a corner where 4 tiles meet...probably the sat passes were at different times and lighting conditions.  It looks pretty bad.  I suspect there are a lot more like that.

Then there's low-level...I did the obligatory flight over my house, which looked pretty good at 2000 ft AGL, then came back and buzzed it at low level, around 200 ft AGL.  The buildings were stretched and distorted like what you might see on the street view on Google Earth...that portends an underwhelming visual environment for helicopters if we ever get them.

The terraforming on airport surfaces is a problem, too.  On my very first MSFS flight, I taxiied the DA62 from near the main terminal at my home airport KCOS, and on Twy H between P and C there was a terrain seam running right along the taxiway centerline that was around a foot high...the plane was dragging the wingtip on the low side, and I had a hard time see-sawing it back and forth to get it unstuck.  I have run into a bunch of other terrain anomalies on airfields, mostly smaller disjointed seams and ditches running across pavement.

There's clearly a "wow" factor when viewed from the right distance and under the right conditions...but I remember that same thought when I first saw photoscenery being demonstrated from high altitude, and before I saw what it looks like down low.  I don't think the book has been closed on this yet.

[Alpine Scenery 1,609]

10 minutes ago, w6kd said:

Well, it probably wouldn't be a bad idea for the others to hold off and see how the limitations of this streaming scenery pan out before we insist all the other sims incorporate it.  I'm seeing a *lot* of aberrant scenery, especially near the seams where the sat photos are stitched together.  As an example, fly northwest up B37 from LVD to SSR (Ketchikan to Juneau AK) and there's an ugly seam off to the left where the colors are all mismatched and the snow and grass textures appear blended to give a blotchy yellow-green snow on the the ground.  Northwest of FZNA (Goma, DR Congo) there's an area that looks like a corner where 4 tiles meet...probably the sat passes were at different times and lighting conditions.  It looks pretty bad.  I suspect there are a lot more like that.

Yah, there are plenty of issues with the Bing Ortho and the color, and the fact it is being downscaled to near 1.2m isn't helping, but the trees are hiding a lot of it, and the detail texture they are overlaying is hiding much of it on the flat valleys. The treeless mountain peaks are the main issue.

As much as Microsoft claimed they used AI here, I see some very minor AI used in the autogen (for sure) and the photogrammetry, but it looks like very basic conditional predicates. I don't see anything here that appears to make heavy use of regression or statistical decision trees or entropy, there might be a little but it wasn't used much.

So there is a ways to go yet, the problem is automatically fixing the orthoimagery involves a sort of very sophisticated color tone mapping algorithm that is incredibly complicated to code. There are only a handful of people that have the know how to do it, but there are many more that given enough time probably could.

Edited August 22, 2020 by SceneryFX

[Tierborn 191]

4 hours ago, tonywob said:

Don't feel bad... Both sims will survive. You're forgetting X-Plane has been around for years and has survived as a niche in a niche. They also have their mobile platform which is doing very well. Many users who don't run Windows also can't use MFS, so they'll be fine.

I didn't get that impression at all. All can happily coexist.

I agree .. Xplane survived when MS flight sim was going very strong.  Xplane first released on Nov 30, 2000  Its been around for 20 yrs now.

[shivers9 1,174]

4 minutes ago, w6kd said:

I don't think the book hasn't been closed on this yet.

Thanks Bob for you inputs. They are spot on as usual. I have wanted to make that point but I am just a Dumb Redneck from NW Florida and I just could not get a paragraph together without using technical terms like "Nit Wit" and almost all the words that you guys have in your "word not allowed" files. LOL!! Jokes aside you post makes good sense.

[Alpine Scenery 1,609]

2 minutes ago, shivers9 said:

Thanks Bob for you inputs. They are spot on as usual. I have wanted to make that point but I am just a Dumb Redneck from NW Florida and I just could not get a paragraph together without using technical terms like "Nit Wit" and almost all the words that you guys have in your "word not allowed" files. LOL!! Jokes aside you post makes good sense.

I agree, Bob should be a scenery reviewer, then we might finally get some valid opinions!

 

[snglecoil 1,828]

2 hours ago, Slides said:

which is that LM doesn't have the skill to do what Asobo did.

Uh...you do know what “LM” stands for, right?

[Kendig 4]

Short version: X-Plane will be fine. For many of us, flight dynamics rules the day.

FS2020 is the shiny new toy and lots of people are checking it out. Many (most?) of the folks who have never bought a flight sim are interested due to the photorealistic scenery, which is quite the achievement even with the (to be expected) anomalies. 
Watching YouTube the past couple of weeks showed the expected dichotomy between “game-play” and “flying the sim” in the titles of the videos. 

I spent a dollar for the first month of game-pass for the PC. If FS2020 makes it out of public beta by the end of that time, I may buy the basic version. I installed it Monday overnight and it has crashed every day this week. On Wednesday I even experienced a BSOD. Yesterday (Friday) did much better. I completed 3 flights and it finally crashed in the the fourth. Total time in the sim was about 4 hours, all in the C172. I find it useful to start FS, let it it update, then reboot the pc before running FS for real.
When it is flying there are lots of minor annoyances. Many settings aren’t sticky; tail number, and such. My biggest complaint about the G1000 is not having VNav.

MFS will succeed due to the gameplay; activities, challenges, badges and the like. (Plus, of course, the scenery.)  XP won’t go away due to the flight dynamics. Don’t forget, the pro version of XP is certified by the FAA for flight instruction.
 

Just my 2 cents.

Cheers

p.s. I’ve been flight simming since the SubLogic days.

Edited August 22, 2020 by Kendig
Added ps