Impressions of the New MSFS...From the Guy Who Developed FSX

A little background. I know Hal Bryan. I’ve drank beers with Hal Bryan. I first met Hal at the International Flight Sim Convention in Denver back in 2006. Hal was the head of ACES Studios and the project lead for FSX. He was also the most surprised person when MS yanked the carpet out from underneath him and his team.

He now works for EAA. Not sure what he does there, but MS invited him back in September to take a look at the early build of MSFS. These were his impressions.

Hal, you’re a great guy. Thanks for the open bar.


That’s all fantastic news that MS has taken this part of their roots to heart. Backwards compatibility, live real world data and all that as well. It’s sounding more and more like a winner.

Thanks for linking PFunk. With all of the details Mr Bryan drops, many of which were new to me, backward compatibility with current third party aircraft for instance, that was an extraordinary read.


Hal is a really interesting guy. I emailed him after ACES got axed and I remember him being pretty upset, but taking it in stride, which is pretty normal for him. I felt awful for the guy, knowing how much work he’d put into it. He got started with ACES as deputy project director for FS9 Century of Flight. They promoted him. He developed FSX. They axed him.

A lesser man would have been destroyed. Hal just kept on rolling.


Wow… That was a great read!

That just rocked my world! I’ve been totally ignoring the MSFS thread. It actually pissed me off. I realize that I am a bit tribal about X-Plane. The Flight thing a few years ago wasn’t helpful either. But there is no denying beautiful. Can’t wait.


Great read, sounds promising!

Interesting…as mentioned above an incredible read - great insights.

As the Mudspike staff curmudgeon…and seeing as how going into may 3rd week having to wear a C-colar and not being able to “fly” has made me extra grumpy…let be post a couple skeptical observations, that in no way disagree or argue with the other replies.

…many of the scenery areas boast a resolution of 3 centimeters per pixel, while the default in FSX was about 1 meter per pixel.

Throw in 1.5 trillion trees, individual blades of grass modeled in 3D, and a complete overhaul of lighting and shadows, and the result is an unprecedented level of detail …

Can you really tell the difference between 1 m and 3 cm form any altitude above about 1,000 ft to 2,000 ft AGL? Or put another way, at what altitude do you stop seeing two objects 3 cm apart as two objects and start seeing them as 1 object; then at what altitude do you stop seeing two objects 6 cm apart as two objects and start seeing them as 1 object; repeat until you get to 100 cm (1 m). From experience (not flying but something else - see “Spot Size” hidden section) that happens pretty fast even with a good, high fidelity monitor…in the current state of VR?

Spot Size

If I may, let me introduce the concept of “spot size”. No, this is not a laundry detergent thing. Spot Size is a concept used in photo reconnaissance and imagery interpretation. Essentially a spot size is the measure of resolution. It is the distance between two separate objects where, due to the resolution, they look the same; typically measured in inches or centimeters, feet, meters.

For example: If the visual difference between two variations of a class of armored vehicle is the placement of two exhaust ports–6 inches apart for vehicle variant A and 12 inches for vehicle variant B–then the spot size needed to determine which type of vehicle it is would be less than 12 inches…i.e. the resolution needed to see two distinct/separate objects (vent ports) on variant B as opposed to a single “blob” on variant A.

Different pieces of military hardware have different spot sizes for different purposes. If you just need to be able to determine aircraft type between transport and fighter, then your spot size is pretty big. Between Mig-21s and Mig-23s then your spot size is a bit smaller. However, if you need to be able to tell the differences between Mig-23 Es and Mig-23 Bs you have a much smaller spot size requirement. There used to be a big reference book of various military equipment and associated spot sizes for determine types, classes, versants, etc.

To achieve a desired spot size you have two main variables to work with: The focal length of your “camera” and the altitude of the aircraft. Some photo reconnaissance systems allow you to “swap out” lenses so you may have two or three focal Len Goths to work with. Other systems you only have one lens so you only have altitude.

There is an equation where you set spot size requited and focal length and kt will give you altitude in feet. Fortunately the Navy had little “Wiz Wheels” - circular slide rules - for figuring it all out…they probably stole them from the USAF but…anyway…

In the late 1980’s I worked in an F-14 squadron that flew the TARPS reconnaissance pod so I had some experience with determining minimum altitudes to get specific spot sizes. Even with a large format camera with a long focal length, getting a 3cm spot size would have driven the mission way down into the AAA treat…replace that with a monitor resolution and a Mark I Mod 0 human eyeball…I don’t have my Wiz Wheel but I’m thinking it is not that high…which is why I sort of roll my eyes when I see a 3 cm resolution touted as it is. Add in seeing individual blades of grass…which I can just barley do from my second story window (I just tried it)…you get the idea.

Microsoft is incorporating a legacy mode that it expects will provide near-complete backward compatibility, so those of us who have huge libraries of old favorites won’t be starting entirely from scratch.

OK…isn’t this one of the factors that killed the original MSFS line? The requirement to be backwards compatible with at least the last version meant they could not really change the engine?

Also missing from the build was ATC and AI traffic, …

…so essentially the same state as the current XP11. :open_mouth:

Thanks to @PFunk for posting this. A very interesting piece!


Usually big numbers are the latest & greatest - but in this case you can’t just let us enjoy our small numbers, can you? :sweat_smile:

Just think of it like this - no appreciable loss of fidelity for those hyper zoomed-in screenshots that we all enjoy sharing!

1 Like

It’s nice to know I’m not the only one.

I’m curious to see where this ends up going, as we get more information about flight model, AI traffic, and real-world practicality with hardware and connection requirements.

1 Like

Flying around Google Earth VR gives some idea of how this could look, minus the weather. I can see where cloud-based terrain could shine relative to XP and where it might come up short. Areas of real topographical complexity really pop. Some areas are simply jaw dropping. No 4K mesh textures I have seen yet in XP come close. But most of the world is like my humble New Jersey. There are spots of brilliance but mostly it is pretty dull. X-Plane does this very well. Their use of OSM data and regionally appropriate buildings (about half the time) create a plausible representation of the real place.

That said, I fear that the days of a scrappy genius coding a simulator in his basement [or] a team of scrappy devs teamed up around the globe will quickly get totally steamrolled by the might of Microsoft. Today’s computers favor works from the mighty. MS will apply its cloud computing and AI might to put the right houses and the right trees in the right places. AI traffic? ATC? Puu-leeeeez! Of course they will get that stuff in.

But maybe I am wrong. Usually I am, fortunately. Maybe I am totally underestimating Laminar’s ability to adapt. The fact that MS courted the EAA signals to me that they intend to appeal to pilots and hobbyists. If so, that will really turn the screws. Now…back to ignoring MSFS!

1 Like

Maybe not. But making the sim look good at lower and lower altitudes has always been the name of the game. Otherwise we’d all be happy with old Mega Scenery that looked great from 40,000 ft!

Traffic pattern work in GA aircraft, bush flying, helicopter operations, etc… all benefit from the great resolution down low.

1 Like

Not to mention youtube streams of horrific crashes. They need that last 20 feet detail.

@Hangar200 I really dug your explanation of spot size. Thank you very much sir!

Truer words have not been spoken. :slightly_smiling_face:

My point was more towards why focus on that? I feel like I’m listening to the car salesman touting the cup holders. He had me at “Throw in 1.5 trillion trees”.

Seriously…1.5 Trillion trees! That is simply awesome. I am more than a little tired with squinting to blur my vision when I look at a forest so that I just see the trees texture and not the few individual trees.

I’ve done a number of scenery areas for FSX (aside from Pilate’s Ghos and ENSB I think that I only have 2 on line). Inevitably you get a small or thin area of trees that just will not sho more than or two trees…and just as inevitably this area is right near a runway…so you end up “planting” your own with individual scenery objects…frustration.

1 Like

That sounds pretty good… :+1:

1 Like

And it’s a reason why the helicopters will not be ready at the release.

If they model thousands of parts of the whole aircraft and calculate drag, lift and other forces for each of these, try doing the same for the fast moving (rotating) airfoils and make it flyable for armchair pilot. I bet there will be much tweaking going on in this department.

No, that’s not it at all. Rumor I heard was that they couldn’t find a master of the dark arts to summon the demon ritual for helicopter flight physics.


I have one here if they still searching…

1 Like

IIRC (do kids still type that?) FSX never modeled helicopters. Those things that looked like helicopters were just fixed-wing platforms with a table of properties which mimicked helo-like flight characteristics.

1 Like

Helicopters are not real. They are a strange manifestation of our conscious minds upon witnessing that which human eyes were not meant to see.