A New Simulator (June edition)

[Staffan 1]

Great news Mathijs.

Long time since we spoke.

My wishlist for a new simulator looks like this:

1. Sloped (curved) runways

2. Cloud shadows

3. Realistic EPR variables

4. Intelligent AI

5. 2D panels

6. A possibility for 2D cockpit sideviews (in FSX they "forgot" about this important function)

Well thats it for the moment. Additional issues may come later.

[Tim_A 1]

It's a brave and ambitious undertaking, and I wish you the best of luck.

I was bitten by the initial rush to buy FSX, as many were, and vowed at that time not to rush out and buy another sim without very careful consideration. Which means you have an uphill task persuading people like me to buy it. I'm now comfortable with FSX, with the kit to run it fairly well. Any new sim is going to need to give me something substantively more than what I already have, or I'm simply going to stay with FSX (just as many still are staying with FS9 since they don't see FSX as giving them substantively more). Which means I think that you've got a long, hard and expensive task ahead.

But all that won't stop me from throwing my ideas into the pool!

Extensible Frameworks and Interfaces. Rather than hard coding systems, define a framework or interface for it. This will give the developers the ability to extend the sim and build areas outside of the core that are either specialist, or that Aerosoft simply don't have the time/budget/whatever to complete. It wouldn't be so important then if (say) bird strike damage isn't implemented -- if there's an interface for it, it can be added by 3PDs

ATC. If you want in-game AI, I don't think its feasible not to have some means of controlling it, and separating it. By and large, the existing ATC system doesn't do a bad job, but has some glaring flaws.

* Separation and sequencing, including holds etc.

* SIDs and STARs

* Route tolerances & corrections (zig-zagging)

* Handoffs (backwards and forwards between lots of controllers)

* Refusals and emergencies, including ATC refusing zone transits etc

* Regionalisation:

- Local procedures for ATSOCAS (ie no universal flight following, but support for FIS, RIS, RAS etc)

- Better support for QNH/millibars (FSX has a strange requirement that to get millibars you must also have kilometres!)

- regional transition levels

- Extensible support for more ground station types including AFIS, Air/Ground, and the annoying little French airfield that closes for two hours over lunch, grudgingly uses English while it's open but insists that you speak in French out of hours, etc.

- Regional accents. Here's where that interface really works, since you don't need to supply all the accents, but just the interface for others to do it!

Interaction with the environment.

* Hitting things should have consequences, whether that's a bird strike / cow /cute bunny-wunny, or a plank of wood placed on the ground to be used as a chock. I guess I'm talking about some sort of mass/weight/physics engine. Right now, if I run off the end of the runway into a hedge/ fence / trees / grass berm etc. I'll just pass through it or trigger a bizarre 2000ft bounce up into the air. But it should create a prop strike, dented wing, gear failure etc. (or other effect defined by its interface...)

* Planes should be chockable, both with "chocks" and scenery objects that can be used as chocks.

* Surface friction for runways & other ground surfaces. In FS9/X, planes skate across the runway as if they are on ice

* Turning circles. Probably linked to the friction issue, but lots of planes can be seen going almost straight ahead with the nose wheel at nearly 90 degrees, regardless of how slowly they are going.

Scenery

While I'm a big fan of a decent photo scenery, I'm not sure that massive areas covered via Google Earth etc are a good idea. Why?

* High resolution coverage is limited to North America, parts of Europe, and small 'keyhole' areas in other places. There are vast areas of the planet where no decent satellite (yet alone high res aerial) imagery exists at all, and won't for decades to come (Africa, Asia, South America...). By adopting photo scenery as a base, you will effectively limit the places to which you can fly, and I certainly won't buy into a sim that isn't global in scope from the outset.

* Colour matching and gradation. Photos as supplied by Google etc are not colour graded, and matching them is a huge task that current dedicated scenery producers are still only achieving partial success with.

* Seasons. Photo reconnaissance typically only happens during the "nice" months, with no seasonal consistency within that band between consecutive tiles, yet alone producing five distinct season sets.

While I'm talking about seasons, the current five seasons in FS could be seen as too coarse. If you fly RW around any area a lot, you will notice that the scenery can change dramatically from week to week. There's the progression of crops from ploughing, growth, flowering, harvest, two may be three times a year, or possibly fallow over the winter. Even general scenery is markedly different from early Spring to late spring, to early Summer etc.

Anyway I could go on, but it's getting late, and you're probably getting bored, so I'll shut up!

[airfrance 0]

A good simulator, you say?

1. Realistic landscape

a. really changing landscape (instead of flying all Europe seeing the same textures)

b. realistic different towns and cities

c. 3D-looking forests (especially at lower altitudes)

d. regional autogen flora (based on realistic biodiversity)

e. regional autogen architecture (no more of the same building types everywhere!)

f. not repeating ground textures (a good algorythm creating fictional but realistc looking texture diversity)

g. no sharp-line boundaries between landclass tiles

h. convincing texture sets (deserts looning like deserts, snow like snow, etc)

i. no ground blurries finally, please!

j. realistic timezones

2. Realistic weather

a. real weather influence (runway friction, icing, humidity, OAT, etc)

b. realistic haze and fog (with gradual visibilty, light effects, etc)

c. diversed clouds (just like ground textures mentioned earlier)

d. windshield wipers visually removing rain and snow

3. Realistic ATC

a. considering SIDs and STARs

b. considering local transition altitude / transition level

b. managing AI very well (no go-arounds, frozen queues, etc)

4. Realistic navigation and flight planning

a. updatable navdata (maybe paid and done by Navigraph?)

b. a realistic flightplanner (something like FSNav/FSBuild)

c. a moving map with selectable details

d. standard FMC (might be a bit simplified) integrated with the ND, instead of GPS

5. Realisting airport services

a. standard pushback and followme

b. basic animated ground staff (marshaller, wing walker, etc)

c. GPU and refuelling really working

6. Realistic traffic

a. AI visible from a long distance (with gradually dissolving 3D contrails)

b. basic ground service for AI (jetways and pushback)

c. basic fictional naval, road and train traffic (seen at low altitudes)

7. Easy scenery adding

A (relatively) simple application letting simmers make and add scenery objects like buildings, local ladclass, mesh, etc.

This way we all could build up the simworld adding our own areas for the use of all simmers.

Now it requires an ocean of applications, conversions and knowledge. Enough to discourage 99% of us.

Of course there will still be a lot of room for professional addons to sell.

If you ask me.

Best regards,

Rafal

P.S. I'll add more if something jumps nto my head.

Sorry for possible matches with someone else's ideas

Well done, a lot of thing has been thought and said.

To complete this post, I would add :

Damage models (thank to the previous posts) :

-tire blow due to overcharge, hard landing etc

-overspeed effects on the extended flaps, extended gear, even on the whole body of the aircraft.

Realistic ATC :

-realistic speeds and level requirements managed by ATC on the approach stage

-possibilty to know the runways in use for landing at the airport of destination more than 150 nm away ? in order to get a realistic descent and approach using STAR procedures.

[Transam1998 0]

I've been waiting for this news. You folks are our best hope for the future of this hobby.

Things I would pay big to have..

1. Realistic weather and effects with real consequenses for poor decisions (flying into a severe Thunderstorm, real effects of weather on aircraft performance, etc)

Fog banks, Poor visibility areas you can see from a distance, rain under the clouds instead of just wherever..

2. weather effects of handling on the ground, including Snow, ice, rain, etc

3. I am a button pusher, I want systems and procedures, with consequenses for cold shock, botched starts, TOT, etc..

4. crashes and crash damage, as well as airframe damage from stress..

5. I LOVE helicopters.. PLEASE talk to the folks at DODO Sim and get their help with helicopter damage and flight modeling..

6. persistent damage modeling would be awesome!

7. Photoreal scenery for the entire U.S. I'm sure that won't happen, so leave us the ability to do so ourselves easily..

8. An easy to use editor for adding custom objects and airports to our world.

I've purchased your products so I know your commitment to ralism within the limits of the current FS. Hopefully, if you folks create your own, you can break through the barriers..

[BPL 39]

This is my first post in the forum. First, I would like to give my opinion of some ideas already mentioned.

First of all, I get the feeling that this sim may be too integrated into the internet. One should be able to use it offline. For example, I have seen a few people say they want built-in ATC to be eliminated, and just use online ATC. I couldn't disagree more; not everyone flies online. ATC should at least be kept and, if possible, improved. Also the internet is often unreliable. I don't want to start up the sim and find out that I can't use half of it because the internet is down.

I also don't like the idea of using satellite images for ground textures. It may look alright if autogen is kept, but just satellite images when viewed from an angle (as they would be in the sim) don't look good at all. Also, many areas don't have high-quality images.

I like the idea of leaving it very open to developers, but please make the SDK easy to use. I would love to make planes, but you practically have to be an engineer to figure it out. Please include a tutorial for those of us who know NOTHING about software development.

Now some of my ideas:

Tow Plane: In FSX you can fly a glider and be towed aloft by another plane. I'd like to be able to fly the tow plane. Also, I would love to be able to do banner tows.

Small Airports: Please don't neglect the small airports like MS did. It would be nice if the runways, taxiways, buildings, etc. were acurately represented. Also, it would be nice if non-towered radio communications were more acurately represented, adding things such as UNICOM services. Also, radio transmissions at non-towered airports should begin and end with the airport's name, not its identifier as they do in FSX (For example: "Burlington Traffic...Burlington" not "Kilo Bravo Uniform Yankee Traffic...").

[I001 0]

It's nice to see someone take the ball & run with it. Much of what I wish to see in a future sim has been listed somewhere in this thread & then some more plus some things I do not wish to see. Maybe more on these later from me.

I've always had the opinion that MS flight sims never really had aircraft ground handling done very well. I refer to the relationship between ground & tyres. MS sim aircraft don't seem to have a realistic hold on the ground. This is most evident when on taxi or landing (when actually in contact with the ground) in a strong cross wind. I'm sure experienced simmers know what I mean.

Ian

[Erik Brouwer 8]

Realistic seasons effect: In winter, whole of the Netherlands is covered in snow in FS, while we only have snow for about 10 days a year. Maybe MS used data from before the global-warming hype, but every time I fly above NL in winter it feels like flying above Siberia.

[skull 0]

I spend most my time flying the same old routes. I'd like to see a random flight generator. This would allow you to input the distance or time you want the flight to be, and what type of aircraft you will be flying, VFR or IFR plan. Then it generates a realistic flight plan somewhere in the world by searching through a database that users could contribute their favorite flight plans to. Each time you fly, it picks a different departure and arrival airport so you never get tired of flying the same area over and over. It would be nice to be able to print out the plan, including approach charts. Of course, as long as the database is their it would also let you seach and pick any flight you want if you didn't choose the random flight option.

Also, the last 10 feet before touchdown is not realistic in FS, most of the time you only need vertical input. We need some better ground turbulence near touchdown to make the planes simulate the quick input needed with ailerons that are often required in real life. I'd also like to see much better stall and spin dynamics simulated.

[DeatHcrusH 0]

Hi all!

I think that it would be wonderful to be able to walk around on board freely and even exit the plane/heli (e.g.: to walk around the machine).

Best regards!

[Aerosoft Team [Inactive Account] 51558]

Follow up from previous topic.

Any I stress that I like to keep this as condensed as possible, try to write short posts please.

[Aerosoft Team [Inactive Account] 51558]

6. A possibility for 2D cockpit sideviews (in FSX they "forgot" about this important function)

Not sure if I understand this one...

[Aerosoft Team [Inactive Account] 51558]

I was bitten by the initial rush to buy FSX, as many were, and vowed at that time not to rush out and buy another sim without very careful consideration. Which means you have an uphill task persuading people like me to buy it. I'm now comfortable with FSX, with the kit to run it fairly well. Any new sim is going to need to give me something substantively more than what I already have, or I'm simply going to stay with FSX (just as many still are staying with FS9 since they don't see FSX as giving them substantively more). Which means I think that you've got a long, hard and expensive task ahead.

Going to comment on this one as it is important.

I sure see the point you make and it is very valid at this moment. But how about in 7 years time? When have you last loaded FS2002?

The difference between FS2002 and what is possible now is mostly based on hardware. If we start a new game based fully on DX10 (or even DX11) things start to look very different. Just compared how other games increased in potential over the last few years. That's not because programmers are better now but because they got better stuff to work with.

[JPSelter 0]

Not sure if I understand this one...

I think he means the rendered 2D views from the cockpit. But it was FS2004 already that just lost this feature. With FS2002 you could look in all directions of the cockpit with nice rendered views, since FS2004 you only have very ugly low-res textured 3D views when looking to the sides or behind you.

Speaking of cockpit views: 2D cockpits need to be optimized for widescreen monitors.

[K_G 0]

Not sure if you have had a comment like this yet but...

I understand you want to improve on FS(X,9,8...) and make it more realistic. By far FSX was not that realistic, Most of the world had that awful brown ground terrain! And the ATC, I know it came in, was it FS2002 or FS2004 but it was awful.

When you talk about simulating the moisture in the air effecting A/C. Would it be possible to have realism dials in the settings so that for example, if I was going to purchase this product, sometimes I would just want to go up and mess around in my 152 without having to worry about carb ice etc and then there are some times when I want to go up for a realistic flight and wish that carb ice etc was possible in FS, I wouldnt want to worry about that.

[Aerosoft Team [Inactive Account] 51558]

Some items copied from the May posts I like to comment on.

1. Flight session restore feature: If for any reason a flight is terminated without using the exit option (power outage, system crash etc) when restarting you would be presented with the option to continue your previous flight or save it for future use. Saving complex panel states in the background may be an issue but the functionality would be pretty cool!

Nice one, I would always like to be able to save the complete state of the simulator and restart it from exactly the same setting. Added to my master lists!

Photo base or generated base scenery base / mesh density

One user commented "and 2m/pixel is not what you'd call good texture resolution. Double the required space if 1m/pixel were used."

There have been a few comments on this I like to comment on. There is no financial or technical problem to do high definition mesh and medium density photo based scenery. The problem is that using the best mesh that's available (free btw) would be about 16 Gb and that doing the terrain as we do in VFR Germany would take about 2 Tb per season, and only day time. To do it complete would take about 2 x (4 (seasons) x 2 (day/night). Now we can cut some corners but nothing under 10 Tb would be possible. Clearly not possible.

And of course going from 2m to 1 meter per pixel, does not double the size of the files but quadruples the size. That's the big issue with anything like this.

One way to let players see the other players' custom planes would be compiling an "extra" lower quality model along with the .mdl/.flt file so this low-res model could be downloaded quickly and thus preventing this downloaded plane from being flyable for players who haven't bought/got the actual product. The model.cfg file will kinda look like:

normal=P-47D23

interior=P-47D23_interior

Multiplayer=P-47D23_MP

Very good idea, and easy to implement.

I have an extremely important question. Will the new sim be compatible with FSX addon scenery by Aerosoft and other third parties? I have spent large sums of money to greatly improve FSX and I Could not even think about starting over (nor would I want to) with this new sim if it was not compatible with the Many addons I have purchased for FSX.

No, not in the base format. That would be impossible for many reason (copyrights one of them). But if the developer worked with up to date professional tools it should be very easy to recompile the product to the new sim. We would for sure be able to do that with nearly all our scenery and we would most likely offer those products for a limited update price.

The FS interface should be the goal. X-plane is far too complicated for the novice.

I don't agree with this. I find the FSX user interface highly unstructured and something between a cheap game and a professional project. My idea is to keep the interface very straight, very simple. Very much default Windows. If you look at it from a distance and without thinking of a game the choices FILE | EDIT | VIEW | TOOLS seem to make a lot of sense for a simulation. File is where you load ans save stuff, Edit is where you configure your enviroment and aircraft, View is where you define how the setting that are set under Edit look.

[vmc 0]

Hi!

ATC:

Implement a simple ATC handling, with similar capabilities as FSX had. BUT, leave the system open, so that individual developers can substitue in-game ATC with an external programm which allows regional specialities/languages and can be improved over time. In the end you could get an add-on external ATC simulator or automatic ATC programm that can simulate ATC, while it's available as ATC for the flight sim. It could even become an ATC simulator stand-alone, which can interface with Flightsim if wanted. User could choose, which kind of ATC he likes.

If the physics engine for the flight sim and the interface is good enough, it might be even applicable for professional ATC simulator use (e.g. DFS academy), which sometimes lacks some realism in airplane behaviour in the training simulation.

Weather:

Visibility different in different directions. When I'm in the air, on my right is visibility 8 km and on my left it's 50km, I want to see things up to 8 km on my right and 50 km on my left.

2D panels:

Maintain 2D panels. For readability and multi monitor use I like having 2D panels much more than 3D panels.

Thanx, Mozart.

[Tim_A 1]

I'd like to see a persistent aircraft state system for damage, wear and tear etc. By which I mean the system will recognise that a plane is due for its 50 hour check, and require it to be "maintained". Too many hard landings will eventually cause the nose wheel to collapse (not today, or tomorrow, but sometime soon...). The same with flap retractions above Vfe... one day the flap motor burns out. and so on.

[Tim_A 1]

I sure see the point you make and it is very valid at this moment. But how about in 7 years time? When have you last loaded FS2002?

FS2002 was before my time, in simming terms, but there are still people I know who are using it. But my point still stands: There is a perceived substantive improvement between running FS2002 on a 7 year old box and running FSX on a current i7. Assuming that whatever you produce leverages (I hate that word with a vengeance!) current and emerging PC technologies from both a performance and visual perspective, you're already along way down the road towards that "substantive improvement" over FSX.

[sparouty 0]

Hi!

Being myself a naval ops fan (and a Carrier designer for FSX/Acc), I suggest you to think about moving carriers and nav ops feature (Optical Landing system, catapults, arresting gears, etc...).

Aces team did a great job with Acceleration addon, but unfortunately the sdk features are not enough opened for building non Nimitz class carriers (European carriers, early US carrier, etc...). If your'are interested in, me and a couple of "furious" carrier builders of my knowledge should be happy to explain/describe you what should be the perfect "Nav ops" toolbox.

Regards

Sylvain

[LufthansaPilot 4]

I have a question, when are you guys planning to start with the new simulator? When would this simulator be finished (appoximately 1 year or 2 years?)?

[stickshaker 0]

Great initiative Matthijs! I wish you every success and hope I can contribute a little to it.

Some ideas (sorry if they have been mentioned already):

1: The possibility to adapt the scenery according to the date. So, if I set as a date April 1917 and fly above Northern France I’ll see the WWI trenches, but they are not there if I fly in 2009.

2: The possibility to use both table-based (FS) and computational dynamics-based (X-Plane) flight model software through some conversion program.

3: The possibility to set the radar cross section of aircraft, probably as a few simple parameters for the six sides of a cubical object.

4: Stabilized HUDs (where the outside world remains in the correct position relative to symbology like the Flight Path Marker, without developers having to jump through all kinds of hoops to make this possible. HUDs become more and more important for civil aircraft also; the A380 and 350 have them.

5: The possibility of helmet-mounted displays that move with the pilot’s head movements. This is important in modern helicopters.

6: Carrier operations (moving carriers) and aerial refuelling possible.

7: The ability to not only tow other aircraft but also carry them (like an X-15 carried under the wing of a B-52) with the ability to jump into any f the cockpits.

8: The ability to give instructions to AI wingmen.

[Fabrizio Reinero 0]

First, thanks Aerosoft to be such a good company.

The main issue is to increase realism and keep the frame-rates low.

I think we have five group of users for a FS:

General Aviation, Commercial Airplanes, Helicopters, Adventures and Multi-Players simmers.

They have their particular needs and it is difficult to realize a product that satisfy every one.

Very good 2D-panels, optimised 3D-panels, a good weather engine, more realistic runways and ATC would already be an huge achievement for many of them. And if possible at friendly frame-rates. (This was the main problem with FSX)

Greetings

[Hiflyer 129]

Well, I can't add much to the wish list that has not already been added except this:

Don't get too lost in all the hi-techery. Yes there are many hardcore simmers who want every single button to be just like the real thing, but don't forget the millions out there who just want to experiance the sensation of flight, try a few missions, fly over thier houses and some big cities (maybe under a bridge) and otherwise just enjoy the scenery.

I suspect this might be a bit controversial to some of the harder-core, but please dont forget or underestimate the casual flyers!

Thanx.

[Sharrow 107]

Photo base or generated base scenery base / mesh density

One user commented

"and 2m/pixel is not what you'd call good texture resolution. Double the required space if 1m/pixel were used."

There have been a few comments on this I like to comment on. There is no financial or technical problem to do high definition mesh and medium density photo based scenery. The problem is that using the best mesh that's available (free btw) would be about 16 Gb and that doing the terrain as we do in VFR Germany would take about 2 Tb per season, and only day time. To do it complete would take about 2 x (4 (seasons) x 2 (day/night). Now we can cut some corners but nothing under 10 Tb would be possible. Clearly not possible.

And of course going from 2m to 1 meter per pixel, does not double the size of the files but quadruples the size. That's the big issue with anything like this.

This is a big problem for sure!

As stupid as this somehow manages to sound may I ask anyway if we actually need 4 totally separate scenery files for seasons at all? I mean how much different is a simulated autumn from a simulated spring anyway AND are the differences which are currently there enough to actually make a real difference in anyone's life?

What about a Seasons Engine which would use only the ONE set of (photo based) scenery and then control the visual seasonal variations through the use of different vegetation autogen and perhaps different colour filters to make the ground a slightly different shade etc. My feeling is that snow cover should really be seperate from the actual scenery files themselves in that it should be rendered as needed OVER existing scenery by the Seasons Engine (perhaps based on live weather data?) - as opposed to having to load a totally different set of texture files which, at best, only approximate the look of snow covered groud anyway.

Then, day vs night textures: To me it seems like a HUGE waste to have to need 2 totally separate sets of scenery files for night and day. It is the exact same thing really, just one is lit up by the sun and the other is not. Obviously it is not that simplistic but my point should be clear enough.

Should it not be one of the fundamentals of choosing a graphics engine going forward which would allow such MAJOR efficiencies to be brought into the system? Clearly packaging 10+ terrabytes of scenery data with any flight sim is as much a non-starter in 2009 as it will be well beyond 2015.

But imagine having just ONE set of master scenery files (hopefully photo based) and that's it! Sure, you would need MANY more autogen variations, more vegetation variations (trees with green leaves, trees with red/orange leaves, trees with no leaves etc etc) and of course the Seasons Engine itself to mix n' match according to season and render snow as required and adjust atmospheric colour filters etc.

This just seems to me like the perfect opportunity for some "out of the box" thinking on this whole issue.

[harun 0]

He,

I think a basic ATC is sufficient, as known from MSFS. But it should be possible to have the offline Addons ATC expand, for example, insert of SID / STAR route from the user.

More important is the first for a stable and powerful base simulator to create the middle class also on PC systems running unnecessary.

If this is done you can have AI traffic, adventure and standard planes worry.