An Airbus in your living room? An Interview with JeeHell’s FMGS Developer

Originally posted on simflight.DE, written and interview conducted by Peter Rosendahl.

Anyone who has built an Airbus A320 home cockpit is probably using the “glass cockpit” software “A320 FMGS”. This software suite connects the home cockpit hardware of different manufacturers with FSX or PREPAR3D. The developer Jean-Luc gave us an interview.

However before we start the interview we would like to give readers who are not yet deeply involved in the topic a little background on the software.

The download of the freeware (for non-commercial users) can be found in the support forum, which is also the largest playground for the JeeHell community. There are really good manuals that you should really read (think RTFM is a good idea anyway), because you are dealing with complex software.

JeeHell is software that is usually installed on several PCs in parallel to the simulator. The setup is very different in the community, but usually a kind of master is installed on the simulation PC, which establishes the connection to the simulator. Several PCs also because, on the one hand, we have to connect a lot of control elements (a side stick, tiller, rudder, etc. per side) and for (excluding one possible DCDU) 5 displays with images (captain / first officer side, MFD, SD and one display each for the MCDU’s).

Displays by JeeHell are highlighted in red

As if that wasn’t all more than enough, JeeHell Ware also has its own piece of software, since it acts as a kind of audio mixer. This provides (in short) space for an intercom and connects the sound sources and outputs for headsets and stick microphones – which of course can also be connected to VATSIM, IVAO and Co. for online flying.

So that we can also see a plane in the simulator, JeeHell is delivering a special A320 model based on the Project Airbus – but without a virtual cockpit, as this is already in the living room and the front view of the scenery is displayed by projectors or screens / flat screens.

But now on with the interview:

Jean-Luc you are very know for your A320 Cockpit Software “JeeHell Ware”. Could you please introduce yourself and tell us a little more about your flight simulation background?

As many know already, I am French, living on the French Riviera in Nice. I work in Nice Airport as an Air Traffic Controller (rated from ground to approach).

I fell in the flight simulation hobby quite a few years ago with FS5.1 as a kid. I’ve owned all MSFS versions since then. My main center of interest quickly became the SATCO network (before VATSIM and IVAO), mainly as an ATC back then. This is how I got interested in my current job! When real life met the FS life, I slowed down for a while as pretending to work got boring… So I spent more time doing real life flying on light aeroplanes.

I came back to simming with the level-D 767 which brought the professional environment I am used to work with, and shortly thereafter started working on my A320 software. Back then I wanted to make my own home cockpit from the ground up, including electronics, but I had to start first with the software. In the end, I only ever made a prototype FCU and MCDU, so the home cockpit is sort of a failed project.

As a developer I can’t really say I have any background, I only ever released the A320 software, some generic gauges (just at the same time as the A320 started), and a very long time ago some application to create sector files for the SATCO ATC software (ProController, which was used long after SATCO went down).

Is flight simulation your main job? Are you an Airbus Pilot in real live or do you work in aviation business?

Flight simulation is still mainly a hobby for me but I am very close to aviation in real life, which drives my need for realism.

I work as an Air Traffic Controller as I said, and I have had my commercial pilot licence for some years now. I also hold a valid A320 type rating and want to switch careers at some point, though the COVID19 episode sort of closed most opportunities for some time. I will keep training on my software until I can achieve my goal of flying the Airbus for real, this has been a great help in getting the type rating in the first place, and it is still very helpful to practice the SOPs between FFS check rides.

Why does a home cockpit builder need your software? Isn’t it possible out of the box with the FSLabs or Aerosoft Airbuses?

Tough question because I do not own these add-ons and have never tested them. I am in fact an almost full vanilla P3D user! I guess the interfacing for almost all available hardware vendors makes it a plus for me, as well as the possibility to run the software on multiple computers and remove load on the P3D PC. Also being freeware for private use makes the jump to making a home cockpit easier.

Regarding realism, I suppose that is something all developers now try to achieve so it is a less of point, but I still receive a lot of good comments on the FBW that would still be above the competition. I don’t really look at the competition in fact, because I feel if the software ias as close to the real thing as possible, then it cannot really go wrong.

What was the motivation of developing “JeeHell Ware” as a freeware project for private use? Will the freeware “ever” become payware?

At the time I started, I did not have the need for additional income, the software was much more crude than it is now and selling it was out of question since I did not want to spend too much time for support on a non working software. Also, I discovered aviation through the MSFS franchise and the really great add-ons. I remember the short 360 by Dai Griffith (If memory is good) this was a huge inspiration. There never was any proper freeware A320, and even payware were not so good. Only the Wilco one which probably was a very nice simulation at the time, though I was still studying and did not want to spend money on FS.

There were also Project Magenta (Boeing mostly) and AST (which just started) back then for home cockpits, and since I wanted to make most of the sim myself, I started programming a few gauges to check the feasibility of my own software. There in fact was ONE freeware alternative, called VASFMC, which reproduced an Airbus like suite, with a rather good 2D FMGS. However the development was not going type specific enough for me, and lacked systems integration, the ECAM was just not alive and there was no roadmap in the direction I wanted.

So I started properly working on the A320, quickly had a 2D FMGS and a poor AP/FBW, and after a while basic overhead systems and ECAM pages. And I felt it was a good opportunity to give something back to the FS community which somehow made me plunge in the aviation industry.

As for the future, I intend to keep a freeware version of the software. There may be a payware option for more advanced features but this is still not clear how it will happen.

Why Airbus and not Boeing? When have you started working on it?

I think the first lines of code were written in 2007, first release around 2009. Back then only PFD/ND/FCU/MCDU were working, and the AP/FBW was like a roller coaster.

I chose the Airbus because I am French mostly… And I did some A320 sim sessions as part of my ATC course back at ATC school. Also the Boeing FS scene was already saturated when I started working on the software.

The great thing about the Airbus series is that it is all computer based systems. It makes it interesting from a developer point of view. You have to fit in a couple consumer PCs the capacity of 20+ dedicated computer systems.

Developing such software really looks like a mess of work. Who are the people behind A320 FMGS?

Mostly myself all alone. And it is a huge project indeed, though I like the challenge and except for a few features which were (or are still) painful to code, I quite enjoy doing the research and coding myself.

All code is from me, except a few libraries I use (for openGL, or even aerodynamics/navigation computation with ADACALC from Hervé Sors many simmers know).

I had some help tuning the flight model and PID controllers for the AP/FBW because it is awfully long to do that and would have taken so much time I would still be doing just that. So most people can thank Tomasz Tobola for his help there.

A few other users sometimes helped on some areas like OVHD graphics (thanks Jacques Zahar), user guide (thanks Michel Picque and Claude Kieffer) or just testing new features, bug hunting etc…(many here, but Chris Paulus from Cockpit concept has been around almost since the beginning).

How do you ensure that the flight characteristic of your simulation software is close to a real aircraft? Is there any big difference to software used in commercial full-flight airline simulators?

Most FFS’s use OEM hardware and software for most parts of the simulation. The difference with the real aircraft is basically the flight model, but the FMGS or EIS are exactly the same as on the real aircraft. The systems like hydraulic are simulated then feed fake sensors connected to real aircraft computers. An Airbus is in this regard easier to reproduce than a Boeing.

In the hobbyist FS scene, however, everything must be “re-engineered”. Aircraft like the A320 or B737 are widespread, documentation is easily accessible. Also as a type rated pilot, I have access to flight manuals. That means I also have hands-on experience on the aircraft. Over the years, I met many people in the industry who have engineering knowledge or access to the right people who can answer my questions.

With the commercial versions of the software, many real A320 pilots have tested the software, given their feedback and that was usually very productive as well as comforting in that I took the good direction most of the time. The accuracy is now so great the software is used on EASA FNPT2/MCC certified devices, in partnership with Simnest (www.simnest.com), an Hungarian company who is responsible for final integration and all the paperwork associated with certification (the so called QTG and validation data). Their devices are really beautiful, I spent a couple hours in one of those just a few days before my last A320 type rating check ride, and it felt just like the big sim.

What kind of development tools do you use?

Nothing out of the ordinary. Most applications are coded in Delphi, it is a fabulous solution to quickly develop applications. Some are in C++ when no alternative is possible (for instance P3D PDK modules).

What were the biggest challenges to implement your software in FSX and PREPAR3D so far? Are there any restrictions you are regularly annoyed about?

When I started, I had to learn basically everything on FSUIPC then later on SimConnect implementation. But past these small hurdles there are no major issues working with FS/P3D. There is a big developer community and the inner workings of the FS platform are well known and documented.

There are a few limitations to FSX/P3D, but on commercial jets you can achieve really high fidelity with P3D. I do not see any real added value to a completely overhauled aerodynamics engine such as advertised in MSFS2020 or X-plane. P3D with minor additions could be as accurate as an FFS simulator. Weather in general could be really improved in P3D (weather depictions by default and weather radar). Airport lighting as well could be enhanced with controls, even if there always are work-arounds.

Some “hacks” I coded are quite innovative and required long development to validate (these are real-time aerodynamic characteristics modifications, and also performance computations from the airfile for VDEV, or even runway lighting control in the pro versions).

Other complex developments were AP/FBW and FMGS though not really tied to a P3D issue.

The biggest issue with P3D is the performance. It is really hard to have stable high FPS in P3D with good weather depictions in all conditions (storms, night etc) while using my software. This is really a problem on devices with projectors trying to achieve 180° FoV or more, and with detailed airport/scenery addons. There are tradeoffs that need to be made… My only wish would be FS2020 visual engines with P3D add-on architecture!

How is the partnership and working together with Hardware-Vendors like Skalarki, CPFlight, e.g.?

I do not have agreements with any vendors. Most of the time, they do not make the interface themselves, which is a shame in my view because I have to use their documentation which is not always so great and is not as low level as I feel it should/could be. Also, I had to make interfaces for hardware I do not have in my hands, that proved to be tricky sometimes though most vendors are generous enough to send at least a small module for testing and that is quite appreciated.

Most of the time however working with them is easy and effective enough, and when possible they implement changes in their firmware when that is required. I only had one really bad experience with a vendor, I will not comment here but those who followed my software development long enough will probably know.

What is the benefit of becoming a commercial customer of JeeHell Ware? Can you tell us something about the costs?

The commercial versions of my software are meant to allow use in a business environment, such as flight simulation leisure centers or even real life training schools. This was not my plan from the beginning and I sort of got there “under pressure” from commercial users. I did not want the freeware version to be used in any business as I support freeware for private persons, but I do not like the concept of using other people’s work to make money without any compensation.

The pricing in this commercial range is prohibitive for hobbyist use, and I will not detail that here. Most customers do not see that a simulator is 50% of hardware and 50% of software. A working device needs both, on the same level. And it is harder to put a price on software, as it is not tangible, so they are always surprised. Microsoft can sell windows at sub-100€ price only because of the volume of sales, this does not exist in the home cockpit market. And the freeware version of my soft sometimes makes people think it is not worth more than a mobile application.

The commercial version brings the right to use in a commercial environment, along with the Instructor Operating Station (IOS) I made for professional training (failures, simulation control and all controls expected from an IOS), and a few other features specially made for training.

In your support forum in the bug tracker you are very active with the freeware community. How important are the customers to you, since they do not bring you any financial benefit?

I do like bug free software, so even if it does not bring financial gains, it allows me to improve the software and have many different user scenarios/environments/configurations I cannot test myself. Some bugs are so “stealthy” I would still be looking for them. In the end it makes the commercial version more robust, and hobbyists also enjoy the stability at some point. The commercial and freeware versions really have the same codebase, and for normal operations there are no big visible differences. Hobbyists almost have a piece of an FNPT2 device at home, at 0 € cost, I think it is still a great bargain even with a few bugs 🙂

I usually do not talk much about the commercial version, because most of the time this brings up the question of how I consider the freeware community. Even if of course the commercial customers have some sort of priority, I spend a lot of time doing support on the freeware user base. I do not always answer all emails or posts on the forums, though I read all messages and try to include fixes as soon as practicable for me.

What are the major development milestones that you are facing within the next 12 month? What about MSFS or support of X-Plane? Any updates to the instructor station?

FS2020 is still a big question mark. I do not have access to the SDK yet, so I have no idea what will be possible or not. As far as I know basic compatibility is possible as the aircraft model should be compatible, and FSX’s SimConnect interface still works as well. Though I have no idea how to do some advanced things like the icing simulation, weather radar and even navigation data extraction (runway/ILS alignment currently done using BGL files). The rumors with web assembly are not so nice for many people, it will require a lot of time for many developers to adapt to the new framework. I think the continuity throughout the P3D versions was really appreciated.

So for some, FS2020 might be a step backwards at first, at least for those who prefer systems realism over graphics depiction.

X-plane could technically be possible, though the aerodynamics are quite different, and I would need to restart there from scratch to make a decent flight model. I am a total newbie here and I don’t really want to face the steep learning curve. I made an interface through x-plane plugins many years ago so that part is covered (it does not even need XPUIPC), but the flight model is a complex thing.

I still have plans for an hobbyist IOS version, though I still don’t really know how to make that happen without stepping on the commercial market. I do not want unlicensed versions being used in training organizations, as I already have to fight that.

In any case, FS2020 will probably take up most of my free time from the 18th of August onward…