A William Reynolds Interview
Mathijs Kok was kind enough to give our readers an insight into the soon to be released Airbus X Extended.
You will recall their first edition Airbus X was previously reviewed by avsim.
How would this new release compare? What will it offer? Let's find out!
Thanks for your time again, Mathijs....
Q: Why the Airbus A320?
MK: As so often, somebody mentions it and then it starts a life of its own. There were no good FSX versions around (we now understand better why that is btw, lol) and perhaps most important, via AirBerlin we had ample access to them at Paderborn Lippstadt airport where our headquarters is located. It’s really a 3 minutes’ walk for some of our developers between their desk and the real thing.
Q: When did development of Airbus X version 1 and then version 2 start? How long did it take?
MK: Not easy to answer as we schedule our resources internally. Sometimes a modeler completes something that is then on a disk for a year before it is used. Aerosoft is not a small company that works on one or two projects, we run 20+ projects at any given time. I am also not really willing to share how much time we spend on this, sometime it is better to let your competitors make the same mistakes as we made. Let’s just say it is about 50% more than planned. No matter what you plan. It’s the kind of project that kills small companies.
Q: What is the development process for a project like this? What were the phases?
MK: We do that a bit different from others as we invite a group of ‘testers’ very early in the process. That links up well with the way we like to show progress on our forums btw. We want the input from users from the very start. However the process is rather straight forward. There are four development tracks, modeling/texturing, flight modeling, systems and last all the smaller stuff (sound, documentation etc). Ideally they all need to finish at the same time but that never happens. Other teams then start to test but in our system the testers test the files we make minutes after they are done. It is really an interactive process and not cycle based as some IT development. We basically release between 6 and 10 new beta versions a day and distribute it via a SVN server to the testers. 7 days a week. To handle that you need very cool testers. But we got those. We got real Airbus pilots who read the beta forums as they fly the real aircraft (long live WIFI on aircraft). There was one aircraft (with passengers) that intercepted an ILS last week at and abnormal angle because something needed to be checked. The real Bus accepted it without problems btw. Our version was not able to stabilize on ILS in time. More work to be done.
Q: What resources did you have/did you seek?
MK: As explained we had full access to the aircraft itself. Even the first time we went for photographs we were left alone in the powered up aircraft (imagine the temptation) and could do as we wanted. That takes care of the modeling and that was very much the core of the product. Finding information about the systems is not hard these days. We got a complete set of manuals (at least 5.000 pages) and the internet is full of documents. Halfway through development we asked two current Airbus pilots to assist us. Resources were absolutely not a problem. If anything we had too much information.
Q: From a developer’s point of view, what is the challenge in coding/simulating the famed Airbus Fly By Wire component?
MK: It’s the high integration of all systems that created the problem. Of course it is hard if the control stick is not connected to the control surfaces and if the throttle is not connected to the engine, but after you find a way to disconnect that in FSX (not easy at all) you can just insert your own code. It is very time consuming and needs months to fine tune but it’s not hard.
The hard part comes when you insert the code into the project. The Airbus is really built as one integrated system where every system communicates with others. Getting those connections correct in all possible conditions means digging at a mountain of bug reports for months on end. And often it is just very hard to understand the logic. From the users point of view it is clear what happens but when you program it you need to know what caused the effect. It’s an old joke but Airbus pilots really are known to say: ‘why is it doing that?’.
Q: Were you satisfied with the original Airbus-X version?
MK: Yes, I was happy with how it ended up and with the sales.
We did make one rather big mistake though and that was not sticking to the original plan. We made it a bit too complex in some parts and that caused confusion with customers. It was never intended as a very high complexity product. It was really intended to be something between the default aircraft of FSX and the far more complex PMDG level projects.
Visually highly developed but intended at users who do not want to spend many hours with complex manuals. Being able to do a complete flight with good realism after just one hour. And we got that, but customers clear expected something more and that was most likely because many systems were very high end. The complete project should have been a bit simpler and we should have matched the modeling to the systems better. It actually was our first big aircraft project (normally external development teams like PMDG make those for us) and there was a learning curve.
Q: If we just concentrate on the new version, how much changed from the original plan you had for V2?? What things got added along the development, or removed during development? What was in your "wish list" that didnt quite make it?
MK: We wanted more complex systems and that needs more CPU cycles but we did not want to lose any framerates compared to the previous version. We sell complex scenery add-ons and we want aircraft to be limited in how much FPS they use. So the modeling was worked on a lot to reduce the load on the computer without looking worst. There was two years between the two projects and we had some new ideas, more experience and better tools. The new version looks better without a doubt but has 20% less polygons. It now has something like 150.000 for the VC and 100.000 for the external.
The systems were almost completely redone. Parts of the ECAM and the PDF are roughly the same but the MCDU, AP, FMGS etc are all fully changed. We now do not use many default functions of FSX. We also wanted the navigation systems to be much more accurate and fully support the new navigation database we will include (it’s backwards compatible with Navigraph but covers many more airport). So with SIDs and STARs, etc. And all of that can be programmed on you tablet computer as the Airbus comes with a webserver included. So the MCDU can be in your iPad.
There is just one thing I personally wanted rather desperately and that we had to drop. I wanted an actual animated and seriously good looking person in the right seat. I never understood why people can complain about the simplest item but are not upset to land an airliner with an empty seat in the cockpit. That’s perhaps the most unrealistic part of any high complex aircraft at this moment. We wasted two weeks on that but the human mind is so tuned to other humans that it is so hard to make it look at the same realism level as the rest of the cockpit. It just never worked. We need a new simulator for that, one that’s faster and has better animation options.
Q: Have the manuals/documentation changed much?
MK: We had to change some to match the new systems but they are really only updates. Still a few weeks of work, certainly we do documentation in 4 or 5 languages.
Q: Any changes to the Flight Dynamics model?
MK: A few tweaks to match the new FBW systems, but the engines were fully redone.
Q: What about the exteriors and sound?
MK: We got one new model (the A320 with sharklets) and there are some minor changes but the externals and the sounds are not changed a lot. The plugin system we have that allows the aircraft to be extended by third parties, does have new sound options though. From audio checklists to a complete audio environment with radio chatter, cabin crews etc.
Q: Has shared cockpit been a consideration at any stage?
MK: Not really. It a nice line on the feature list but it really is only used for a very small group of people. Not worth the investment and limitations.
Q: Do you have plans for a Service Pack in the future? If so, what will it have?
MK: I am sure we’ll have to do some bug fixing and perhaps to change some things, but we do not plan any major changes after release.
Q: You stated the original Airbus-X targeted a specifc type of simmer...what is the target audience for the new version?
MK: The simmer who likes to simulate the job of the pilot in the left seat. We do not simulate simulators so we did not spend a lot of attention to failures, they are so rare that if we would do them realistically the average pilot would have to fly hundreds of thousands of hours before something happens. Why make an APU fire simulation when it has never happened on a real aircraft? We think that we got all the things a pilot does on a flight are simulated realistically and that’s all we wanted.
Q: There are other A-320 simulations in the market, some released quite recently, and some to come...what does Aerosoft's version offer the Sim public that is the differentiator?
MK: I think we got a lot of functions that they lack (like a complete RAAS and digital flight data recorder) that really add to the realism. I am also sure our modeling is better than any other A320 out there and that our aircraft is lighter on FPS and on your wallet. After that, we’ll let’s see what the reviews say.
Q: Any confirmed expansions for it? Will it be tested in Prepar3d? Will there be an X-Plane version?
MK: We will for sure release the P3D version. Probably the same day as the FSX version. No X-Plane version planned for now. We do not plan any extension, at Aerosoft we are not big fans of that.
Q: Finally...yourself personally...how satisfied are you with it...did it hit the mark? is there more than can be done?
MK: There is a lot more that can be done. But as said we set ourselves limitations and have different ideas on these projects than for example PMDG. They aim to simulate the aircraft to perfection, we aim to simulate the job of the captain. That creates different products at a different price levels. So far I am satisfied with how it is going, we are turning to final at this moment and though the whole team is very tired, we feel it will work out. If it hit the mark? That’s something customers will have to say. As always some will be happy and some will expect a different kind of product, it’s the amount of each that matters!
Mathijs, once again, thank you very much for your time, we look forward to a good rendition of the A320 and the continued work from Aerosoft to expand our virtual skies!.