AFAIK, what the author did was just translating a French article published in
a paper (one from Bordeaux IIRC), but in the original there was no mention at
all about the rendering engine.
Since my role here is to provide support to developers and not really to
debate with them, I’ll try to keep it short:
- “the issue I see here is the lack of commitment from Asobo”: this is curious since Microsoft and Asobo announced from the beginning that MSFS was a 10+ years project. The regular (and free) Sim & World Updates also seem to contradict your statement - unless you meant something else?
- “They are discussing it for realism, more in-depth simulation”: again, this is a curious statement in a topic that discusses virtual cameras. Granted, they would be nice to have for some planes but they are hardly related to “in-depth simulation”.
- “I understand there are contractual obligations, revenue generating features”: which “revenue generating features” are you referring to? I am sure you can make your point without speculating about contractual obligations.
- “when will Asobo take things seriously or provide support for more in-depth simulation?”: although you may not agree (or understand), we have been doing so from the start.
- “Asobo said they fixed it and provided API yet it’s no where near what its needed”: not sure if the “fixed” part of this statement refers to the weather system or the API. I think we said quite early that we would not allow custom weather systems. We know that the Weather Radar Map API is not enough for some people - but please note that it is for others - and we may work on it again in the future.
- “these issues start from the fact that you guys look at things from graphical and perception point of view, not realism or actual simulation aspect of it”: this is probably the most wrong part of your post - there are more people in the studio working on planes (art, flight model, systems) than on rendering.
- “Asobo does not accept feature requests from 3rd party developers that are related to improvement of simulation aspect of it”: we regularly review suggestions/ideas and use them to both drive our roadmap and make adjustments to existing features. However as you know making a request does not mean it will be satisfied (and there could be many different reasons for this).
Best regards, Eric / Asobo
As RXP said, it looks like the author of the article you’re referring to
translated bits of this article:
<https://objectifaquitaine.latribune.fr/business/l-actu-de-la-french-tech-
bordeaux/2020-09-17/avec-microsoft-flight-simulator-asobo-studio-s-envole-
pour-au-moins-dix-ans-856752.html> But for some reason he then decided to add
fake information (maybe because the full article is behind a paywall?). A
quick search on the web returns several discussions about these made up
statements. Best regards, Eric / Asobo
Hi Eric @EPellissier, Do you think it would be possible to do something among
those lines you described (toned down details, forded LoDS, low res, etc.) to
create rear view mirrors in MSFS? Currently for many airplanes were mirrors
are required such as the F-18 for example we are unable to configure this
properly, we end up creating a metallic object surface to have some
reflections, but since reflections are very close for that type of material
most of the time the mirror effect looks wrong and lacks the desired results.
What you described would be a much better experience vs what we currently
achieve and offer to end users, and perhaps the performance implications could
be avoided by using the tricks you described in combination with new rendering
techniques such as DLSS etc. Let me know your thoughts, if this is something
that you believe could be acceptable then please let me know to submit this as
a new idea so you guys can discuss it further. Thanks for taking the time to
answer our questions and concerns on this subject. Best Regards, Raul
The rear view mirror example is interesting: in most (if not all) racing games, the rear view mirror shows a (sometimes very) degraded view of what the main view could display (potential optimizations are a limited draw distance, toned down FXs, lower LODs...).
Sure, mirror won’t ever require to have the same amount of detail of a normal
view, it’s overkill and not really required and for applications like
“cameras” that project over a texture, it’s perfectly acceptable to have a
lower resolution, with no effects and perhaps limited shading. However, RTT
camera views are just one of the possible applications for RTT, another big
one would be programmatically draw on a texture surface without having to
necessarily project an existing view. Something like using one of the methods
available in Gauges, like NanoVG, but over an arbitrary texture in the scene
that is not necessarily a Gauge in an airplane Panel, with the feature
possibly accessible through a stand-alone WASM module that register to some
kind of callback whenever such texture is defined ( like the $-prefixed
textures available since FSX ) In that case, it will be entirely the add-on
developer responsibility of being careful how complex the drawing is, exactly
like a Gauge developer is already supposed to do now.
Hello Eric, I appreciate your response. This is not a debate, I would consider
more as communicating the frustrations and POV. I’d like to reply to your
bullet points in order:
- What I meant from “lack of commitment” is the not Asobo and Microsoft’s support for the next 10 years. What I meant by that is the lack of commitment in improving the sim to be used for more professional aspect or in a way that is closer to real life. For instance in the MSFS forums, there is a thread for “MSFS Pro” with thousand+ post. If you look at the forum there are people who even want to pay for a more simulation and realism focused version of a MSFS (obviously there would not be a version but this is not the point here). Yes what you guys are doing are great but it’s lacking substantially in areas such as realism and in-depth simulation. This is what I mean. You guys are committed yes, but only to simpler, arcade-ish version of MSFS. By version, I mean usage. That’s how you want this sim to be used, not to be used by people who want to use products that simulate real life planes to the exact bit and behaviour.
- It is related to in-depth simulation. Since we are discussing about this in virtual cameras topic, I’ll give an example that is related both to the topic and what we are discussing. B777-300ER. One of the most popular and best-selling add-on in flight simulation market. PMDG did it 8 years ago to a platform that was released 16 years ago. Yet, when they develop B777-300ER to MSFS, they will not be able to simulate this as it’s not supported by the platform. As you can see, the lack of support results in less in-depth simulation of 3rd party addons. And this is only ONE example out of many. I just wanted to give an example that is related to this subject
- Yes. It’s business. Revenue generating features or improvements will always be prioritized over other development areas. This is just business 101.
- I’m sorry but no. Only improvement regarding more in-depth simulation was the CFD. Nothing else. Can you point out what Asobo did within these 2 years that increased the in-depth simulation of MSFS and 3rd party addons?
- “Fixed” part comes from what you guys wrote in the Wishlist. If a “wish” is “fixed” then that means Asobo considers the work is done and completed. This is actually what I’m referring to when I say there is a larger issue of how you guys perceive things. “I think we said quite early that we would not allow custom weather systems” this is fundamentally and logically proving my point. You do not want to develop the in-depth simulation and realism aspect of the game. I’m not going to call it simulator because for you guys this is a game. Continued, “We know that the Weather Radar Map API is not enough for some people - but please note that it is for others” it is enough for small GA planes as they share similar systems and align with MSFS’s agenda of keeping things and simpler. However this was not why developers and users requests an API in the first place. It was so that the complex planes can use this data and simulate a real weather radar, for instance Collins WXR-2100.
- Again, either I was not able to get my point right or you didn’t understand it correctly. In fact, the example you give is again art and I’m not talking about default planes.
- I understand this part completely. Not every feature request will make it to end product however when developers provide answers to questions like; “Why we need it”, “Why it would be useful”, “What we want to achieve”, “How MSFS could benefit from this” etc. Shutting it down immediately is not what we want to see.
Thank you, Alp
I'm sorry but no. Only improvement regarding more in-depth simulation was the CFD. Nothing else. Can you point out what Asobo did within these 2 years that increased the in-depth simulation of MSFS and 3rd party addons?
Sure, happy to. Here’s just a small sampling of the various simulation depth
features that were added since the initial release:
- Ability to alter the center of lift pressure
- Numerous additions to the flaps configurations (lift pressure center adjustments)
- Ability to directly adjust ground effect by mach number
- Addition of simvars and events for master caution/warning tracking and states
- Additional ITT cooling time constant and tuning cfg values
- 4D high-speed segment-based propeller simulation
- Crosswind behavior tuning params
- Tire sidewall/rotation resistance tuning params
- Expansion of possible turboprop beta behavior and bindings
- Numerous prop drag adjustments
- Prop feathering and windmilling simulation improvements
- Ability to get RNAV approach minima data on approach navdata
- Addition of RF leg support
- Ability to get runway slope, independent elevation, and threshold lengths from the navdata
- Ability to adjust ground and air spoiler drag values independently
- Expansion of fuel flow calculation configuration options
- As you mention, the CFD simulation (massive)
- Overhaul of the altimetry simulation to support correct non-ISA pressure and temperature errors (first sim to do so)
- Addition of autopilot override simvars to allow third parties and advanced airplanes more direct autopilot control for building complex AP simulations
- Independent upward and downward elevator trim angle maximum support
- APIs for requesting nearest METARs by lat/lon or METAR by facility ICAO
- Updates to the electrical system to better support non-uniform battery types
- Addition of the ability to have up to 4 nav and com radios
- Simvars for getting the state of piston engine superchargers
- Simvars for adjusting the scaling of both the CDI and GSIs independently
- Configuration settings for developers to control the icing simulation on their aircraft
- A continuous stream of Live Traffic accuracy, density, and datasource bug fix updates
- Updating the icing simulation based on addition source research material and whitepapers
- Added the ability for pitch autopilot mode to capture altitude if desired
- Additional overhaul of the flap system to support multi systems with different number of levels and max angles
- Ability to completely customize the weather radar simulation graphics colors based on precip density, allowing for correct colors and gain simulation
- Addition of a branch new visual effects system which greatly expands the capabilities of the previous visual effects system
- Ability to configure individually the lift coefficients of the elevator, ailerons, and rudder
- Ability to adjust the fuselage lateral Cx to allow for more or less drag while in a slip
- Overhaul and update of the sim magvar underlying data and system
- Removal of the limit of the number of possible terrain/weather maps per aircraft
- Improved tools to tune prop drag on constant speed & turbo propellers
- Removed the old FSX era wrong ITT simulation when prop is feathered or reverse
- Improvement of water rudder capabilities and simulation
- Greatly increased the granularity of the ambient density data and simulation
- Ability to add negative wing camber to aircraft
- Added TACAN simulation, navdata, and associated simvars
- Added the ability to allow fuel system triggers to be based on CG values
- Ability to scale the amount of engine wash on the wings to tune engine wash contribution
- Addition of more autobrake cfg file params to allow for modeling all the available autobrake arming modes and behaviors
- New cfg file params for adjusting and tailoring stall behavior, plus debug screen
- Addition of cfg params for adjusting the rigidity of the fuselage and connecting response (softbody simulation)
- Ability to adjust the spool up N2 rate while starting turbine engines
These are just the items that I personally remember. 
Hi Raul, We would have to test this in-house but I am not convinced that
simply degrading the secondary view(s) would be enough to overcome potential
performance issues. A few things that such an optimization would not change:
- The additional GPU RAM required by these views for their rendering
- The culling calculations which need to be redone for each view (which could potentially trigger the loading of other objects which were not in memory yet)
- Additional CPU/GPU work
Anyway, I suppose we would have to add some kind of “quality” setting if we
were to provide an external camera API. Best regards, Eric / Asobo
Hi @virtuali, Actually the NanoVG API provided through WASM is already doing
some kind of RTT - we already discussed internally ways of extending its use
for scenery textures but the topic is not very high on our list of priorities.
Maybe this is something we should re-evaluate. Best regards, Eric / Asobo
Hi Eric, @EPellissier Thanks for the explanation, I get all you guys are
saying. Just trying to find ways to make better airplanes mirrors, because as
it stand we can’t really make them work nicely. They are key for formation
flying among other things. Just trying to find ways to improve the end user
experience overall. Thanks again for your time to answer. Best Regards, Raul
Thanks for very detailed answer mattnischan. Everyone here wants this sim to
be the best it can be. I want to add one thing that, still, most of the
improvements are for general aviation aircraft, still not lot of
improvement/development regarding airliners. Which is something that we want
Asobo to focus on. Even Jorg or Sebastian (don’t remember who exactly) said in
one of the Dev Q&A;'s that they did not even anticipated this much demand and
interest to airliners and they wanted to focus on GA. Kind of explains their
point of view about these these topics 
This, I think, is where I get confused, since I would say better than 80% of
the items in that list also apply to airliners, and some of them were
absolutely brought especially at the request of airliner developers. Like the
much more complex flap system, the autobrake enhancements, fuselage Cx
changes, jet engine param additions, autobrake stuff, splitting the spoilers,
the altimetry changes (important for temp comp situations), n2 spool rates,
fuselage rigidity, visual effects system (prominent airliner devs wanted
better landing gear particle effects, contrails, etc), the addition of 2 more
com and nav radios (GA planes having but two, nearly always), icing sim
changes (again at the behest of airliner devs reporting that it affected large
aircraft too much or too strongly), lots of the turboprop stuff (requested by
regional airliner devs), center of pressure work (far more important for large
aircraft), live traffic work (almost entirely focused on commercial airliner
traffic). It’s obvious to me that the team has prioritized very strongly
features which are core to the airliner flight experience. Perhaps we are just
disagreeing about whether external cameras, a feature available on very few
airframes and used for an extremely small portion of one’s airliner flight
session, constitutes an important core feature.
Alright, how about the WXR then It’s not just about one simple feature and
why it’s not in the sim, it’s about how they see things and develop it.
Biggest example is the WXR fiasco. They said SDK is coming, and we got a band-
aid, unfinished and useless SDK. Yes they did great improvements in the sim
yet still no where close to airliner experience you can have in P3D or XP…
This thread is about virtual cameras and has already diverged quite
significantly from the original topic - there’s already another topic about
WXR so I don’t think it’s worth discussing here.
https://devsupport.flightsimulator.com/t/4735 Best regards, Eric / Asobo
Indeed it did and yes I agree we need to keep things separate. I just wanted
to give the WXR situation as an example. Anyways, I hope that gets attraction
as well, since there is still no response there. Thanks for all the work Eric.