Developer Q&A Live - March 23rd

[jcomm 6,121]

2 hours ago, scotchegg said:

I finally got around to watching the full Q&A today, and one thing seems particularly interesting about the new thermal modeling; if done with sufficient fidelity and data, could the sim not start generating its own weather rather than waiting for it to be injected? If the data is there for temperature and air movement, does that not mean there's scope for at least convective clouds to be generated?

Bear in mind I don't have the slightest capacity for mathematical science / physics before laughing at the naivety of the question...

Excellent question!

Please see my answer as a totaly unbiased one, from someone who loves flightsimming, MFS, XP, IL2, .... a glider pilot who has been using "soaring weather" and trying to learn it since 4 decades and earns his monthly fee from the MetOffice IRL 🙂

Both XP12 and MFS are bringing changes to this particular form or lift, apparently and from what I heard Sebastien saying in this last Q&A, in more detail and with a more sound approach in MFS.

There was a soaring sim using albedo based on satellite data to calculate convection and that was Silentwings. The authors did a very interesting implementation of soaring weather as far back as 2005 when they started serioulsy working on it until the sim was released imn 2007. Condorsoaring uses a slightly different approach, more script-based.

There was a great add-on for fs9 / fsx - Peter Lürkens' Flight Simulator Tools (t-online.de) - that also used a similar aproach to simulate convection in the form of updrafts. It was the best available for fs9 / fsx by that time.

In XP12 I heard Austin talking about taking cloud bases as a new reference for the calculation of the reach of the rising columns of air, but he apparently forgot about "blue-days" thermals, and other factors. I am willing to see how XP12 is also going to deal with the effect of wind on the development of thermals.

Ideally a model of soaring weather should use tephigrams to help properly determine convection, and weather models convey the base information for that, so, if ASOBO's approach is going to calculate radiance / albedo based on their satellite ( Bing ) data and other weather model parameters, like wind, saturation, temperature lapse rates, etc..., we can only expect it to become promissing for future development of such an approach, including 3pd fine tunning provided access to that data is allowed.

The "Soaring" tittle is going to be the 2nd "Theme", after the Reno Air Races, and I do hope it get's a lot more of interest from users and, who knows, can call the attention of many simmers to the enjoyment / fulfillment of flying gliders !

BTW:

meteoblue weather in new Microsoft Flight Simulator2020 - meteoblue

Edited March 29, 2022 by jcomm
typos and addtional info

[wim123 769]

2 hours ago, scotchegg said:

I finally got around to watching the full Q&A today, and one thing seems particularly interesting about the new thermal modeling; if done with sufficient fidelity and data, could the sim not start generating its own weather rather than waiting for it to be injected? If the data is there for temperature and air movement, does that not mean there's scope for at least convective clouds to be generated?

Bear in mind I don't have the slightest capacity for mathematical science / physics before laughing at the naivety of the question...

no. thats a great question..i would love to see the weather develop by itself , say from some preset.

in fsx with a preset you could set the likelyhood of it changing to another.

that would make more interesting weather with presets.

[SubtotalGuide 133]

9 hours ago, jcomm said:

BTW:

meteoblue weather in new Microsoft Flight Simulator2020 - meteoblue

Look at the date of that post, Asobo has since stopped using meteoblue forecasts and metar is king once more. Say goodbye to those realistic scenes they mentioned. They can't let the weather develop by itself cause then people will complain it doesn't match ma metar!

Would be interesting to hear what that guy who enthusiastically explained how it all worked has to say about the sim weather now. Sadly MS most likely gagged them with an NDA.

Edited March 29, 2022 by SubtotalGuide

[SierraDelta 1,383]

4 minutes ago, SubtotalGuide said:

Asobo has since stopped using meteoblue forecasts

Source?

[SubtotalGuide 133]

12 minutes ago, SierraDelta said:

Source?

Have you used live weather lately? Meteoblue would not deliver such a mess compared to their previous offering. MS will never admit it of course, the only proof is in the pudding.

[threexgreen 3,261]

23 minutes ago, SubtotalGuide said:

Have you used live weather lately? Meteoblue would not deliver such a mess compared to their previous offering. MS will never admit it of course, the only proof is in the pudding.

Apart from general weather forcasts, Meteoblue provide weather data for all sorts of applications, one of which happens to be MSFS. The simulator is taking that data to feed its own weather simulation. What Meteoblue does is provide raw data, that's all. The interpretation and depiction is done by the simulator. Meteoblue doesn't have anything to do with what Asobo/MS choose to do with it beyond that. If they implement METARs and how well that is done is purely Asobo's/MS's business.

Also, they are still using forecast simulation in the weather. This is being mixed with METAR information now to provide more accurate weather at airports and other weather stations, but that in no way implies they have abandoned using forecast calculations. The METARs are basically a guidance for live weather to know where to go with the simulation to keep it closer to the real weather.

After SU5 (I believe) the weather was indeed messy and reeked of a business plan pushing for a premature release, but it's a lot better now and work on in it still ongoing to make it better. They also have acknowledged the weather issues after SU5 multiple times by the way.

[MrBitstFlyer 3,944]

1 hour ago, SubtotalGuide said:

Asobo has since stopped using meteoblue forecasts

Love to know where you got this news from - or did you just invent that?

[SubtotalGuide 133]

@threegreen I like your optimism, however the weather turned sour after SU7. The post-SU5 cloud visuals is one thing, it's the rest that's the big problem now and only getting worse. Popcorn anyone?

 

As Seb said during a previous Q&A, metar is being blended by meteoblue before being sent to Asobo. Why is it then so hard for Asobo to ingest and render all of a sudden?

Contradictions like these make it very difficult to know who to believe when the results continue to speak otherwise.

All we hear is soothing talk about how it's all going in the right direction, just give them time. Time to let the money keep on rolling in.

[jcomm 6,121]

2 hours ago, SubtotalGuide said:

Look at the date of that post, Asobo has since stopped using meteoblue forecasts and metar is king once more. Say goodbye to those realistic scenes they mentioned. They can't let the weather develop by itself cause then people will complain it doesn't match ma metar!

Would be interesting to hear what that guy who enthusiastically explained how it all worked has to say about the sim weather now. Sadly MS most likely gagged them with an NDA.

I believe you're wrong. 

What changed is the way "aerodrome weather" is fed into the weather system in MFS. AFAIK their initial approach was based only on GRIB / GRIB2 data from Metoblue Weather Forecast Models which caused some discrepancy due to lack of consistency with the hourly observations reported in METAR and SPECI, because for sure the forecast model can't have the resolution required in aviaton meteorology for TAF and observation in METAR.

Since SU7, if I'm not wrong, they now merge actual METAR data with the "aloft" model weather and IMO they're doing it quite satisfactorily. 

Is it Perfect ?  Surely it isn't, but it's shapping up nicely, specially after what I heard Sebastien mention in the last Q&A regarding the modelling of various local  / Meso-Gama effects.

 

Edited March 29, 2022 by jcomm

[RXP 2,338]

12 hours ago, MarcG said:

But yes the main issue is the information is not being passed over to the team good enough as you say, the Devs appear to be in their own bubble so remaining oblivious to many issues. The regression in VR on recent updates it's testament to their ignorance, even though Jorg & Martial state they play in VR...I find that hard to believe.

I just catched up the latest QA and speaking of a bubble:

Quote

Could there be a sharpening option in the menu if you want to un-sharpen or sharpen your graphics? 1080p monitors, they say, have an over-sharpened look. Timestamp

Seb - We have a sharpening postprocessing filter linked with TAA [temporal anti-aliasing]. I don’t think there’s an option right now. It’s definitely feasible to add a sharpening slider. We should add that to the list and see when we can do it. I think it was in the options file at some point. (Maybe or maybe not; maybe I’m wrong.) I’ve definitely seen it in our code options. We can add that to our list of things to do: Sharpening or un-sharpening. People may want to have it both ways.

It is not as if it wasn't asked or commented about before though (see below)...

 

57 minutes ago, SubtotalGuide said:

Contradictions like these make it very difficult to know who to believe when the results continue to speak otherwise.

All we hear is soothing talk about how it's all going in the right direction, just give them time. Time to let the money keep on rolling in.

Indeed, such contradictions are perplexing sometimes. For example about the "sharpening postprocessing", here is what I wrote since January 2021...
[BUG/FEATURE] Provide Color Space, Tone Mapping, CAS Shader Strength and post-processing effects controls in VR - Virtual Reality (VR) / VR Wishlist - Microsoft Flight Simulator Forums
 

More specifically about the CAS Shader Sharpen setting:
https://forums.flightsimulator.com/t/bug-feature-provide-color-space-tone-mapping-cas-shader-strength-and-post-processing-effects-controls-in-vr/345703#cas-shader-source-code-14

But what can we do about this when the moderators are even locking me out of my own topics and posts....

 

PS: it is even feedback-logged but it just seem like something new to discuss about in the QA...

Edited March 29, 2022 by RXP

[VFXSimmer 1,314]

2 hours ago, SubtotalGuide said:

@threegreen I like your optimism, however the weather turned sour after SU7. The post-SU5 cloud visuals is one thing, it's the rest that's the big problem now and only getting worse. Popcorn anyone?

 

As Seb said during a previous Q&A, metar is being blended by meteoblue before being sent to Asobo. Why is it then so hard for Asobo to ingest and render all of a sudden?

Contradictions like these make it very difficult to know who to believe when the results continue to speak otherwise.

All we hear is soothing talk about how it's all going in the right direction, just give them time. Time to let the money keep on rolling in.

We can surely continue to be optimistic so long as this forum is frequented by super-sleuths like the ones we've seen in evidenced in this discussion.

Without thoroughly researched, thoughtful, fact-based analysis like this Asobo might have been able to pull of their dastardly scheme and fool us all.  I mean the motive for them to be deceptive is clear... it... well escapes me at the moment, but I'm sure I'll think of it soon...

Fight on noble knight!

[SubtotalGuide 133]

12 hours ago, RXP said:

But what can we do about this when the moderators are even locking me out of my own topics and posts....

You can appeal a ban by emailing support@flightsimulator.zendesk.com, but I wouldn't be surprised if the community managers are the same people behind zendesk. The castle is heavily fortified.

Seriously, people want to help make the sim the best it can be. Why must there always be so much politics, deception and other BS involved? I started out having the utmost respect towards MS/Asobo for bringing out this amazing sim. Along the way their actions have ruined it all.

[Pastaiolo 1,690]

Excited about Italy finally coming, hopefully with some custom autogen and pine trees all over coastal areas. Since i started using flight simulators in 1998 this is the first time in which Italy actually looks like Italy. What a time to be simming 🙂

[jcomm 6,121]

1 hour ago, Pastaiolo said:

Excited about Italy finally coming, hopefully with some custom autogen and pine trees all over coastal areas. Since i started using flight simulators in 1998 this is the first time in which Italy actually looks like Italy. What a time to be simming 🙂

Yeah !!!!!  Finally able to glimpse the "Pizzeria" next to Pasta's home !!!