MSFS2020 photogrammetry makes building looks melted

[ha5mvo 254]

1 hour ago, lwt1971 said:

Sorry that's a blanket statement that is very exaggerated. So you're saying that the PMDG 737, Leonardo MD80, Fenix A320, Bae 146 don't "even remotely" fly by the numbers? The PMDG 737 "flies like a pig"? Ok then 🙂
 

Let's be more specific, the Fenix did all they could with the *engine* physics/modelling and now are seeking to do an external engine model as already articulated well by @Aamir on another thread (https://www.avsim.com/forums/topic/620353-fenix-a320-growing-on-me/?do=findComment&comment=4793628). With the current solution using the default engine physics they're able to get to within 6% of performance numbers. The Fenix still uses the MSFS core physics and aerodynamics engine for its flight model and dynamics, that is not external, nor is it even possible/practical to run the core flight dynamics entirely outside the sim. So those are various loaded and exaggerated statements (i.e. the Fenix "shakes about like a nutshell in white waters", riiiight). 
 

 

You sound like a decent fella. Now do you find it reasonable to believe that FENIX will spend months, if not more, of its time to develop  external dynamics engine for the sake of refining the model from a 6 to say..3 percent error? Can you honestly say that this sounds reasonable.

Fenix, with admirable honesty and transparency said out loud that they have given up on the sims default engine. Majestic did the same with p3d's turboprop dynamics when they wanted a serious simulation.

You don't need to take my word for it. Experiment yourself! set a crosswind of mighty 3 knots gusting to 5 at the opposite direction and see how it flies!

I stumbled upon this video:

[P3D] FSLabs A320 Pitch and Power Basics - YouTube

 

The numbers he's quoting are accurate. See if you can recreate it with the fenix, and see if you can land with unreliable airspeed indication. These are bread and butter basics!

I'm not trying to attack Fenix. Actually they are the most decent in acknowledging the errors but there's something fundamentally wrong with the sims flight dynamics which - and its their words, not mine - is beyond the scope of just polishing up the airplanes dynamics.

[lwt1971 7,417]

35 minutes ago, ha5mvo said:

You sound like a decent fella. Now do you find it reasonable to believe that FENIX will spend months, if not more, of its time to develop  external dynamics engine for the sake of refining the model from a 6 to say..3 percent error? Can you honestly say that this sounds reasonable.

Fenix, with admirable honesty and transparency said out loud that they have given up on the sims default engine. Majestic did the same with p3d's turboprop dynamics when they wanted a serious simulation.

You don't need to take my word for it. Experiment yourself! set a crosswind of mighty 3 knots gusting to 5 at the opposite direction and see how it flies!

The numbers he's quoting are accurate. See if you can recreate it with the fenix, and see if you can land with unreliable airspeed indication. These are bread and butter basics!

I'm not trying to attack Fenix. Actually they are the most decent in acknowledging the errors but there's something fundamentally wrong with the sims flight dynamics which - and its their words, not mine - is beyond the scope of just polishing up the airplanes dynamics.

I was trying to point out that the area where Fenix is now going to write something entirely external is for the aircraft engines, that's it.. they have put out a current implementation using MSFS's default aircraft engine modelling, but it's not good enough for the level of simulation/depth Fenix is seeking so they're going external for that because that is something an aircraft add-on *can* go external of the sim with. And I'm not saying that they are doing this *only* for the 6% fuel flow discepency as I naturally would not know that now would I, just going by what's been relayed by @Aamir here and on their discord... For example, this is what Aamir said on their announcements discord channel https://discord.com/channels/736572376967282769/736710909887643679:

"There's two big ones I'd like to address: engines and DUs. Engines are a fun one, MSFS gives us a single point of tuning for fuel-flow - either it puts out some Saturn V flow figures in climb at low altitude, or it sips fuel in the cruise - there's just no winning. We've fretted over it as best we can, and we're seeing about a 6% margin to the real numbers, balancing all aspects of flight and then rounding up on the slightly more efficient side so you land with a few gallons more than you need, not a few less...! Ultimately we think the best way forward is a fully custom engine model. We actually already started poking in this route with start-up/shut-down, EGT, spool times etc - but to go much further and do it properly requires some SERIOUS work, and might not prove fruitful for some time. Given we're so close to the numbers already, holding back for an undefined period of time (at least a few months) doesn't seem worth it. Quite frankly, to the average eye nothing will appear amiss, but you guys are far from average aren't you? Once post-launch clean-up calms down, we'll get cracking on a fully 'Fenix' engine solution"

The above has nothing to do the with the core MSFS aerodynamics/physics capabilities which the Fenix is still built on top of for everything else in its flight model and dynamics (i.e. everything outside aircraft engine modelling).

You can understand how I struggle to take you seriously when you make broad and comical statements like how the PMDG 737 "flies like a pig", or that there is not a "single tubeliner that even flies remotely close to the numbers" (including 3rd party aircrafts) in MSFS, or this about the Fenix: "Try to run that very same FENIX, the current pinnacle of MSFS development , in a variable "gale" wind of 3-5 knots and it will shake about like a nutshell in white waters" ... care to tell us how you think the PMDG 737 flies in that wondrous flying-on-rails environment that is P3D? :)

Edited June 11, 2022 by lwt1971

[ha5mvo 254]

You're right that p3d doesn't simulate, or at least convey, air movement very well. Xplane does it much better, MSFS exaggerates it...

How does the fact p3d "flies on rails" as you put it, improve Microsofts dynamics?

You don't have to take my broad and comical statements at face value - test it for yourself! I did....

I have put out values and numbers in previous posts but that obviously didn't matter to those who are entrenched in the position that its all within a couple of percent margin. Heck, not even the developers statement will make them budge.... 

[lwt1971 7,417]

23 minutes ago, ha5mvo said:

How does the fact p3d "flies on rails" as you put it, improve Microsofts dynamics?

I should've clarified, given that you think the PMDG 737 flies like a pig in MSFS, I was curious as to how you think it flies in P3D.

 

23 minutes ago, ha5mvo said:

I have put out values and numbers in previous posts but that obviously didn't matter to those who are entrenched in the position that its all within a couple of percent margin. Heck, not even the developers statement will make them budge.... 

Again, I'm not saying that the Fenix engine modelling issues are *only* about the 6% fuel flow discrepancy or other similarly small percentage margins.. I quoted Fenix's complete statement on this topic above. Where I take issue with your take is that it's somehow due to MSFS physics in general. The Fenix's flight dynamics and handling are very good in MSFS, and similarly for the PMDG 737, Leonardo MD80, Bae 146 and other well implemented aircrafts. Not perfect by any means, but far from the serious issues you seem to think they have. We'll have to agree to disagree I guess.

Edited June 11, 2022 by lwt1971

[The Moose 1,177]

I could have sworn this was once a thread about photogrammetry....

[toby23 138]

This video talks about the state of photogrammetry in London. In short, it looks like the blitz just happened and no amount of moaning (or tweaking) will fix it..

It is what it is, some cities have great photogrammetry data available, some don't. It's not (normally) about the speed of your internet connection. 

Here's an idea (from a London helicopter pilot) - don't fly over London, it hammers your performance, doesn't look great up close, and there are much prettier places to fly - try EGQI and fly an emergency mission in Neofly with the freeware R44 and AirlandFS around the Cairngorms National Park for a change. Imagine you're there, walking over the hills, head in the clouds, breathing in all that fresh air and enjoying the marvellous views.. YMMV