In this scenario, I do not want moving fog banks that I can see from any angle. I want the entire planet to be covered in one solid fog bank, with density and thickness that I specify, and no variability. This is how sim training takes place.
If from a practical standpoint it’s accomplished by a circle of fog with a 10 mile radius that is centered around my airplane and moves with it, that’s fine. I don’t care how it’s done, I’m just looking for the ability that any actual simulator has. It’s kind of sad that old FSX could accomplish this better than MSFS.
Now I’m a fan of MSFS’s live weather system. I greatly enjoy watching it in action when my intent is to simulate an entire flight.
But when using the sim for maneuvers or procedures proficiency, a user needs to be able to define weather with the expectation that the sim will depict exactly those conditions.
Since a simulator is intended to be used in a variety of ways, both for line-oriented scenarios and “slew to or load up in position and go” scenarios, the ability to manually control all aspects of weather is pretty important.
You’re right, FSX (and even FS98!) accomplished using distance fog by not having a volumetric atmosphere and having no other options at the time. Even Doom and Duke Nukem accomplished having distance fog. Now that we’re in the modern era with much better hardware, volumetric fog in a volumetric atmosphere is the new standard, which is far superior in every single way, especially for training. Kinda like how having AI-generated 3D cities with satellite imagery is far superior to grey polygons on the ground representing cities. Now we just need to be able to fine-tune the atmosphere grid\voxels so that we can have precise control over cloud density, temperature etc of each grid cell. That would allow us to set up scenarios with low visibility where we need it to be. In my opinion, that would be much better than going back to the stone age.
For now the best you can get is to create a thick cloud layer and start it as far below the airport as you can, then add max precipitation. It’s not idea, but the results are pretty good. This is for a runway at about 250’MSL
If the sum total of a training goal is to fly an approach using a particular panel configuration to 200ft and 1/2sm, how is the new system any better than what active sky produced in FSX? The approach lighting system still fades into view in exactly the same way.
The only difference is, in FSX when I set this weather I knew that it would exist exactly as I set it.
In MSFS, not only can I not set weather parameters with the same precision, they actually vary over time. I can set up overcast skies with cloud bases on the ground and max precip and get some pretty low vis… but if I shoot 5 approaches, the outcome of each will be different. Some I might miss because I never see anything, and others I might get the rabbit at 500ft.
This is not a sim training environment. Variability is fun when using live weather, but if I’m setting the weather, I want it to be what I set. Not with thin and thick spots.
But I don’t understand why you perceive a necessary difference here. Why can’t the sim set a global weather system, exactly as I specify, and it still be volumetric?
If you’re training for real life flight, then it stands to reason that a more realistic approach is superior to a less realistic approach that was only used because hardware couldn’t handle a more realistic approach at the time. I’m quite happy that hardware and software optimizations has evolved to the point where we don’t have to settle for those old workarounds anymore.
Because it’s volumetric. You can see the fog bank in the distance, fly into it, fly out of it and see it disappearing behind you. You can watch it dynamically move in from over the ocean or form in basins in the distance, so weather phenomenon like the Southern California marine layer or inland smog\fog bank (see screenshots) is pretty accurately depicted with MSFS, but not in FSX or any of the older sims.
Volumetric fog is far more precise than fog that extends to infinity with a circle of visibility that follows you around until it’s time to transition to the next global visibility that covers the entire world. I agree though that we need more control over it.
The atmosphere in MSFS is made up of a 3D grid that is rendered almost as far as you can see (within a reasonable altitude). Each cell\cube of that 3D grid within your line of sight plays a role in how the light is scattered and how the atmosphere is rendered overall. Every individual pixel of the sky is dynamically governed by the conditions of each atmosphere grid cell\cube within your line of sight, instead of the old system which was just a texture on a sphere. Since the atmosphere in older sims was not volumetric, your plane was the reference point and there was a single “visibility” value that extended out from your plane. Asobo correctly realized that that’s not how real life works and implemented a more realistic volumetric atmosphere system. To precisely control a volumetric atmosphere, you need to be able to set the conditions of the individual cells or use some sort of reference points because the conditions don’t follow you around. They’re actually tied to the world now, not your plane.
But I actively want none of this to happen. I want to create a weather scenario where I never am not in clouds and low vis, until I break out at the minimums I specify. I don’t want variability in this scenario.
There’s no reason fog can’t be both volumetric and 100% coverage, right?
Not at all. It’s entirely dependent on WHAT the training goal is. There are times when, yes, the goal is to simulate a flight in its entirety, with all the pop-up threats and variability of a dynamic environment. MSFS is good at this.
But there are many times when the training goal is to accomplish a specific task to specific parameters. As an example, at work on maneuvers val day of recurrent sim training in the level D boxes, we will fly three cat III approaches. One will be an autoland in the minimum 400ft RVR. One will be a handflown Aiii approach using the HUD, to exactly a 50ft DA in 600ft RVR. One will be a missed because the weather is below mins.
There will never be any variability in this. These are the scenarios the FAA has approved, because they expect pilots to demonstrate their ability to fly these approaches all the way to legal mins.
Whether I’m wanting to use the sim to maintain proficiency at a 6 pack scan since I don’t often fly non-airline IFR, or I want to use the PMDG 73 to maintain my Aiii HUD scan… I want to fly these approaches exactly to mins. Breaking out early, or flying in and out of a fog bank just so I can see it come and go… these things break the training goal of flying a particular approach full IMC to mins.
This is really a simple standard in the entire simulator industry, and an area that needs a large amount of improvement in MSFS. That would not be a step backwards.
100% coverage can already be done using a preset. That’s not what we’ve been talking about. What we’ve been talking about is being able to specify a precise ring of visibility that follows your plane through fog that extends to infinity, in other words, downgrade the current engine to technology from the 80s and 90s. I have not heard a single benefit to this vs keeping the current system and just being able to edit the grid that makes up the atmosphere to set up a low visibility scenario around the airport you’re training at, or being able to modify METAR reports and do it that way. You could set up any scenario you choose with far more control vs what you guys are requesting here…
You’re missing the point. I highly doubt the FAA commands that a volumetric atmosphere is not allowed and the simulator must use a ‘distance fog’ rendering method that creates a ring of visibility that follows your plane in an unrealistic way that doesn’t represent how it actually works in reality. Any sort of scenario that can be done with a non-volumetric atmosphere can be done far better and more precisely with a volumetric atmosphere… I don’t get how anyone can think otherwise…
It was a simple standard across most 3D engines, but not anymore. Just like a fixed function pipeline was the standard, realtime raytracing was unheard of, deferred rendering was not a thing, VR was only in movies (and virtual boy), and everyone would have laughed at the thought of AI-based upscaling or AI-based scenery generation. Things are constantly advancing in the world of software development and computer science in general. It’s up to the old simulators to keep up, not for new simulators to hold themselves back.
A ring of visibility that follows your plane would be a step backwards by the vary definition of the saying… It’s moving from modern technology to old technology for no other reason than “but that’s how it used to be!”.
Practicing low visibility operations requires accurate control of ceiling and RVR, once more I don’t care how it achieves this in the background but flying an approach to airport X it must be possible to set weather exactly to CAT-I, II, III minima. Its the most basic functionality of any flight simulator.
In the context of a volumetric atmosphere simulation, the solution is having precise control over each individual cell of the atmosphere so you can create a low visibility scenario in the area you’re doing your training. It’s a very important distinction and completely different than “give us a visibility slider with a ring of visibility that follows your plane”. I doubt Asobo will ever respond to that because it’s nonsensical in the context of the current volumetric fog\atmosphere system that they worked very hard on. They probably just shake their head every time they see it. However, if instead people kept asking for precise control over the weather grid, maybe in the form of a map (like the manual cache) or via modifying METAR reports, they may actually listen and these training scenarios could be set up. There’s no need to go back to the stone age of distance fog, which would look ridiculous next to all the advanced rendering techniques the sim uses.
I’m not sure you properly understand what “volumetric” means.
Yes, the FAA does “command” that a “distance fog” rendering method is used, in the sense that they’re dictating precisely what distance you’ll be able to see when below the ceiling. Until minimums, you get a visibility of exactly zero. It does not vary from zero until you reach your decision altitude. At that point, it fades into whatever visibility you’ve specified will exist below the ceiling. With no variability. Because this is how a training scenario of an approach to mins has to work.
Have you flown many instrument approaches in low vis in a real airplane? You just… don’t see anything, until you break out. I mean, nothing. All your windows are painted white, opaque. If you find a digital rendition of that boring because it doesn’t leverage all the newest rendering technologies, well, I don’t know what to tell you. But that IS realistic.
Now, like Nijntje91 says, I don’t care how this is accomplished. If Asobo wants to let me specify the weather for a 50nm ring around a given airport, in which I can specify every parameter and they’ll all be rendered precisely and with zero variability… that’s fine. If these conditions are just rendered globally the way they are in level D sims, that’s fine. If my weather is a bubble that follows my plane, that’s fine too. I’ll experience all three of those scenarios the same way, so to me it makes sense to just take the easiest / lowest coding workload approach.
But we need to be able to precisely specify a visibility distance that then needs to render precisely, with 100% coverage and with no variability. This is a given requirement for any simulator training.
Its very simple, for something to be considered a training device or simulator, precise control of all weather features is essential. That includes, visibility / RVR, ceiling, wind direction and speed, QNH and temperature. Without that its useless as a training device for instrument flying. For low visibility training you need to be able to precisely control RVR and ceiling in order to break out of the clouds and have the runway / approach lights in sight at exactly the right moment (or not). I don’t care what it does in the background to achieve this, aerosol density, temperature / dewpoint split, relative humidity, compensation for runway / approach lights intensity, day versus night, etc. I have never used any simulator (approved synthetic training device or not) which does not have that capability. Apparently it is able to do just that in live weather, so it must be able to do it using weather preset as well? Instead of getting visibility info from a METAR it gets it from a slider instead? And instead of creating fog in and around one airport, the fog layer should be worldwide, just as you will have scattered clouds world wide when using that preset, or a 20 kt wind for example.
I think you’re projecting, considering you don’t seem to understand why a single global “visibility” setting with no fixed reference point is nonsensical in the context of a 3D volumetric atmosphere simulation, despite me explaining it in simple terms multiple times. I even included images, which apparently didn’t help. I also went as far as to try to reword your suggestion to make it applicable to the current volumetric engine (being able to modify individual cells of the atmosphere) and even that didn’t help.
That is not what defines distance fog and can be achieved far better by having precise control over the atmosphere grid. Distance fog is simply setting the transparency of objects based on the distance to the viewer, which makes it look like a fog effect. The color of the fog is based on the color used to clear the buffer. This was a simple method used because there were no other options at the time and it looked better than objects popping in out of nowhere, not because it was ideal or had anything to do with realism or “simulation”. Thankfully we’ve moved far past that.
You described a single scenario where distance fog somewhat looks okay and conveniently didn’t mention the long list of cons that come with a one-dimensional distance fog system with a ring of visibility that follows the viewer. That long list of cons is why it was ditched a long time ago in favor of the far more realistic volumetric fog\atmosphere, which is thankfully what we have now.
Those are radically different solutions that require radically different discussions and radically different changes to the simulation, which is my entire point. If they do address this (which I hope), I hope they do it the right way, and not the “old standard way” that other simulators use, which was the least-realistic way of doing it.
Which could be done if we could precisely control the atmosphere grid and precisely set the weather around an airport and position your plane wherever it needs to be to satisfy those requirements, which is the point I keep trying to make. Everything you want can be done without having to downgrade to the old antiquated system that made it possible to have a single “visibility” slider.
Great that its possible! How would you set visibility / RVR without a visibility slider? Voice recognition? Its not all that simple I would imagine as RVR is the distance a light source (e.g. runway or approach light) can be seen by an observer. For the same visibility, RVR is lower during daytime compared to night time. RVR is roughly visibility x 2 during night time while it is x 1.5 during daytime with high intensity approach and runway lights.
Being able to set the individual cells of the 3D volumetric atmosphere is how you achieve that in the most precise and realistic way possible.
Live weather doesn’t have a single “visibility” value. Live weather receives a complete 3D grid of “virtual atmosphere” from meteoblue surrounding your plane and then blends in data from nearby METARs into that virtual grid. Virtual rays of light travel through that virtual grid of atmosphere, it gets scattered\obstructed by virtual water particles in each cell based on the atmospheric conditions of the cell, and then that’s used to render each pixel of the atmosphere. It can be foggy\cloudy in a single atmosphere cell, it can be foggy in a group of cells, it can be foggy at the base of the mountain and not the peaks, or whatever the case may be. It’s completely dynamic, based on pseudo-real life physics, and is why the lighting effects are so convincing and we can have effects like dynamic rainbows. Everything is completely localized and 3D.That’s how visibility is currently handled. How could you possibly have a single “visibility” setting as part of a preset? With a volumetric atmosphere, you would create these precise conditions by modifying the individual cells of the atmosphere, or some system to automate the process, like being able to modify METAR reports and it automatically does the hard work of filling in the atmosphere cells. That’s what we actually need.
The same way the live weather does it, and similar to how it works in real life. The parameters of the individual cells of the virtual atmosphere have to be set to the required values to give you low visibility at the location where you’re training at, which can’t be done with a global visibility slider in a weather preset. There HAS to be a “position” component in the context of a volumetric atmosphere simulation.
Still don’t get it, I want to perform a CAT-II approach, with a RVR of 300 m which corresponds with seeing the approach lights at 100 ft decision height. Without a visibility slider, how do I set 300 m RVR? I’m a dumb pilot so as simple as possible.
You could have a visibility slider in theory, but it has to be tied to a reference point in the world. It can’ t just be a “global visibility” in a preset unless you want a visibility circle to follow your plane through infinite fog. Like I mentioned above, one simple approach would be being able to modify the METAR reports and set your own values, which I would imagine wouldn’t require much extra work. Or be able to specify coordinates, a radius, some parameters, and the sim automatically fills in the atmosphere based on the parameters and radius you set. Another option is a map that allows you to set each cell of the atmosphere, creating more advanced scenarios that could be useful for bush flights or missions.
Why not? This is how all other simulators - to include all certified training devices and simulators - work.
Now, why do you think they work that way?
Because those of us that use them to train pilots in the real world need them to work that way.
So why wouldn’t you want MSFS to be able to do this as well?
You seem to be using “volumetric” to mean that the weather cannot or should not be constant throughout the entire world. This isn’t really the literal definition of the word, but if I’m understanding correctly I can see how you’re using it. But we keep explaining to you that in many training scenarios, we NEED the weather to be constant throughout the world.
Now, how would YOU accomplish this? I want to tell the sim I need to remain solid IMC with not a single glimpse of ground, blue sky, cloud structure or lighting change, until seeing the first approach light at the exact moment I specify. I want to be able to fly 100 approaches in a row to one airport without any variability at all, then divert to another airport 50nm away and see the exact same conditions.
How would you have the sim accomplish this? And given that it’s a requirement of real training vs entertainment flight simming, why wouldn’t you want the sim to be able to accomplish this?
Again, we aren’t talking about changing how live weather works. We’re talking about allowing the user to actually control the weather when necessary - something we’ve always been able to do in previous sims.
Nice. I guess my biggest annoyance is that MSFS is capable of rendering accurate vis, and beautifully… it’s just that the vis value must come from a metar.
How hard would it be to give the user a way to specify it? Even as a variable in a preset, vs the (small) work of adding a slider to the UI. I mean, how hard can it be?
