December 27, 20205 yr Clearly I know nothing about this stuff, but I would think naively that a simple if/then statement in the code would avoid these gross artefacts (which I do not understand were not noticed by the team when testing the last update), something like "If crazy elevation jump in the mesh over small area, just flatten it out" Just flew out of Torino and the four giant spikes right north of the airport are not pretty... and they show up everywhere I've been flying since the update (NA and Western Europe)
December 27, 20205 yr 2 hours ago, AnkH said: Typical case for MSFS and a nice example how we are dependant on the quality of the streamed data. Basically this is the prize you pay for world wide coverage like we have in MSFS... Hope they fix this soon, I am already tired about those mandatory updates rendering the sim worse than before... Literally every update tho, I can finally play again since the Tokyo update so there’s that.
December 27, 20205 yr 25 minutes ago, Turpentine said: Just flew out of Torino and the four giant spikes right north of the airport are not pretty... and they show up everywhere I've been flying since the update (NA and Western Europe) Torino was corrected in the mod, I get the complaints guys (I really do), but if we are going to complain, can we please complain about places that haven't been corrected 🙂 https://flightsim.to/file/5198/elevation-fixes-around-the-world Edited December 27, 20205 yr by Alpine Scenery AMD 5800x | Nvidia 3080 (12gb) | 64gb ram
December 27, 20205 yr 10 minutes ago, Turpentine said: Clearly I know nothing about this stuff, but I would think naively that a simple if/then statement in the code would avoid these gross artefacts (which I do not understand were not noticed by the team when testing the last update), something like "If crazy elevation jump in the mesh over small area, just flatten it out" Just flew out of Torino and the four giant spikes right north of the airport are not pretty... and they show up everywhere I've been flying since the update (NA and Western Europe) Since the water itself is spiking, there should be variables in there that can be conditionally evaluated to prevent it from spiking, but it's most likely someone wrote some bad spaghetti code trying to mix client/server interaction for corrections and the code isn't easy to update because you nail down one board and the other side flops up (hence fix one elevation, another one pops up), so no simple if/then without changing the spaghetti code most likely. Other than doing it conditionally, Its a trig problem, it's not that hard, but you have to take the slope and calculate it against a probability of the nearest prominence of the slope next to it. There are probably at least 6 different ways to solve it mathematically though I'd guess, I certainly don't know them all off the top of my head. Edited December 27, 20205 yr by Alpine Scenery AMD 5800x | Nvidia 3080 (12gb) | 64gb ram
December 27, 20205 yr 1 hour ago, Alpine Scenery said: Just fly in the US or some airport that has been corrected or doesn't have an issue, there are still plenty of those. I just do short flights, it really doesn't bother me too much, but it's a good time to explore new areas if too many of your flights are in places with mesh errors. I haven't seen any serious mesh errors in the US that remain uncorrected after the mod. There are still crawling water and coastal issues, but you can ignore those pretty easily. Most of Mexico, Central America, and South America looked clean too, though admittedly I didn't check all of it. There are plenty of places to fly, and we have mods, so don't let it get to you is my advice... Maybe pick up some Greece airports and pretend you are playing Assassin's Creed 🙂 I'm off to do a flight out of Mexico City, haven't really explored that area too much. I agree! There are still some beautiful places to explore: MSFS
December 27, 20205 yr I still believe this is not a mesh issue, but rather a software issue. Both X-Plane and Prepar3D have really nice default meshes, it just doesn't make sense that Asobo don't have access to a similar high-quality mesh. It makes a lot more sense that it's rather a processing issue on the software side, not a raw mesh data issue. This also means manually changing elevation data would just be hiding the issue, instead of eliminating the underlying cause. It's still better than nothing until Asobo comes up with an actual fix though. Edited December 27, 20205 yr by BiologicalNanobot PC specs: i5-12400F, RTX 3070 Ti and 32 GB of RAM. Simulators I'm using: X-Plane 12, Microsoft Flight Simulator (2020) and FlightGear.
December 27, 20205 yr There's another bump just north or Cordoba city (Spain). I'm not sure, but I think there are more spikes and bumps near cities. Am I all right?
December 27, 20205 yr 1 hour ago, BiologicalNanobot said: I still believe this is not a mesh issue, but rather a software issue. Both X-Plane and Prepar3D have really nice default meshes, it just doesn't make sense that Asobo don't have access to a similar high-quality mesh. It's a math issue, when we say "mesh" we are talking about the end result. It should have been corrected, but someone mixed spaghetti math with spaghetti code is what it is. I know because I've been in that exact situation before, and even though the problem is simple to understand, it's a lot of work to solve because you are coding around someone elses spaghetti dinner, especially if you are worried about breaking something. It's an error in the discreteness and/or rounding of the quantizer to translate the water's coordinate system into an area of pixels to the game engine's round earth coordinate model within that tiny confined area of those water cones. There's a mismatch or a data error somewhere in their data which is either null'd out or just too big of a number, and they have spaghetti code trying to fix it. My guess is there some bug deep in the code that is supposed to look for null values around water or improper coordinate rounding, and it's replacing the null with a big number, which is why we see the spikes. That's why it happens only in certain areas, it's a mathematical anomaly from multiple roundings that triggers it, or it's just corrupt data that is being provided. Edited December 27, 20205 yr by Alpine Scenery AMD 5800x | Nvidia 3080 (12gb) | 64gb ram
December 28, 20205 yr Saw 3 spikes just flying from San Jose KSJC to Sacramento CA. These were never here before the update. Jim
December 28, 20205 yr I can confirm, I'm seeing spikes, taking off from Torino Caselle (LIMF) airport on rwy 36. Spikes starts to appear as soon as my wheels aare off the ground. I asked to a friend of mine to perform the same take off, and he reported the same behaviour (different pc with different graphic card.).
December 29, 20205 yr The latest update of mappamundi's Scenery Glitch Fixes on flightsim.to is a knock-out. They claim that it should fix these spike issues globally, and from the places I've spot checked that seems to bet the case. It's a somewhat big download, but worth checking out.
December 29, 20205 yr 1 hour ago, kaosfere said: The latest update of mappamundi's Scenery Glitch Fixes on flightsim.to is a knock-out. You are not kidding! Vancouver Island.... volcanoes are gone 🙂 Totally amazing! Wonder what Asobo does next?? Bert
December 29, 20205 yr 5 hours ago, kaosfere said: The latest update of mappamundi's Scenery Glitch Fixes on flightsim.to is a knock-out. They claim that it should fix these spike issues globally, and from the places I've spot checked that seems to bet the case. It's a somewhat big download, but worth checking out. Yup, it's good but it's also depressing that the CGL method worked because it means Asobo is meandering about and not productively fixing stuff that isn't hard to fix. I saw someone mention a CGL method before, but I never tried it thinking it would be too much trouble, or that if it were "that easy" Asobo would have fixed it. For the US, there are still many areas that are problematic after that "global mod", so I left the US one active. I redirected all international users to try that one first instead of mine, and I discontinued it. I have some experience using those GIS tools, but I guess I should have read more into the CGL method, not sure where he got the documentation on how to fix it, possibly what he did was use the mesh guide that one guy recently posted on a Github and he just exported the differences (only the conflicts). I always thought in terms of replacing the entire mesh is just too big of a file, but I never thought of just exporting the differences, which is what I am assuming he did. Edited December 29, 20205 yr by Alpine Scenery AMD 5800x | Nvidia 3080 (12gb) | 64gb ram
December 29, 20205 yr On 12/27/2020 at 7:22 PM, Alpine Scenery said: Torino was corrected in the mod, I get the complaints guys (I really do), but if we are going to complain, can we please complain about places that haven't been corrected 🙂 https://flightsim.to/file/5198/elevation-fixes-around-the-world No, let's complain about everything because it's not up to the community to fix bugs and it's not up to the users to download every ***** small area from a 3rd party website before starting the game. So much for the "USA update". 7800X3D | 2x32 GB DDR5-6000 CL32 | RTX 5080 | Alienware OLED 34" | 1 Gbps fiber
December 29, 20205 yr 6 hours ago, kaosfere said: The latest update of mappamundi's Scenery Glitch Fixes on flightsim.to is a knock-out. They claim that it should fix these spike issues globally, and from the places I've spot checked that seems to bet the case. It's a somewhat big download, but worth checking out. The first comment says there are still issues in USA after installing this mod. We want a global fix from Asobo or the possibility to revert to previous builds disabling multiplayer. This downward spiral in the quality of the game update after update is not acceptable for an AAA title. 7800X3D | 2x32 GB DDR5-6000 CL32 | RTX 5080 | Alienware OLED 34" | 1 Gbps fiber
Archived
This topic is now archived and is closed to further replies.