@buzzbomb said in Another F16 entry coming "someday"?:

Yes, it IS a competition. Every time I decide I want to fire up a flight sim, I have to choose: BMS or DCS, or X-plane or MSFS, or Prepar3d, or various legacy sims. Which will give me the experience I want at the moment?

Every person who has more than one sim installed makes this choice, too.

The moment you think you aren’t competing, you’re not. And that makes your product obsolete.

I realize that adding these things to BMS will be no trivial effort. But it’s going to decide how relevant and competitive BMS remains in the foreseeable future.

The flyable world will surely be the biggest effort, particularly when you start adding in 3D constructs. (Buildings, vehicles, etc) Hand crafted areas of interest which will be focus points for future battle campaigns will certainly require extra effort. But even that is getting easier to implement all the time. It was beyond consideration a few years ago. It’s working now in MSFS, X-Plane, and Prepar3d .

BMS isn’t competing because we don’t need to. We aren’t driven by anyone’s money nor will so we don’t need anyone’s support in order to keep do what we do

And speaking for myself (i.e off-record as BMS dev) - You need to understand - Before I’m coding for you or for him, I’m coding for myself. I LIKE flying Falcon/BMS, even alone, even SP, even forever (And consider I didn’t actually flew for the last 4-5 years because of my current development journey and it’s incompetence with existing state). At this point in time, I can tell you that I will keep coding and developing Falcon/BMS even if I’m the only one in the world using it. So, let’s get over that specific point.

Now that we are clear about that let me add a few things about BMS compared to the rest of the world:

1. We don’t “speak loud” in general. Well, I do, sometimes, but still not really… if we wanted the world to know where we are, then with a few clicks you will have much clearer answers, but that’s not the point and not how we work. Again we don’t need money, so while hype is nice for keeping interest, it’s a momentary thing eventually. We look for the long run.

2. BMS isn’t for everyone. BMS is being developed with the faith that it was meant to be used for “as real as it gets” (i.e dead serious) simulation of a F-16 pilot in a war environment. We don’t mean it to become some “Digital Cinematic Simulator”. For people with that kind of purposes, there are other products, probably much more suitable…

3. Considering all the above said, and while we lack in the GFX department (for now, but being practical, probably always will, at some amounts), we compensate on that with other stuff. Graphics sell, and that’s why you see all commercial products taking care of that first, but when you look under the hood, that’s where the things that matter will be, and I believe there we still have a strong word.

If you or anyone else will decide to not use BMS anymore because he think we are too slow, not advanced enough with chasing technology (We are doing this at our free time yes?), then we will be very sorry to hear that but not much we can do more than we are already doing.

Jp already explained the root cause, if we “remove” the wall you will see basically “nothing” there.

To the more technical explanation:
The gray wall is actually not a wall, it’s the background.
The renderer is currently using a “Fog Box” that covers the entire viewport, then
on top of that then we render sky and terrain.

There are 2 critical problems why this can’t be easily fixed:

1. The SkyBox of BMS isn’t a full dome but a half dome - That is wrong
2. The terrain rendering range won’t cover enough ground in the distance so even if we would create a full dome, it’ll look somewhat wrong, squared-shaped probably at the edges

So for now it is what it is, ugly yes sure we know.

And BTW the terrain rendering range isn’t 10NM but 64KM, that’s not a guess, that’s a number we know. Of course it feels less because by nature the fog is eating some of it towards the viewing camera.

Anyway, we intend to fix ALL of those issues for 4.37.

@AWmk1 Yes competition is always there and by nature I think some of us have very competitive personality (I know that’s true, at least for myself). But what I meant here in my words is more in the sense of: We don’t need to compete, cause we already won

And let me explain before I’m being jumped upon: We won because Falcon was supposed to be dead probably 10-15 years ago. I (personally, and speaking as a “Falconeer” here) find myself lucky that Falcon is still alive and kicking, many folks are using and enjoying it, flights are running daily, PVP sessions are executed on different theaters every now and then, and: It’s being flown the way we aimed for! So even though everything I said above about “I’ll keep developing even if I’m the last one using Falcon”, of course it’s warming our hearts to see a big and strong community!

Also yes sure, I agree that cinematics do sell (As I stated above, GFX sell). But we are out of this game, for various reasons. I can talk here for hours but instead I could post a single vid or screenshot which would explain much better where we are currently and where we are headed, but we don’t do that. Things will come when they will.

@Buzzbomb Yes, trust me I totally agree with you that GFX (and sounds) are part of realism and immersion, that isn’t a question. So basically all the things you have in mind, be sure we don’t miss anything and we know where current technology is. So while it doesn’t mean we will release or implement everything in a single version, it’ll probably take more than that, we don’t intend to miss anything that the DX11 API can be used for and which all the great sim engines out there already have.

Regarding VR it was stated more than once by myself and possibly by others as well - We know that VR is the future, and it’s already not a question of “if” but a question of “when”, only.

@b0bl00i said in On what grounds would you wish 4.37 to be developped?:

-New renderer that actually looks like something from this century while support VR. Good and realistic lighting, proper height maps (sharp mountain peaks and ridges) crisp and sharp textures, ground decals, volumetric clouds, plenty of forests and buildings, huge draw distance, scalable spotting, updated plane models (Doesn’t have to be UE5 good looking but a heck lot better) ** highest prio

You are probably confusing us with M$ or that other sim which keep delivering unfinished alpha versions while the game itself stays unfinished and actually sandbox, forever (Yes forever, you read correctly).

BMS will do the graphical jump but don’t expect it to be a-la M$FS. We aren’t there and even if assuming we can get close, it’ll take time (i.e years, so things will improve but not as fast as you think). We don’t have teams of Devs to develop every small graphical feature. What we can offer though is a graphical improvement that will come on top of an already working sim.

Regarding the 3D models: On what models are you talking really?
Our ugly F-16s? https://i.imgur.com/T56B7jd.png
Ugly B-52? https://i.imgur.com/DbAPGtl.jpeg
Ugly Rafal?: https://i.imgur.com/9AWm4C3.png
Ugly Flanker maybe? https://i.imgur.com/3CPZcOQ.png
Ugly EF? https://i.imgur.com/lB0Rj0e.png
Ugly F-4s maybe? https://cdn.discordapp.com/attachments/263474413909835776/968185715353608192/2022-04-25_192347.png

Really I need to know, because our 3D modelers are doing awesome work and you should show some respect for getting all that, for free…

I’d be willing to chip in on the dev costs through Patreon.

Ha no - We don’t take money, and for sure we don’t work for anyone. This is a hobby for us and we do it for the fun and challenge.

Just to clear what is practical for us let me say that:

• Regarding everything but terrain mesh, we aren’t MS and we don’t see FS2020 as something we can compete with (I don’t think anyone can, I mean not really) - We don’t intend to ever get to that level of details, we don’t have the manpower nor the means to hold/stream this amount of data.

• The way ED are doing their terrain is probably the worse I can think of (Hold chunks of “3D models” and render them)

• In order to create detailed terrain you need DEM data, a flat grid and then use HW tessellation to displace the vertices. That’s the only sane way I know of and literally all engines/samples (And I bet all modern sims/games except DCS) I’ve ever seen are using this way to create a detailed terrain mesh. Of course then when you get into details there are other difficulties to tackle - For example a main one is the tessellation factors that DX support are limited to 1-64. The practical meaning of that is you can only create geometry at scales of 1-64 from the most detailed to the least. It sounds a lot but for a flight sim terrain, this will not be enough and so there are more complications.

• When I started working, I thought the main concern will be the terrain mesh - Well after some years I can tell that the mesh is really just the tip of the iceberg and there are many many more complicated challenges

• Adding details “by hand” to the general mesh can be performed but this isn’t a main goal and will remain a “later work” probably. Maybe also some randomness can be created in order to add even more details when very close, but this is for later.

• There is still time until it is ready for release, but we are working hard to make it happen and I have faith in the team and the excellent guys supporting this work to establish the goals and beyond.

Regarding those “20GB of RAM”, well that was just an illustration I made to let people understand the meaning of holding a 10m mesh data. Of course we aren’t going to hold 20GB of RAM, more like a tenth of that However, loading DEM in chunks isn’t a bad idea, but something that I still didn’t tried. It’ll be a challenge due to view distance, and the ability to “jump” between distanced locations by a key press, action camera etc… I’m not sure how much it’ll worth dividing actually, but that is a low priority for now anyway, there are higher priorities to tackle first

@kyros said in Done:

I’m absolutely done with BMS and the disrespectfull and egodriven characters in this forum and especially in the ‘development’ team. With each new release more bugs than new features are introduced. The fact that there have been (and still are) so many serious bugs in the most basic systems makes me wonder how many people fly this game in a realistic fashion. Actually how many fly this game at all in this rapidly shrinking community. There’s a serious lack of proper testing, wrong priorities and lack of documentation. Serious questions and bug reports are met with sarcasm, rude comments or completely ignored. This community is littered with ego and some individuals are really sick. Like many before me, I’m out.

I took the time to run through your “Nightmare bugs” threads and I’m sorry, but you sim to hang on to tiny details and you aren’t getting convinced when a Dev tells you: “But this is how it’s working”. So in that case I don’t have any sympathy to your post - I honestly expected to read that you couldn’t load the sim, getting continuous CTDs all over, or some showstopper - But if what makes you leave is some “issues” (Mostly in your head) with avionics or AI LGBs falling a couple of seconds longer or some trees at runways… then we can’t help you much.

If you call the BMS Devs “sick” then I suggest you to move on to DCS, I’m sure there you will find everything working as you expect ROFLMAO.

Good luck and don’t let the door hits you on the way out…

@kouzi - Hi, First of all let me say that we integrated your work into the BMS install by default for 4.36 Update 1 that will come sometime in the coming weeks/months. Big thanx for your work!

Regarding the difference between 4.35 and 4.36, the problem was that trail0.dds apparently isn’t part of 4.36, so the effect didn’t show because of the missing texture (PS code won’t run the effect if the texture isn’t defined). Use vapor0/1 instead and it’ll show fine.

Now, regarding the large cloud effect, TBH I didn’t liked it so didn’t integrated it. The code will be in ParticleSys.ini but is disabled (run command commented).
What I do suggest however if you want such effect, is to define it as an additional chain (in .dat file) rather than calling a trail from PS ini. A trail feels very “rough” (IMHO at least) as it “moves by” naturally due to trail behavior, I think a static chains definitions that will be very gentle in alpha and above the wing as a “cloud” (But not too large, means you will probably need more than one) will look better I believe. But anyway I wouldn’t start such effect at level 0, maybe level 1 or maybe even just level 2.

Cheers!

<blockquote>Hi, I would like to know the computer configuration requirements of 4.36 and 4.37.</blockquote><p>4.36 - same as 4.35<br /><br />4.37:<br /><img class=“ql-image” src=“https://i.imgur.com/OUaQBJb.gif” /><br /><br /><br /></p>

Correct, and this was already fixed internally. Contrails will start a bit behind the engine position.

@airtex2019 There is no need to “convince us”
If there is indeed a memory leak of 50-85 MB/minute, I sure hope it would be a more well known effect among many. It’s really hard to tell from 1 report or 2 if there is indeed something real. I’ll anyway forward this report internally and will see, we do have ways to track memory leaks and usually serious ones are found if a coder take the time to inspect closely and use tools that help. I can also tell that in last years we use more and more less risky data structures - More std structures like vectors and less self-allocated arrays or whatever, and so I hope it should lower the chances for a leak.

Cheers!

@dema They should see the other one

@Kevstosmart said in BMS Crashes after time spent in flight - Memory Usage warning appears moments before crash - Memory Leak issue?:

@I-Hawk I can add to this, I have never seen .36 go over 6gb’s and I do not ever restart bms, I’ve been told to do it in .35 due to crashes but never experienced those either.

6GB of BMS makes perfect sense to cause his system to fail. Guys he only has 8GB RAM + 2GB VRAM. If you say BMS on your system is using 6GB at times, and I assume at least 2GB, that leave ~2GB of RAM for Win 10 and everything else, and no reserve for a case where VRAM is overloaded and need to take some RAM to survive.

And BTW 2GB VRAM isn’t even that, if you will check closely you’ll see that the VRAM can never really be entirely filled and there is always some usage even with no 3D app running.

8 + 2 isn’t enough to hold BMS 4.36, so it seems.

@ArcherAC3 said in BMS Crashes after time spent in flight - Memory Usage warning appears moments before crash - Memory Leak issue?:

It shows Display Memory (VRAM): 2010 MB and Shared Memory: 4036 MB.
Now 2010MB VRAM is simply not enough for BMS anymore?

2GB is nothing for texture memory yes. Here is Dev version VRAM usage:

Yes that’s ~7GB. So you can consider 2GB as “very low”. Show me any full game/sim around that is using less than 4-5GB.

Regarding leaks, guys please stop speculating, if BMS had a SERIOUS leak we would know about it and fix it already, I’m 100% sure that BMS has SOME leaks, but those aren’t critical and happen mostly probably at entry/exit to sim or 3D. But not of those are critical, because if they were, we would know.

BMS code is getting better and safer with the years regarding memory allocations, so chances for new leaks are really low, if there are any existing leaks they are probably there since long time and so nothing that could cause a crash.

@ArcherAC3 your case is simple: Texture memory has no more room, the meaning is that the system cannot allocate new textures, that will cause a crash eventually with any app/system. You need more VRAM.

@airtex2019 said in BMS Crashes after time spent in flight - Memory Usage warning appears moments before crash - Memory Leak issue?:

I think the question is, does BMS min-require 4Gb of dedicated video ram … or should this work, and there’s a leak somewhere (eg. managing texture memory as it loads from system ram to video ram).

I don’t know enough about the DirectX programming model, to comment intelligently … is there even an opportunity for such a leak in BMS code – or does DX library manage all the texture-memory swapping between system ram and video ram?

BMS Engine doesn’t leak texture memory. If it would I believe we would know for sure (Our Dev version uses much more texture memory and it’s loading/unloading fine )

@airtex2019 said in BMS Crashes after time spent in flight - Memory Usage warning appears moments before crash - Memory Leak issue?:

@I-Hawk said in BMS Crashes after time spent in flight - Memory Usage warning appears moments before crash - Memory Leak issue?:

no dedicated VRAM

I think 2 Gb. But BMS looks like it uses about 3-3.5 Gb so your point stands … and for a leak rate of 100-200 Mb/min texture data is about the only thing I can think of, to achieve that.

2GB may be the “announced VRAM”, it doesn’t mean that it’s not shared from the main RAM. I could be wrong though, it depends on this specific laptop and GPU model. I have a laptop with GTX-1050 and I know it has dedicated 4GB VRAM. To know how much video RAM you have overall, you may go to Nvidia driver panel --> Help --> System information and there you should see something like:
Dedicated video memory: <amount>
System video memory: <amount>
Shared video memory: <amount>

Those 3 will help understanding how much total VRAM you have. On a normal system (dedicated GPU) the total video memory will be combined from the dedicated amount and half of the total RAM of the system. If you run dxdiag command you should see under Display tab the amount of “Approx Total Memory” which will be the amount of dedicated + shared, that basically will be the max amount of VRAM that your system can use (Example here with RTX-3060TI 8GB and 32GB RAM I have in that field ~24GB of VRAM total which is combined from 8 dedicated and 16 from the RAM)

I can clearly tell you though that BMS 4.36 doesn’t use 6GB of VRAM and not even close to it, at any moment… I bet that your 0 texture memory problem is the consequence of lack of RAM to cover both RAM and VRAM needs.

And regardless, if BMS was leaking 100MB VRAM per minute, it would crash for everyone all over the place with every long flight… I seriously doubt that.

Hi, you are running with 8GB on a laptop that probably has no dedicated VRAM. So your VRAM is being probably eaten from the main RAM and those same 8GB. As you can see in the message the texture/video memory has nothing free, and so the next load of a texture will be fatal.

4.36 may become “heavier” in terms of textures usage and so may be different than 4.35 or anything previous in that aspect.

In general, and to prepare yourself for the future, I’d suggest upgrading to at least 16GB of RAM. In the future 8GB won’t cut it anymore to eve run a campaign mission I’m afraid.

Guys first of all let me apologize for my apparently wrong numbers previously… yesterday I was at work and via AnyDesk connected to my PC, and I don’t know why but my FPS were MUCH lower (My GPU is mining Eth all the time when idle, and I do change my MSI afterburner settings on it for that, it may be not so great returning to normal when I tell it to ).

Right now in the Benchmark TE my numbers are:
4.36: ~67 on the runway and ~100 after 1 minute of taking off
Dev version with new terrain and a lot more stuff: ~54 on runway and ~73 after 1 minute of taking off (Yea Dev version is heavier overall, 4.36 even though less efficient still has a LOT less triangles, Dev version FPS in the air is

Here is a screenshot capture from my 4.36:

So yes ~40 sounds too low for 10850K and 3070. I’d first check:
Nvidia driver
Settings (e.g better set antialiasing from BMS Multi-sampling than Nvidia driver
Make sure your windows is updated (I’ve seen shit in the past from that) and restart before the test if you haven’t done so for a long time

@GP1508 - No hard feelings and that’s why I posted “No offense” at the beginning of my post. I just wanted to correct what I saw inaccurate assumptions in your post. In last years I’ve gained some experience in all the CPU/GPU relations so I have some idea (usually better than the average user/gamer) of what’s happening under the hood.

Cheers!

@GP1508 said in FPS DROP IN THE NEW FALCON BMS 4.36:

@Kevstosmart @Kevstosmart mate, I think this is the one that I learned hardway in my case of Video Streaming Encoder, there are two part that matter,

1.CPU
2.GPU

Both of this part needed to be balance in term of speed ability. The CPU prepared all the matter for the GPU to render. Some people have very powerful GPU, but still experiences frame drop / in BMS can visually seen as FPS drop / stutter. It is caused by, not enough material prepared by the CPU to feed into the GPU to be rendered. The cause of the CPU, is very much complex nowadays, but commonly, the CPU is underpower to prepare the graphic material that we already set. That’s the one main cause of problem. At first time I play BMS, I do experiences FPS drop at the ground, and going smooth after airborne. And this is all the step I do to made the ground / at the ground FPS is capped and steady at 144FPS.

First, I do a denial one, because mostly, my CPU is well okay doing the intense calculation at very heavy gaming at the ULTRA graphic setting, so I think, this is all about the CPU setup at the regedit part Win32PrioritySeparation. Why this one? Because I do also intense calculation in the background using the tracker. It is CPU burden also, so that I have to made it perfectly suited to setup the CPU scheduling so that the BMS FPS steady, both at the ground, and airborne. I do intense math check and come up with this 9 possible setup that I’ll try at those time one by one to see the improvement.

This setup inputed at regedit HKLM/System/CurrentControlSet/Control/PriorityControl/Win32PrioritySeparation
By default it is set to 2, but next I’ll give you all my list to test at those time
Please note, input with hex radio button selected
2A hex
29 hex
28 hex

this three is short fixed around 60ms fixed Quanta at the CPU. 28 is no foreground application boost, 29 medium foreground boost, and 2A High Foreground Boost

next is the famous number strongly sugested by many people
26 hex
25 hex
24 hex

this three is short variable quanta with 24 is no foreground boost and 26 is high foreground boost and 25 is medium

next I usually use these three setup when doing a very heavy CPU calculation while recording and mixing around 700 tracks with all plugins on, and have to get the latency not more than 4ms

1A hex
19 hex
18 hex

this three is long fixed quanta, mean that it gives the lonnger time for CPU to schedulling the thread, with 1A is high foreground boost (it is the best when in term of my recording latency), then 19 is medium, and 18 no foreground at all.

And I come up with this result, 2A in my machine work best, the SIM FPS stabilized, not jink up and down, but, I still see some room of improvement here.

Side note, mostly for BMS, you do not need High END GPU, except you use the very heavy AntiAlias, anisotropic, at the GPU side, not the software side. Term GPU side is, you turn off all the antialias and anisotropic filter in BMS, and use spesific setup at “CARD DRIVER” for BMS program only, that’s when your expensive GPU will kick in. In term of stream, it called, hardware encoder.

So right now we know the three FPS counter
SIM
ren (mean rendered)
FPS

while SIM is the actual FPS that your CPU prepare, it goes into the rendered, for your GPU to perfectly sync and polished to be seen at the FPS (this is the output one).

Side note again, do not expect the GPU to rendered more than the SIM (CPU) feed into the GPU. This is the main problem. No matter how powerful the GPU is, all is CPU dependant. Also note about any trackIR, it is heavy CPU also.

Right now, we understand that, CPU is essential here, while GPU is just, well, so so. At this moment when I see this, I come into conclussion for my machine at those time, if I need 140FPS SIM at the ground, mean that I have to lighten the load of my CPU, or using 4x time faster CPU roughly. From 30 - 60 FPS SIM at the ground, into 140FPS SIM at the ground. Two things I can consider, buy, or lighten the load.

I pick lighten the load, and started to goes into config, eliminate all the graph until minimum visual I can accept, those also lower the resolution. I do using no antialias at the CPU side by turning it at the minimum into the BMS config, and turn it max at my GPU driver page.

After I did this, well, ground FPS SIM CPU is roughly near the same while airborne, and my GPU sync it with monitor perfectly.

Sometimes, gamer tends to kick the graph into something that their CPU can’t chew, assumed that they have powerful GPU. If the game have the hardware option to make all the graph related fully at the GPU power, it is good to have high end GPU, but, in case of BMS, it is CPU dependant.

Let’s conclude this, 2A for CPU, and lighten the load. Lighten the load start with shader, Resolution, and detail, while at the same time all filtering like antialias and anisotropic have to be off and turn it on at the GPU driver page. You might find yours, but I hope it’ll help you.

-GP-

No offense, but you have some mistakes in your assumptions:

1. If you think a modern CPU as 10850K cannot feed the GPU fast enough, then you are wrong. The problem about the GPU not 100% utilized is because inefficiency of the 3D models and terrain systems.

In short details: GPU like batches, it doesn’t like to get feed only a few tris at a time, so if you want it 100% utilized, you’ll need to fed it right. BMS currently don’t do that good enough.

2. Do you really think BMS internal multi-sampling setting is doing anti-aliasing on the CPU?? Come on man this is 2022… The multi-sampling setting is fed into the Back-Buffers and is handled by the GPU in real time.

3. A more general advice - Chasing performance can get you to many placebo effects. I suggest to listen to people who know how things are working in BMS and/or have knowledge of game engines.

@Kevstosmart said in FPS DROP IN THE NEW FALCON BMS 4.36:

@GP1508

I have a 10850k , 3070 and my sim fps is around 40-50fps on the bench mark TE on ground

This game will suffer if it gets a terrain upgrade

Soon as I use janhas models SIM fps tanks on ground
4 f16cm40s for my package
4 f16cm40s escorts

---

Janhas models 32-40fps
Vanilla 48-60

Gpu usage never goes over 50%

In air it goes to normal usually 144 fps…

This Benchmark TE is an extreme case of crowded 3D models/objects very close to you. At some point these will take your FPS down no matter what. Even weather isn’t really a factor and also when you get airborne you will still see relatively low FPS until you get far away enough from the AB. Something there isn’t optimized and it’s a benchmark TE especially for testing that impact and not much more. Janhas models will cause additional hit due to number of tris and draw calls compared to vanilla models.

New terrain engine has not much additional impact here. In 4.36 I’m getting ~42 FPS on the runway in this TE and with a Dev version with new terrain engine I’m at ~38. Not much of a difference, mainly because the bottleneck is elsewhere.

My system: 12700K 3060TI 32GB 2560x1440 and Multi sampling quality 4

With new terrain engine your GPU will be 100% utilized, don’t worry about that.