Jump to content

1.14 development


Captain Kitten

Recommended Posts

11 hours ago, arthursgon said:

Why xEnviro team don't do as Hifisimtech and insted of release 1.14 to public as "final version" release as "Beta 1.14" to avoid issues like on 1.12 that were fix on 1.13? Receiving some feedback from community and just after that release the "final version", i think that it would be better.... I do not mean a real public beta, just the "final version" called as Beta for a short while for general feedback. 😉

 

Because it is:
1: not needed
2: way too much work to keep running (you just get noise)
3: we have a beta team
4: No matter how much precautions you make - there will be issues
5: What's even the difference of releasing a "public beta" and then fix for release VS release, then fix a bug with a hotfix?
Basically just how you choose to name it...

We have a plan and a method and we're sticking with it. Any changes just robs us of development time.

 

16 hours ago, Bakr said:

Here are my 10 cents::

  • +10 extra FPS for 1.14
  • +10 extra FPS for 1.15 (post Vulcanization)
  • +10 new features from the wish list gets materialized for 1.16 (to be added as re-utilization of new headroom)

I guess, a lot of trans-formative upgrades are going to happen within the next three months ... let alone the six months to match day zero of respectful FS2020 - or ever OUTSHINE in some respect... as i suppose they (MS) have frozen the environment features already.

 

  • 10 FPS for 1.14? Even we don't know what the end result will be.
  • Vulkan won't help us. Period. It will only make us work hard to even get xE working. No one knows the result.
  • Different features demand different performance. We haven't decided on which path to go. It would be silly to choose anything before we know what 1.14 will bring.

We gained 10 fps now - one of five paths we are tryoing out. Maybe the result of all will give 30fps - maybe we'll end up with just the 10 in the end. We don't know. There are no 10 cents out there that hold their value atm. Just the way it is.

  • Like 1
  • Thanks 2
Link to comment
Share on other sites

Currently I am getting fairly descent frame rates with xEnviro 1.13.  Anywhere from 20 in dense areas up to almost 60 in rural areas.  So for myself, I would be happy to get the same frame rates with 1.14 as 1.13 but without the stuttering and hopefully a little less haze.  It's frustrating watching the ground scenery stutter as it moves by even though I'm getting 45+ frames per second.  Perhaps I have set my bar low but If  1.14 can fix the stuttering, improve the haze a bit and keep the same performance I will be satisfied.

Edited by jfwharton
  • Like 1
Link to comment
Share on other sites

As a follow up, we need not forget about the missing baby elephant in the room that could have went out of sight (i guess)

Whereby -  the performance improvement stems from:

  • (A) XE's code optimization paths - in progress.
  • (B) XP's out-of-box gains and entitled performance benefits for switching to next-gen OpenGL (Vulcan)

Whereby (A = Direct Cause) and (B = Indirect Cause)

More on (B), the fact that X-plane start to make less burden on stress on single core, means leaving more headroom for CPU cycles to serve the computation needs of the third-party addons, such as XE as far as CPU is the bounding factor.

We realize as per previous post in this forum that XE is mostly GPU bound, nevertheless we hope a better overall cumulative experience.

---------

I have spotted a very interesting discussion laminar's developer blog :::

image.png.262bb81c2cd16dd1624b15de5cd2a58e.png 

What i get from this, is that Laminar itself is still in early stage to smooth many rough edges with the Vulcan engine .

And it is clear x-plane will require another year cycle to fully tune and reap the benefits of vulcan... however, even laminar itself interfaces with vulkan via their own internal API which translates/interfaces to metal or vulcan sub-system - which means to me (as a developer) that third-party developers will still have to talk to x-planes own API.

It is like there is a guard on the door.

---

Wishing  everyone a happy new year.

 

Link to comment
Share on other sites

On 12/27/2019 at 8:23 PM, Captain Kitten said:

Laminar has done a LOT of right throughout X-Plane's life line. But there are moments where I shake my head and wonder if they either have something extraordinary up their sleeve or if they're just stubborn beyond belief. I seriously hope they have a Royal Street Flush hidden in Austins Cargo Pants because a quartet of Aces is about to land on the scene.

 

I think it is the latter. The community has beggin for years for better clouds, a better weather system and overall a much more realistic rendition of the sky/atmosphere. What did they change from xp10 to xp11 ? they made their ugly 2d clouds white instead of grey...WOW!
Look at all the addons that are out there! ASXP, FSRWG, SKYMAXX, UWXP, XAMBIENCE, XENVIRO (and many more) all trying to alter the rendition of the sky. And they all fail (to a certain extend) because of the bad underlying structure of XP I mean shouldnt this number of addons that are targeting more or less the same area tell you something as a developer ? No ? Well I know what I would be developing...

I think it is really sad to hear that you guys have no opportunity (as it seems) to have a constructive dialogue with them.

Xplane has some really good things going for them, night lighting is very good, the whole mesh system and osm data is good, the standard airports,  you name it...but with msfs2020 around the corner and beeing a part of the tech alpha I think they do indeed need
a Royal Straight Flush in their hands...

  • Like 1
  • Confused 1
Link to comment
Share on other sites

  • 2 weeks later...
On 12/29/2019 at 9:29 PM, jfwharton said:

Currently I am getting fairly descent frame rates with xEnviro 1.13.  Anywhere from 20 in dense areas up to almost 60 in rural areas.  So for myself, I would be happy to get the same frame rates with 1.14 as 1.13 but without the stuttering and hopefully a little less haze.  It's frustrating watching the ground scenery stutter as it moves by even though I'm getting 45+ frames per second.  Perhaps I have set my bar low but If  1.14 can fix the stuttering, improve the haze a bit and keep the same performance I will be satisfied.

I agree 100% in a flight sim any frames above 20 I'll deal with but the stuttering is killing me especially on gusty crosswind landings 

Link to comment
Share on other sites

  • 2 weeks later...

Breaking the silence. Sort of.

As mentioned  in an earlier statement, January is a very busy time in all of the staffs lives and time is limited to work on xE. But that doesn't mean that it is not being worked on. Its just slow.. xE has been in pieces in the lab undergoing performance enhancement treatments. We had an increase of 5-10 fps depending on what system you are on. This was clear already before the end of last year but now we have prepared xE for the next 4 paths of performance enhancements and we're now rebuilding it to finish test version runs.

So no real news at the moment other than that we are one step further to letting the beta team lose on 1.14.
The images below are from 1.14t12 and is still an incomplete assembly.

What I have stated abnout Vulkan has not changed. Read earlier statement on why the introduction of Vulkan won't break xEnviro.

2095937745_2020-01-2621_55_30-AsPowerBar.jpg.109046c47cd94745b85f9aa1820d15c1.jpg833934839_2020-01-2622_01_35-AsPowerBar.jpg.559d581da1fe7eb21025179341155bf3.jpg556429218_2020-01-2622_02_31-AsPowerBar.jpg.68b0aa8678a3f6bb6c325365d2cb981d.jpg1505943979_2020-01-2622_03_26-AsPowerBar.jpg.cf6c6d4b6723c5091f1e09ca611983f3.jpg170163212_2020-01-2621_52_35-AsPowerBar.jpg.0038e0ae887c8cd980b16e1eed471816.jpg

  • Like 4
Link to comment
Share on other sites

I have some bad news. 😪

592831032_2020-02-1312_50_10-unknown.png(PNGImage25601080pixels).jpg.46a0babbe66f59dfab7ff80a0790c94a.jpg
1.14t26

I have been dreading this message to you and postponed it for too long. We have made a decision now that will affect the path going forward tremendously and things will never be the same. Some of you will be furious, others will be annoyed but I hope that the majority of you will understand. To the ones that will be triggered right away: There would not have been much time difference in development of 1.14 anyway and the only reason you're furious is because you think we're changing directions once again. But we're not. We have planned this for a long time - we're just jumping in the deep end of the pool once again to save some time for what will be part of the foundation of xEnviro 2.0 (That is far down the line still - don't worry). So not really bad news - just different and an appology for the development taking so long. 😊

I honestly thought 1.14 would be a 4 week task. But this angry cat (xEnviro) that we're trying to cram in to the shoe box moving in 320kn (X-Plane) in a 30m/s crosswind (real life) has shown it self to be a challenge.

So... Nope - this is not about Vulkan. Not directly anyways. xE will work in Vulkan. We're talking with LR and now it's just a matter of how much of the features will work. 2/3 of our needs will very likely be covered so most features of xE will be working off the bat once we direct xE towards Vulkan engine. But we're not rebuilding for Vulkan yet. 1.14 will be released independently from Vulkan release and we'll monitor closely which release will be the most suitable for a Vulkan release. Might be 1.15 or might be 1.2. We'll have to take that decision as we go along.

What are you on about, Magnus? Get to the freaking point!

Remember full atmospheric volume rendering? I have mentioned full volume rendering a couple of times before. A version of xEnviro that will be unbound by cloud layers. This is where, when having enough headroom we can shape the scene with what ever we want. (granted we have performance headroom).

We have been testing full volume rendering and semi full volume rendering lately. Full volume will take back too much of the performance we have built up until now so we have decided to go for semi full at this stage. One step in the right direction and towards features like thunderstorms and towering cumulus.
 

282906076_2020-02-1121_36_32-X-System.jpg.79d7d7e6459cbf18cc243559391efb3e.jpg
1.14t21

Why "full / semi full" now?

We're struggling to get our render passes per layer optimized and don't really see the benefit of limiting ourself to defined layers when the struggle to get them all optimized is this hard. Especially in OpenGL - with Vulkan we will go full steam with full volume rendering anyways. And the stuttering 5% of you are experiencing may be related to some of this as well. Full volume rendering is not without risk, but it is more controllable than where we are now. And it will open up for many features that are in the pipeline anyways.

What about FPS?

This is a hard one. We think our FPS gain that we managed to pull off now at 5-10fps gain (depending on equipement) is where we'll stay. My own 1070 has a solid 10fps gain - but it varies from system to system. We will still have four optimization paths left to go through and play with when the build is finished. How much they will gain is not known. We hope that each one will give some return but we have to walk the path to know it for sure.

When?

Not entirely sure. I hope well before Q2 - this is sort of the bad news. We just don't know yet and as per usual ETA's won't be given. xE is yet again on the lab table - but it's not a huge task to get it ready for test version 'ing again. If anything this might even be easier than trying to control the layers. It will definitely make the future easier to handle.

What about bug X, Y and Z?

We take notes of the support tickets that come in and we're aware of stuttering, cockpit lighting bug and so on. Those won't exist in 1.14. The reason why we don't fix those now is that (those two in particular) will need work that is done in 1.14 anyway - but not in a way that makes it easy to provide an easy fix. So we have all men on deck for 1.14 semi full volume rendering now so we can get that out the door as soon as possible.

2094879606_2020-02-1310_51_05-unknown.png(PNGImage25601080pixels).jpg.09f11f22df5dd339f1740a0358f9a51b.jpg1008549909_2020-02-1120_31_00-unknown.png(PNGImage25601080pixels)-Scaled(66).jpg.64bb7aaf9b1a5c3e4d59f5e95581c950.jpg
1.14t26

177585388_2020-02-1312_51_28-unknown.png(PNGImage25601080pixels).jpg.9fc5ac1991db4a4243e3b2cf55d7d45c.jpg

1029650664_2020-02-1312_50_38-unknown.png(PNGImage25601080pixels).jpg.3e1dbe77e13a8afdab4719ff1d5daad8.jpg1417479285_2020-02-1312_51_06-unknown.png(PNGImage25601080pixels).jpg.18ee49be94eeda00a752c3a70c85c45e.jpg
1.14t26

Cloud shapes have been reworked somewhat. Semi full volume rendering gives us some more headroom and tools to get this in. It's still a balance between grain and definition. This will be an ongoing battle until we gain enough performance headroom to double the step sampling resolution. Main focus now is known issues and performance. That has not changed!

Test versions are in final stages (I hope) then we're going through a deep beta run before release. Progress has definitely been made.


(What's the difference between full and semi full volume rendering? That's a trade secret 😉 )

Enough babbling from me. Back to work 😃
1959148022_2019-10-2718_46_53-(2)SneakPeek!andUpdateonScrappy!-YouTube.jpg.61728ad540ab9c930c8df33537e3c9a3.jpg
Mike, you're an inspiration of a higher dimention 😄

  • Like 6
  • Thanks 1
Link to comment
Share on other sites

  • 2 weeks later...

Is funny to see that volumetric clouds on FS2020 has the same problems that xE have (except perfomance). Grain is way worse than in xE. Pics of 1.14 looks amazing.

Any chance of an open beta? More feedback could make the final version more stable.

  • Like 1
Link to comment
Share on other sites

i hope the benchmark performance of system spec including gtx 1050ti , because that my gpu, and also that convinced me to buy xEnviro from the info you give, keep it up, everyday check for update for performance, i have to reduced the resolution to 720p to get above 20fps.

Link to comment
Share on other sites

17 hours ago, Nathos said:

Is funny to see that volumetric clouds on FS2020 has the same problems that xE have (except perfomance). Grain is way worse than in xE. Pics of 1.14 looks amazing.

Any chance of an open beta? More feedback could make the final version more stable.

Nathos, can you shed more light in your comparative findings thus far between the two platforms, 
are you having access to alpha MSFS2020

 

The last screenshots posted above - promise a bigger expectation from Xe, as they are gorgeous 

Edited by Bakr
Link to comment
Share on other sites

2 hours ago, Bakr said:

Nathos, can you shed more light in your comparative findings thus far between the two platforms, 
are you having access to alpha MSFS2020

 

The last screenshots posted above - promise a bigger expectation from Xe, as they are gorgeous 

Sadly i can not say more, but if you watch some alpha techs videos in 4K you can easily see some problems with clouds, like flickering and grain. Also, it seems that the problem get way worse when running the sim on lower resolutions.

Link to comment
Share on other sites

In fact, am not sure if the XE CLUB members have really noticed those last screen-shots.

Update from Kaptain Kitten sounded apparently as bad news -- I SAY IT IS ACTUALLY GOOD NEWS (this is a MILESTONE update actually)
based on recent discussions in this forum, it is now established by now that it is the same technology being used to render the noise-based clouds (XE vs MSFS2020)

Here are my observations:

Before Today's Dev Update

no doubt the clouds in MSFS2020 looked far superior than XE 1.13 by measures, in terms of 

  • Cloud shapes: 
    The cloud puff in MSFS2020 looked deeper and more authentic than XE, especially from distance. 👍

    though I was not convinced with clouds in MSFS2020 which were very close to airplane, as -they looked less impressive -  rather like *white fade in the eyes*, less defined (probably this is the trick how they juice more performance, so that the visuals looks great from far distance, giving that sense of huge cloud definition coupled with plausible light-scattering-effects and color shades )👎
     
  • Light Scattering:
    The rendering and scattering and shadows casting shade on each other is clearly superior in MSFS2020, and also thanks for the multi-layering which plays nice along.

After Todays's Dev Update

  • Cloud shapes: 
    The reworked cloud shapes in XE 1.14 - according to my eyes - are sharper, more defined, and having finally that Deep Cloud Puff feeling, giving the cloud a better visual volume and is sharp to extent of /satisfying the retina/ ----- thus i would proclaim at this point that XE (private alpha Clouds) are better than MSFS2020 (public alpha clouds).. 👍

    From far and also low Distances, the new reworked cloud shapes in XE 1.14 are SUPERIOR - (based on above screenshots)
     
  • Light Scattering:
    It seems that XE teams had done some color-shading tuning in the lab, to make the cloud capture light in a much better algorithms than 1.13
    The overall tune is improving, to be almost on a part with MSFS2020 - the gap has narrowed considerably - overall they have crossed the barrier of my taste of satisfaction - Kodus on that. 👍

----------------

Big Thanks for the DEV Team on their achievement, if we look way back to 1.10 and compare with 1.14 alpha screenshots - and considering the plan going forward, i expect that in few months time, X-Plane + XE 1.14 is going to set stage and give us a taste of competition...
especially coupled with a proper airplane such as FlyJSim's Q4XP (https://www.flyjsim.com/q4xp) , which proofed that other airplane developers where not really making justice to what x-plane can do today. see the screenshots and judge for your self.

I dont see any further disruptive changes for the XP 11.x platform - except we have to wait for 12.0 that should address the limitations which are now a show stopper going forward - Laminar knows that.

They are doing good effort already to live the transformation phase, to cleanup their table of priorities and takes the lessons to re-engineer the Next-Gen engine for XP.

 

Edited by Bakr
  • Like 1
Link to comment
Share on other sites

12 hours ago, Bakr said:

Is it possible to post a video recording for 1.14t26, we miss the days of 1.10 alpha/beta and the beautifully commented sneak-peak videos.

Yeah tempt us more with videos! Proud to say I like what I see...  V1.14 with the fps increase you witnessed wasn't releasable as it was?
Performance is the main reason the community uninstalls it... just feedback I'm sure you've heard before.

I think I need to be on the beta team... where will i find the time.

Edited by ashmanmedia
Link to comment
Share on other sites

29 minutes ago, dkabob said:

I do hope this X-Enviro 2.0 you mentioned in the update @Captain Kitten will be a free update to existing users.  Kinda been sitting on broken software since this project started and would like to see a finished product.

2.0 will not be a free update.

But ---> "...what will be part of the foundation of xEnviro 2.0 (That is far down the line still - don't worry)."
 

--> So don't worry.

Link to comment
Share on other sites

Guest
This topic is now closed to further replies.
  • Recently Browsing   0 members

    • No registered users viewing this page.



  • Don't have an account?

    An account is required to interact with and download content from the Threshold Forum.
    Forum is currently also locked down for guests because of heavy bot invasions. Sorry about that.

  • Threshold News

  • image.png.5d3ebd5638179a05dc5cfa67eaf4bf3a.png

×
×
  • Create New...

Important Information

Please read the Terms of Use