Project WIN Ascension Ultra BETA test thread

[Notebook]

Stock DG is the one I mostly use, its always there, and I'm used to it. Always put the three stock vehicles in my bases, just seems the proper thing to do.

N.

 

[wehaveaproblem]

I almost exclusively use the stock DG for different tests, but I don't test the Ascension Ultra.

ha, you didn't ask to.

But I tend to use the stock DG for testing too... but I don't test UMMU.

 

[PDF]

FYL I would like to be a beta tester & also I'm currently in works to updating my logos on my website,

 

[wehaveaproblem]

Thanks PDF, I've replied to your PM

 

[Hlynkacg]

I generally use the stock DG for testing as well but have been using "Spider" my lunar lander project to test the UMMU functionality

 

[Screamer7]

Tom and Friedrich, will it be a major undertaken to make this base compatible with the D3D 9 client?
Or is it more in the line for Jarmo do develop it?
In the D3D 9 or even the D3D 11 client, there are many more possibilities.
At least I think so.
The transparency issue, frame rates and overall anesthetics can be improved.

 

[wehaveaproblem]

Tom and Friedrich, will it be a major undertaken to make this base compatible with the D3D 9 client?
Or is it more in the line for Jarmo do develop it?
In the D3D 9 or even the D3D 11 client, there are many more possibilities.
At least I think so.
The transparency issue, frame rates and overall anesthetics can be improved.

There was some talk about this in the AU dev thread I believe. Face can tell you if it's a 'major' undertaking, but I know it's certainly an undertaking, and not a priority now. But I hope it will be compatible one day, cos it does make things look a lot sweeter.

 

[Face]

Tom and Friedrich, will it be a major undertaken to make this base compatible with the D3D 9 client?
Or is it more in the line for Jarmo do develop it?
In the D3D 9 or even the D3D 11 client, there are many more possibilities.
At least I think so.
The transparency issue, frame rates and overall anesthetics can be improved.

http://orbiter-forum.com/showthread.php?p=363148&postcount=161

 

[csanders]

FYI, "glideslope MFD" doesn't like the "tab" between the lon/lat location values in the "Ascension Ultra.cfg" - it causes a CTD when one starts the glideslope MFD.

Technically it's a "glideslope" bug though (it's not parsing white-space correctly).

 

[NukeET]

Thanks for that feedback, it's good to get at an idea of the fps numbers. Seems satisfactory so far.. agree?

Most assuredly!

I'll see if I can sort that out without having to tear the hangar to vertices...

:rofl: Laughing about the way you phrased that statement...NOT about the actual mesh re-work. I know how bloody that can be.

Cheers for the vids, much obliged.

Seemed to be the most efficient way...

 

[RisingFury]

Thanks for that feedback, it's good to get at an idea of the fps numbers. Seems satisfactory so far.. agree?

I can't say I agree. Even decent / top range computers are somewhat struggling with AU. If AU was the only base of its type in existence, it'd be ok, but if you're hoping the AU technology becomes standard for base making, then we have a problem... pun not intended...

On Sunday when I get back to my computer, I'm going to test what happens if you put 5 or 10 AU bases all over Earth. I think 5 to 10 top range bases based on this technology is a good estimate.

I'd request every tested does this test as well and posts the FPS.

So the test is:
If you spread 5 AU bases randomly across the Earth at distances greater than 1000 km or so, what happens to the FPS. After that, repeat the test with 10.

 

[storm]

I can't say I agree. Even decent / top range computers are somewhat struggling with AU. If AU was the only base of its type in existence, it'd be ok, but if you're hoping the AU technology becomes standard for base making, then we have a problem... pun not intended...

I have some crude data:

Dual-Core E5700 @ 3.0 GHz 3.0 GHz
4.0 GB RAM (3.0 usable)
32 bit Windows 7 Ultimate
nVidia GT 240 512MB RAM
Running Orbiter in 1400 x 875 window mode

Starting off with the default beta scenario with the standard camera view using D7 client (performance appears to be better in D9) the frames start at 42 FPS. As you progress through the graph you will see random sharp drops, those are due to the scenario editor UI. But basically after creating 4 other bases and placing them randomly throughout Earth, making sure that no other base is in view of the first one, we stabilize around 24 FPS.

In the graph you can see where I start deleting the bases. These are the FPS spikes following the stable frames of 24 FPS. Somewhat strange thing is that after deleting the 3rd base the frames went back up to 42 FPS, even though there is one extra base remaining (the 4th base and AU main). Deleting the 4th base has no impact on the frames.

I saved the scenario with 5 bases and reloaded, and they once again stabilize at 24 FPS with the default camera view.

I don't know how Orbiter's rendering works, so I don't know how the FPS are affected if you add some complex mesh in the session even though it is not being rendered in the current scene. The only other thing left is the coding overhead from the 4 other bases, and I wouldn't expect it to have such a dramatic affect on the frames. Adding 4 other bases effectively cut the frames in half, even though none of the other four bases are in view.

 

[Raven]

Would it be possible to add cargo storage/management for the hangars, similar to woo482's UCGO warehouse, but controlled by TowerMFD?

 

[Screamer7]

I also try to spawn multiply AU bases around the globe.
My frame rate with 1 AU is 33 FPS in the D7 client.
With 5 more AU bases scattered around the globe, it drops to 10 FPS.
In the D3D 9 client I started at 55 FPS.
With 5 Au bases it stabilize at 36 FPS.
Interesting.
My PC specs:
Intel core 2 duo E 4400 at 2 ghz.
2 GB RAM
nVidia 9600 GT with 512 GDR 3 ram.


I done several experiments with the standard AU base.
I filled every hanger with various winged aircraft.
XR1,XR2, XR5, DG 4-2, Stock DG, Delteglider EX.
At he vertical launch facility, i spawn a AMSO saturn and the Energia from Kultch.
With this "full house" my frame rate does not drop significantly.
And the best news is there is no CTD at all.
I launch them all and returned with the winged aircraft to AU without problems.
I used the air traffic mode mfd several kilometers from AU and it worked flawlessly.

 

[wehaveaproblem]

Just a quick one to say thanks for the fps stuff gents. I'm busy this weekend, but will come back to this with a proper answer shortly.

---------- Post added 09-07-12 at 13:14 ---------- Previous post was 08-07-12 at 19:01 ----------

Okay, so lets address this fps stuff:

I can't say I agree. Even decent / top range computers are somewhat struggling with AU.

Unfortunately that's not true. My rig runs AU in excess of 70-100 fps even with a ton of XR fleet vessels added. Tech-wise my rig is actually about a year old, although quite new to me: Intel(R) Core i5-2500k @ 3.6GHZ, 8GB RAM, GeForce GTX 580, Win7.

If AU was the only base of its type in existence, it'd be ok, but if you're hoping the AU technology becomes standard for base making, then we have a problem... pun not intended...

This is a valid point, but I think it's important to remember that not all bases need be as complex as AU wrt meshes and textures. I warrant the code over-head is not the issue, the polys, textures and transparencies are. Therefore, multiple bases are only a problem if all bases are as detailed as AU. And then, it's only a problem for less up-to-date pcs tbh. I stated from the start of AU dev that some users may struggle with the gfx demand, and although I have tried to be as efficient as possible, I don't intend to change that. But that does not undermine the concept of multiple bases using the code tools. It just undermines the ability for some users to have all the pretty bells and whistles, unfortunately. Also bear in mind that LoD options will enable people to reduce the demand in exchange for less detail. This will also allow us to isolate exactly which elements cause the most demand: meshes, textures, transparencies, code, beacons etc etc..

I have some crude data:
Starting off with the default beta scenario ...snip... even though none of the other four bases are in view.

Thanks for that detail. It makes me want to ask the following question, to all of you: What fps do you normally get/expect from Orbiter on average, and how does AU compare with that?

I also try to spawn multiply AU bases around the globe.
With 5 Au bases it stabilize at 36 FPS.
My PC specs:
Intel core 2 duo E 4400 at 2 ghz.
2 GB RAM
nVidia 9600 GT with 512 GDR 3 ram.

Thanks for the numbers. 36 fps with that rig would be pretty good right?

I done several experiments with the standard AU base.
With this "full house" my frame rate does not drop significantly.
And the best news is there is no CTD at all.

Good news. It's interesting to see that extra vessels had no impact on fps. It would seem that AU is playable for you then?

Also, it is worth noting that I think nearly all users will experience a prettier and smoother experience in the dx9 client, as the dx7 engine has limits as you all know. Admittedly it is not compatible atm, but worth bearing in mind for the future I think.

Would it be possible to add cargo storage/management for the hangars, similar to woo482's UCGO warehouse, but controlled by TowerMFD?

I'm not that familiar with woo's addon, but the TA crane will offer cargo management options. Are you asking for the ability to spawn cargos directly into the hangar? or?

 

[NukeET]

This is a valid point, but I think it's important to remember that not all bases need be as complex as AU wrt meshes and textures. I warrant the code over-head is not the issue, the polys, textures and transparencies are. Therefore, multiple bases are only a problem if all bases are as detailed as AU. And then, it's only a problem for less up-to-date pcs tbh. I stated from the start of AU dev that some users may struggle with the gfx demand, and although I have tried to be as efficient as possible, I don't intend to change that. But that does not undermine the concept of multiple bases using the code tools. It just undermines the ability for some users to have all the pretty bells and whistles, unfortunately. Also bear in mind that LoD options will enable people to reduce the demand in exchange for less detail. This will also allow us to isolate exactly which elements cause the most demand: meshes, textures, transparencies, code, beacons etc etc..

Most assuredly!

There's a point where pushing beta testing to the extreme edges of the envelope becomes pointless.

Sending you a PM with further comments.

 

[Screamer7]

Thanks for the numbers. 36 fps with that rig would be pretty good right?


Please note that the 36 FPS is under the D3D 9 client.
AU is not design for D3D 9 client, so it is not valid.
BUT as NukeET stated, it is also rather pointless to have such many bases.
It is more realistic to have many vessels at that base, without to suffer on FPS in the D3D 7 client.

 

[wehaveaproblem]

Cheers Nuke, and cheers for PM.

And thanks for clarification, Screamer7. So, no ctds or fps drop with multiple vessels aside, was the fps acceptable on your rig?

 

[RisingFury]

@WHAP: I know the code isn't the problem. I imagine the code is tedious, but not CPU heavy.

The problem is that the base's mesh is loaded all the time, doesn't matter where on Earth you are. I've proposed a solution, but through further testing we'll see if it's needed. You have to keep in mind that bases aren't the only things running in Orbiter and there are some other CPU heavy add-ons.

That said... I just ordered parts for a new computer, so I'll be able to do some more testing next week when it arrives.

 

[wehaveaproblem]

@WHAP: I know the code isn't the problem. I imagine the code is tedious, but not CPU heavy.

The problem is that the base's mesh is loaded all the time, doesn't matter where on Earth you are. I've proposed a solution, but through further testing we'll see if it's needed. You have to keep in mind that bases aren't the only things running in Orbiter and there are some other CPU heavy add-ons.

That said... I just ordered parts for a new computer, so I'll be able to do some more testing next week when it arrives.

Fair point again about other addons. But, when finished, all the meshes won't be loaded all the time due to the LoD. So perhaps it's best we return to this issue later in the dev when we can start looking at LoD and fps in more detail. In the short term it's nice to have gotten some feedback and numbers now though, and good to know it seems to be running pretty much as I expected at this stage.