Board index FlightGear Development Spaceflight

Space Shuttle

Discussion about development and usage of spacecraft

Re: Space Shuttle

Postby Thorsten » Wed Oct 28, 2015 2:43 pm

I know the polygon have a processing cost, but we are in 2015 and not from 2005 to 2010, the graphics cards are very much improved. For example, a few weeks ago I posted the test in a post of this forum, I have verified experimentally that it was possible to obtain a minimal reduction of performance with graphic objects composed of 500,000 vertices (about 40-50 MB of .ac file). At this point I think it is not a problem to increase the vertices, at least in the points with high visual contrast, those in which tends to pause the eye of the seer.


I'm sorry to say, but that's just misleading. You did a test on a high performance graphics card, for which this is true - unfortunately we may not assume that everyone uses a high performance graphics cards. Vertices are and will remain an issue, and using 500.000 vertices will floor most computers used to run FG except those at the high end.

The Space Shuttle has too many vertices for my taste and most of our efforts are vertex reduction, however the problem is not in the exterior model mesh but rather in the cockpit.

Such as rounded corners (the portholes and windows) few dozen extra vertices can make an incredible difference, absolutely no change performance!


Personally I feel the absence of the RMS controls is more of an issue than a few vertices, but as I said, the exterior mesh is mostly the original NASA public domain 3d model, so feel free to get involved and round the windows.
Thorsten
 
Posts: 11605
Joined: Mon Nov 02, 2009 8:33 am

Re: Space Shuttle

Postby Richard » Wed Oct 28, 2015 3:45 pm

abassign wrote in Wed Oct 28, 2015 12:51 pm:...too little detail...


The whole back section of the cockpit is something that I know isn't nice - and needs some attention; as does the joining of the models (the new cockpit and the parts that I've taken from the old model aren't well aligned and need some work on the faces and positioning). You can see all sorts of nastiness at certain angles looking down at the top.

However the cockpit is full of vertices. Too many of them. At one point we had 400k now we're just under 200k as I've been removing them (without destroying the detail) but it's a slow process.

When I originally joined the cockpits models up I made a few errors of judgement to try to get the vertex count down. I've redone the front section (because that's what you usually see) and the rest will get done when time permits. At the moment I'm working on getting the PFD display working which is more important IMO.
Richard
 
Posts: 772
Joined: Sun Nov 02, 2014 10:17 pm
Version: Git
OS: Win10

Re: Space Shuttle

Postby wlbragg » Wed Oct 28, 2015 6:12 pm

EDIT: Richard, I posted this after I saw your last post, so please disregard everything that you are planning on taking care of.

OK, I have figured out a couple things that are wrong but I need input from everyone working on this shuttle. I don't know who did/does what and don't really want to dig through the commits to find out so I'll just explain and you all can decide what we need to do.

First let me start by saying that the z-fighting in at the top of the shuttle cockpit is a conflict between two objects in cockpit_ste5.blend (cockpit.ac) and shuttle_o2.blend (shuttle_o2.ac).
Specifically in cockpit_ste5.blend the objects "Roof_Assembly.001" and the "back-wall" group which contains the trouble object "spaceshutt_spstob_1". They z-fight with "spaceshutt_spstob_1" object in shuttle_o2.blend. Don't confuse the object names being the same as meaning they are duplicate objects, they are not.

I'm not sure how much these picture help but at least it should get you in the vicinity.
ImageImage
These are the object/s in cockpit.ac that are causing the z-fighting or need to be scaled?.
Image


Second the blocking windows to the cargo bay. I already touched on that but that object is in both cockpit_ste5.blend (cockpit.ac) as "spaceshutt_spstob_4.0" and shuttle_o2.blend (shuttle_o2.ac) as "spaceshutt_spstob_4". There is also a "spaceshutt_spstob_4" in cockpit_ste5.blend (cockpit.ac) which is jest along one edge of "spaceshutt_spstob_4.0".
Kansas(2-27-15)/Ohio/Midwest scenery development.
KEQA (2-27-15), 3AU, KRCP Airport Layout
Intel i5 3570K AMDRX480
User avatar
wlbragg
 
Posts: 5531
Joined: Sat Aug 25, 2012 11:31 pm
Location: Kansas (Tornado Alley), USA
Callsign: WC2020
Version: next
OS: Win10/Linux/AMDRX480

Re: Space Shuttle

Postby wlbragg » Wed Oct 28, 2015 6:18 pm

wlbragg wrote in Wed Oct 28, 2015 11:43 am:
Thorsten wrote in Wed Oct 28, 2015 7:06 am:please let me know whether someone will address by then
* the blocked view into the payload bay

To correct it immediately I just hid it and exported the model again and the problem is gone.

Thorsten, to answer the one outstanding issue you are concerned with. The above stands. I see no issue with removing this object for now by hiding it.

Richard, if there is something I can do to help you with the back and top side of the cockpit let me know. All I guess I would need is some specific instruction from you as to how you want me to proceed.
Kansas(2-27-15)/Ohio/Midwest scenery development.
KEQA (2-27-15), 3AU, KRCP Airport Layout
Intel i5 3570K AMDRX480
User avatar
wlbragg
 
Posts: 5531
Joined: Sat Aug 25, 2012 11:31 pm
Location: Kansas (Tornado Alley), USA
Callsign: WC2020
Version: next
OS: Win10/Linux/AMDRX480

Re: Space Shuttle

Postby wlbragg » Wed Oct 28, 2015 10:54 pm

@ japreja, I changed the back wheels to 1.14m (44.88") and they are not sticking through the wing, can you verify that this is the spec size per NASA or the size you used originally? Actually it says 44.9 per http://artifacts.nasa.gov/shuttle_facts_figures.htm

What I am wondering is if you remember what you originally had them size to in blender when they stuck through the wing? Does this sound right?
Kansas(2-27-15)/Ohio/Midwest scenery development.
KEQA (2-27-15), 3AU, KRCP Airport Layout
Intel i5 3570K AMDRX480
User avatar
wlbragg
 
Posts: 5531
Joined: Sat Aug 25, 2012 11:31 pm
Location: Kansas (Tornado Alley), USA
Callsign: WC2020
Version: next
OS: Win10/Linux/AMDRX480

Re: Space Shuttle

Postby abassign » Wed Oct 28, 2015 11:41 pm

@Thorsten

Whenever I try to stimulate to improve the display of a model there are always some reluctance. I understand the problem of the number of vertices, but what I am saying is to use cunning to put in the vertices. Surely it is important to the correct visualization of the aircraft, but if we try to assess which are the "focal" on which the eye dwells here is that you can give the impression of having made a plane extremely detailed, but in reality it is almost a whole illusion!
The example of the windows of the aircraft is exemplary, the eye of the viewer looks at first the corners, but not through the window! The reason stems from the physiology of the eye (it is the same reason for which, if you read a word, unconsciously observe only the first and last letters and the length of the word!), However I do not think this is the place to make a depth study. For these reasons only takes a few dozen vertices that the window assumes an apparent fidelity without any real slowdown.
abassign
 
Posts: 835
Joined: Mon Feb 27, 2012 5:09 pm
Location: Italy (living 5 Km from airport LIME)
Callsign: I-BASSY
Version: 2018.3
OS: Linux Mint 19. x

Re: Space Shuttle

Postby japreja » Thu Oct 29, 2015 1:38 am

wlbragg wrote in Wed Oct 28, 2015 10:54 pm:@ japreja, I changed the back wheels to 1.14m (44.88") and they are not sticking through the wing, can you verify that this is the spec size per NASA or the size you used originally? Actually it says 44.9 per http://artifacts.nasa.gov/shuttle_facts_figures.htm

What I am wondering is if you remember what you originally had them size to in blender when they stuck through the wing? Does this sound right?


I got the size data from Air Michelin, the rims and gear itself is from design drawings I've posted links to in previous posts which also uses the same sizes as the Michelin website.
japreja
 
Posts: 331
Joined: Thu May 07, 2015 11:05 pm
Location: MT, USA
OS: Windows 10 Pro 64bit

Re: Space Shuttle

Postby wlbragg » Thu Oct 29, 2015 2:07 am

That appears to be the same dimensions as the source I used. I wonder why they fit now when they didn't before? Unless the size Blender is reporting and the FG world is different?

So if Blender says 1.14m, that is 44.88" which is close to what shuttle spec says. That should be the correct size, right?

Thorsten, do you want me to commit the wheel change so you can compare the profile again with the real thing?
Kansas(2-27-15)/Ohio/Midwest scenery development.
KEQA (2-27-15), 3AU, KRCP Airport Layout
Intel i5 3570K AMDRX480
User avatar
wlbragg
 
Posts: 5531
Joined: Sat Aug 25, 2012 11:31 pm
Location: Kansas (Tornado Alley), USA
Callsign: WC2020
Version: next
OS: Win10/Linux/AMDRX480

Re: Space Shuttle

Postby Thorsten » Thu Oct 29, 2015 5:46 am

@abassign

Whenever I try to stimulate to improve the display of a model there are always some reluctance. I understand the problem of the number of vertices, but what I am saying is to use cunning to put in the vertices.


Sorry, I don't have the time for this right now. Please just read again what I have written, I believe it outlines my position and I believe you didn't understand it the first time.
Thorsten
 
Posts: 11605
Joined: Mon Nov 02, 2009 8:33 am

Re: Space Shuttle

Postby Thorsten » Thu Oct 29, 2015 5:49 am

Thorsten, do you want me to commit the wheel change so you can compare the profile again with the real thing?


I'm not quite clear on what you did - did you just change the wheel size (my original impression) or did you change the contact point location? Only contact point location would change the profile of the OV on the runway - if you changed the wheel size and compensated by strut compression, everything should be automagically okay.
Thorsten
 
Posts: 11605
Joined: Mon Nov 02, 2009 8:33 am

Re: Space Shuttle

Postby Thorsten » Thu Oct 29, 2015 9:42 am

Richard, is there a way to set the HUD color/brightness? It looks like currently colors are non-emissive for some reason.
Thorsten
 
Posts: 11605
Joined: Mon Nov 02, 2009 8:33 am

Re: Space Shuttle

Postby Thorsten » Thu Oct 29, 2015 12:10 pm

@wlbragg:

@ japreja, I changed the back wheels to 1.14m (44.88") and they are not sticking through the wing,


Currently the gear isn't rendered unless gear-pos-norm is non-zero, so the only time they'd stick through the wing is in the split second gear starts to deploy.
Thorsten
 
Posts: 11605
Joined: Mon Nov 02, 2009 8:33 am

Re: Space Shuttle

Postby wlbragg » Thu Oct 29, 2015 6:07 pm

Thorsten wrote in Thu Oct 29, 2015 12:10 pm:Currently the gear isn't rendered unless gear-pos-norm is non-zero, so the only time they'd stick through the wing is in the split second gear starts to deploy.

That explains it.

I'm not quite clear on what you did - did you just change the wheel size (my original impression) or did you change the contact point location?

I only changed the size. It has not been commited.

if you changed the wheel size and compensated by strut compression, everything should be automagically okay.

But you are trying to gain height in the profile, correct?

I don't understand why or how the compression of the strut is at issue, I don't think I did anything to the strut other than animate it. Unless the animation itself is at issue, as in its range of motion? I never did anything to bogey or contact settings that would change compression.

So I am unsure how to proceed?

I will check the wheel size in FG during deploy and see what I see.
I will also look at the wheel sink into the ground with the new size.

Beyond that what do I need to do to make any change in profile height that you want. Do you want me to change contact point if necessary?
Kansas(2-27-15)/Ohio/Midwest scenery development.
KEQA (2-27-15), 3AU, KRCP Airport Layout
Intel i5 3570K AMDRX480
User avatar
wlbragg
 
Posts: 5531
Joined: Sat Aug 25, 2012 11:31 pm
Location: Kansas (Tornado Alley), USA
Callsign: WC2020
Version: next
OS: Win10/Linux/AMDRX480

Re: Space Shuttle

Postby Richard » Thu Oct 29, 2015 6:23 pm

Thorsten wrote in Thu Oct 29, 2015 9:42 am:Richard, is there a way to set the HUD color/brightness? It looks like currently colors are non-emissive for some reason.


Normally via the emission control in the model xml;

e.g.
Code: Select all
    <animation>
        <object-name>???NAME???</object-name>
        <type>material</type>
        <condition>
            <greater-than>
                <property>/fdm/jsbsim/systems/electrics/ac-essential-bus1</property>
                <value>0</value>
            </greater-than>
            <not>
                <property>/sim/rendering/rembrandt/enabled</property>
            </not>
        </condition>
        <emission>
            <factor-prop>sim/model/f15/controls/HUD/brightness</factor-prop>
            <red>  1</red>
            <green>1</green>
            <blue> 1</blue>
        </emission>
    </animation>
Richard
 
Posts: 772
Joined: Sun Nov 02, 2014 10:17 pm
Version: Git
OS: Win10

Re: Space Shuttle

Postby Richard » Thu Oct 29, 2015 6:41 pm

wlbragg wrote in Wed Oct 28, 2015 6:18 pm:Richard, if there is something I can do to help you with the back and top side of the cockpit let me know. All I guess I would need is some specific instruction from you as to how you want me to proceed.


Wayne, the earlier post you made is pretty much a worklist of what needs fixing. I've got no changes planned to the model.

All of the rear section (from the Cicruit breaker bulkhead directly behind the seats) needs fixing; I think I'd decided to move the larger structural elements onto the main model to just leave the interior detail in the cockpit (my rule for this is stuff that is normally visible from the outside as well, such as the window surrounds);

The window sizing / positioning didn't originally line up at all well. I adjusted things so that the two models sort of line up as they do now; but I can't see any reason to have both sets of geometry (except for the amount of time it'll take to fix it all properly).

So if you're willing to fix this up I'd be very happy; and please accept my apologies for the mess that is the back section (and parts of the ceiling). I'm absolutely snowed under with work at the moment - any spare time is getting the PFD working - and it's looking like I'm going to have even less spare time in the next couple of months so please do fix it if you can.
Richard
 
Posts: 772
Joined: Sun Nov 02, 2014 10:17 pm
Version: Git
OS: Win10

PreviousNext

Return to Spaceflight

Who is online

Users browsing this forum: No registered users and 1 guest