View Issue Details

IDProjectCategoryView StatusLast Update
0001313FSSCPgraphicspublic2008-04-07 06:03
Reporterthesource2 Assigned Totaylor  
PrioritynormalSeverityminorReproducibilityalways
Status resolvedResolutionunable to reproduce 
Summary0001313: Lucifer preview model causes severe slowdowns in taylor-0227 if seen after other model previews
DescriptionIt rotates with periodic and frequent 2-second freezes. Try to see other big ships in tech room and then switch to Lucifer (checked in fsport)
TagsNo tags attached.

Activities

taylor

2007-03-04 01:11

administrator   ~0007790

Works fine here.

thesource2

2007-03-04 02:53

reporter   ~0007792

Then I'm confused

thesource2

2007-03-04 21:57

reporter   ~0007797

Note that it happens only in fsport and only with SD Lucifer preview (either in tech room or in briefing). SD Lucifer of fs2 is displayed normally.

VA

2007-03-05 13:34

reporter   ~0007804

This is the HTL Lucifer right?
And if so, does the same thing happen when viewing the Triton or Aeolus?

If there is an issue here, I suspect it'd be because of the 2048 res texture. If other big ships work, it's obviously not polycount, and the Triton & Aeolus are the only other ships with similarly huge textures.

thesource2

2007-03-05 21:52

reporter   ~0007805

1. I repeat for the third time already - this is FSPORT SD Lucifer, not mvp one.
2. mvp SD Lucifer is displayed normally along with Aeolus and Triton.
3. It happens ONLY in preview and not in mission.
4. This is not low framerate but periodic freezes (like it rotates smoothly for 1/4 of round and then freezes for 1 second)
5. Finally, this happens ONLY in taylor-0227 and not in official 3.6.9

VA

2007-03-06 15:00

reporter   ~0007808

Saying it's the port lucifer means nothing to me - I don't have a clue if it uses my HTL version or not.
In fact you said: "SD Lucifer of fs2 is displayed normally.", which to me implies you checked the problematic behaviour of the HTL one against the original FS2 one. :p

As it's the original one with the problems though, nothing I can help with here.

thesource2

2007-03-06 15:14

reporter   ~0007809

FSPORT Lucifer is neither original fs2 nor mvp Lucifer. It is a special model with shield system (looks great!). And the other fact is, like I said before, this happens only in latest built. Can anybody at least confirm this?

ARSPR

2007-03-07 01:47

reporter   ~0007810

Last edited: 2007-03-07 01:56

I'll test it ASAP. (FSPort Lucy is less detailed than mediavps one BUT as you've said it has a "skin" shield mesh).

VA

2007-03-07 11:25

reporter   ~0007811

Might that be causing it then?
I mean, if there's no problem with the other capships in the port (which I presume are also low poly ones), and this one is for all intents and purposes identical to the original except for the shield mesh, that's where I'd start my search. :)

thesource2

2007-03-07 11:54

reporter   ~0007812

The other ships are not low poly since thay are not something special and therefore replaced with mvp models when possible. So the only thing you have to look into is special SD Lucifer model.

ARSPR

2007-03-07 23:20

reporter   ~0007814

Last edited: 2007-03-07 23:26

Just tested. It works fine on my PC with Taylor's 0227 and FSPort v3. (I've tested F3 lab, Tech database and briefing ship render).

thesource2

2007-03-08 00:34

reporter   ~0007816

And I just had this problem with GTB Boanerges (mvp one this time). I'm really confused.

taylor

2007-03-08 01:23

administrator   ~0007818

It just sounds like an issue with your computer (video card issues, not enough memory, etc.) and not any code related problem.

thesource2

2007-03-08 01:27

reporter   ~0007819

Errr... what do you mean not enough memory? HTL Lucifer is displayed normally and simple bomber is not because of memory? And why in official 3.6.9 everything is fine then?

taylor

2007-03-08 04:29

administrator   ~0007822

3.6.9 had bugs which have since been fixed. Because of the bugs not all textures are properly put into memory, and memory is no longer handled the same way anyway. OpenGL is just making better use of various things now and so it behaves a little differently.

That no one else can reproduce this problem means that it is most likely just on your end. Lack of VRAM or a slow video card perhaps, but not a code issue. If someone else ends up reproducing this then we'll have more to go on, but so far I see nothing that indicates a code problem.

thesource2

2007-03-10 18:24

reporter   ~0007829

I just had this problem with Perseus after playing several missions in a row. But after I restarted freespace all was fine. It looks like a memory leak.

thesource2

2007-03-11 01:16

reporter   ~0007830

Yep, it really looks like a memory leak - after several missions with large number of different objects in a row I started to encounter texture losses.

taylor

2008-02-27 08:20

administrator   ~0008912

* BUMP *

Is this still an issue with newer builds?

thesource2

2008-02-28 18:38

reporter   ~0008919

Can you please attach pilot file, that has fs1 lucifer model open in tech room? I've lost mine.

taylor

2008-02-28 22:16

administrator   ~0008920

Just hit CTRL-SHIFT-S and it will show all possible ships.

thesource2

2008-03-01 10:26

reporter   ~0008925

Well, the issue is changed. Freezes became much less noticeable but now they happen with any ship in tech room, not only with Lucifer (at least with fs1 mod).

My hardware is: Intel Core 2 Duo 2.4 GHz, GeForce 8800 GTX (with latest drivers), so this is not a hardware problem - my hardware is far more than enough to run FSO

taylor

2008-03-01 10:42

administrator   ~0008926

Please post a debug log from a session that demonstrated the problem.

thesource2

2008-03-01 15:33

reporter   ~0008927

Can't reproduce problem with debug build.

taylor

2008-04-05 00:30

administrator   ~0009184

Does it still happen with the release version of the same build?

thesource2

2008-04-06 13:57

reporter   ~0009196

Runs smoothly in Goober5000-20080321.
I guess you may close this.

Issue History

Date Modified Username Field Change
2007-03-04 00:43 thesource2 New Issue
2007-03-04 01:11 taylor Note Added: 0007790
2007-03-04 02:53 thesource2 Note Added: 0007792
2007-03-04 21:57 thesource2 Note Added: 0007797
2007-03-05 13:34 VA Note Added: 0007804
2007-03-05 21:52 thesource2 Note Added: 0007805
2007-03-06 15:00 VA Note Added: 0007808
2007-03-06 15:14 thesource2 Note Added: 0007809
2007-03-07 01:47 ARSPR Note Added: 0007810
2007-03-07 01:56 ARSPR Note Edited: 0007810
2007-03-07 11:25 VA Note Added: 0007811
2007-03-07 11:54 thesource2 Note Added: 0007812
2007-03-07 23:20 ARSPR Note Added: 0007814
2007-03-07 23:26 ARSPR Note Edited: 0007814
2007-03-08 00:34 thesource2 Note Added: 0007816
2007-03-08 01:23 taylor Note Added: 0007818
2007-03-08 01:27 thesource2 Note Added: 0007819
2007-03-08 04:29 taylor Note Added: 0007822
2007-03-10 18:24 thesource2 Note Added: 0007829
2007-03-11 01:16 thesource2 Note Added: 0007830
2008-02-27 08:20 taylor Note Added: 0008912
2008-02-28 18:38 thesource2 Note Added: 0008919
2008-02-28 22:16 taylor Note Added: 0008920
2008-03-01 10:26 thesource2 Note Added: 0008925
2008-03-01 10:42 taylor Note Added: 0008926
2008-03-01 15:33 thesource2 Note Added: 0008927
2008-04-05 00:30 taylor Note Added: 0009184
2008-04-06 13:57 thesource2 Note Added: 0009196
2008-04-07 06:03 taylor Status new => resolved
2008-04-07 06:03 taylor Resolution open => unable to reproduce
2008-04-07 06:03 taylor Assigned To => taylor