View Issue Details
| ID | Project | Category | View Status | Date Submitted | Last Update |
|---|---|---|---|---|---|
| 0002945 | FSSCP | nebulas | public | 2013-10-27 23:35 | 2014-06-20 23:09 |
| Reporter | AdmiralNelson | Assigned To | MageKing17 | ||
| Priority | normal | Severity | minor | Reproducibility | always |
| Status | resolved | Resolution | fixed | ||
| Platform | Windows | OS | Win 7 | OS Version | 64 |
| Product Version | 3.7.0 | ||||
| Summary | 0002945: No EMP/Lightning sound in debug mode | ||||
| Description | Using 3.7.0 or 3.7.1 debug I do not get any EMP effects or lightning sound effects in nebulae. Standard 3.7.0 or 3.7.1 builds have the expected behavior. | ||||
| Steps To Reproduce | 1. Select 3.7.0/3.7.1 debug in wxLauncher 2. Choose 'No Mod' 3. Start 'As Lightning Fall' mission 4. Note no EMP/no lightning noise | ||||
| Tags | No tags attached. | ||||
| related to | 0003067 | resolved | MageKing17 | Nebula flashing not following player view point |
|
|
I believe this is actually a known issue/limitation of debug builds, although perhaps less known than it used to be. Or I'm just imagining this, but I swear I've heard of this problem before. |
|
|
I can't find much on the forums other than this thread: http://www.hard-light.net/forums/index.php?topic=77889.msg1543193#msg1543193 in which someone mentions that lightning sounds are disabled in debug builds. |
|
|
Well this is odd - just tested with r9986 DEBUG and I get EMP & lighting sounds in 'As Lightning Fall'. I'm running Linux, so that may affect it? Also, which OpenAL were you using, Creative OpenAL or OpenAL Soft? |
|
|
I tried both openAL version initially. Just put in r9986, same result. I don't think any of the 'nebula' features work (EMP, radar range, etc) |
|
|
Just wondering, is it just the sound that is missing, or are all the lightning and EMP visual effects also missing? |
|
|
I see lightning effect, but EMP effects such as HUD scrambling are missing. |
|
|
Just my personal opinion but if this is changed then IMO it should be something that could be toggled on or off (possibly via cheat codes ?). Trying to locate a bug or something with EMP effects going on just sounds rather annoying. |
|
|
I'm most likely going to flag this as a "won't fix" unless somebody can offer a compelling reason elsewise. Nebula EMP/Storm effects have never to my memory ever affected Debug builds and to be perfectly honest, I don't think they should. I get that it's jarring that the SFX have now also decided to take a hiatus, but debug builds are not "playing" builds, they for hunting down issues or validating a mission before using a standard build to actually play. |
|
|
I agree with Zacam, but I just took a look at the code and I don't see an explicit disabling of EMP effects in debug builds. Stuff like HUD scrambling shouldn't be affected by the sound code. Before we close this bug, someone should trace through the debug build and see why EMP effects are not being played. |
|
|
I just took a look into this and found that if you comment out the screen position conditionals in code/nebula/neblightning.cpp at lines 966 and 1013 (leaving three interior lines at 971 and 1018), debug builds get the screen flashes (and lightning sounds, and EMP effects; they're under the same conditional as the screen flash). I'm not sure why those conditionals are always returning false, but only under debug builds. |
|
|
2945.patch (1,271 bytes)
Index: code/nebula/neblightning.cpp
===================================================================
--- code/nebula/neblightning.cpp (revision 10824)
+++ code/nebula/neblightning.cpp (working copy)
@@ -1010,13 +1010,13 @@
g3_project_vertex(&cap->vex[idx]);
// if first frame, keep track of the average screen pos
- if( (c->vex[idx].screen.xyw.x >= 0)
- && (c->vex[idx].screen.xyw.x < gr_screen.max_w)
- && (c->vex[idx].screen.xyw.y >= 0)
- && (c->vex[idx].screen.xyw.y < gr_screen.max_h))
+ if( (cap->vex[idx].screen.xyw.x >= 0)
+ && (cap->vex[idx].screen.xyw.x < gr_screen.max_w)
+ && (cap->vex[idx].screen.xyw.y >= 0)
+ && (cap->vex[idx].screen.xyw.y < gr_screen.max_h))
{
- Nebl_flash_x += c->vex[idx].screen.xyw.x;
- Nebl_flash_y += c->vex[idx].screen.xyw.y;
+ Nebl_flash_x += cap->vex[idx].screen.xyw.x;
+ Nebl_flash_y += cap->vex[idx].screen.xyw.y;
Nebl_flash_count++;
}
}
Index: code/render/3dmath.cpp
===================================================================
--- code/render/3dmath.cpp (revision 10824)
+++ code/render/3dmath.cpp (working copy)
@@ -86,7 +86,7 @@
dest->world = *src;
dest->codes = 0;
- dest->flags |= PF_PROJECTED;
+ dest->flags = 0;
return 0;
}
|
|
|
Okay, digging a little deeper, I've found that if you memset() all l_section objects to 0 immediately after they're initialized... well, if you do nothing else, it "seems" to fix nebula lightning in debug (which is to say, you see flashes and hear the noises and get EMP'd). If you then change the (c->vex[idx].screen.xyw.x >= 0) checks to be "> 0" checks instead, both release and debug stop flashing the screen (and playing lightning sounds and EMP-ing the player because, as I mentioned in my previous note, they fall under the "screen flash" conditional). Here's my best guess for what's been going on (since I can't get usable data out of VS Express 2013's debugger :/ ): - l_structure is being initialized with junk data (in both release and debug, but whatever debug's initializing with makes c->vex[idx].screen show as "off screen" while whatever release's initializing with does not (at least some of the time, anyway)). - g3_transfer_vertex() sets the PF_PROJECTED flag, making g3_project_vertex() do absolutely nothing. - The conditionals are therefore checking this initialized junk data, resulting in screen flashes in release builds and nothing at all in debug builds, but these flashes have nothing whatsoever to do with the actual coordinates of the lightning. The lowest-impact fix would seem to be to change g3_transfer_vertex() to clear flags (like g3_rotate_vertex()) instead of setting PF_PROJECTED (this logic error would seem to have been present since g3_transfer_vertex() was added in SVN revision 709; it wasn't a problem with nebula lightning until SVN revision 956, which started using g3_transfer_vertex() for HTL). While I was at it, I noticed that the conditional at line 1013 should refer to "cap", not "c", so that's fixed in this patch as well. |
|
|
Patch looks good to me. Tested in Linux release & debug and they still work as expected. I'll commit later tonight if no one else gets to it 1st. |
|
|
Fix committed to trunk@10828. |
|
fs2open: trunk r10828 2014-06-19 05:24 Ported: N/A Details Diff |
Fix mantis 2945 (From MageKing17) Don't set PF_PROJECTED in g3_transfer_vertex() as it's not projecting the vertex Note: displaying nebula emp/lightning used to depend on unitialised data! |
Affected Issues 0002945 |
|
| mod - /trunk/fs2_open/code/nebula/neblightning.cpp | Diff File | ||
| mod - /trunk/fs2_open/code/render/3dmath.cpp | Diff File | ||
| Date Modified | Username | Field | Change |
|---|---|---|---|
| 2013-10-27 23:35 | AdmiralNelson | New Issue | |
| 2013-10-28 03:32 | chief1983 | Note Added: 0015364 | |
| 2013-10-28 03:39 | chief1983 | Note Added: 0015365 | |
| 2013-10-28 10:45 | niffiwan | Note Added: 0015368 | |
| 2013-10-28 12:36 | AdmiralNelson | Note Added: 0015369 | |
| 2013-10-30 02:25 | niffiwan | Note Added: 0015373 | |
| 2013-10-30 15:21 | AdmiralNelson | Note Added: 0015374 | |
| 2013-11-23 16:01 | Wanderer | Note Added: 0015460 | |
| 2014-03-02 23:16 | Zacam | Note Added: 0015634 | |
| 2014-03-31 15:50 | Goober5000 | Note Added: 0015693 | |
| 2014-06-16 13:05 | MageKing17 | Note Added: 0015881 | |
| 2014-06-17 06:56 | MageKing17 | File Added: 2945.patch | |
| 2014-06-17 06:56 | MageKing17 | Assigned To | => MageKing17 |
| 2014-06-17 06:56 | MageKing17 | Status | new => assigned |
| 2014-06-17 06:56 | MageKing17 | Note Added: 0015889 | |
| 2014-06-17 06:56 | MageKing17 | Status | assigned => code review |
| 2014-06-19 02:21 | niffiwan | Note Added: 0015894 | |
| 2014-06-19 09:04 | niffiwan | Changeset attached | => fs2open trunk r10828 |
| 2014-06-19 09:04 | niffiwan | Note Added: 0015895 | |
| 2014-06-19 09:04 | niffiwan | Status | code review => resolved |
| 2014-06-19 09:04 | niffiwan | Resolution | open => fixed |
| 2014-06-20 23:09 | MageKing17 | Relationship added | related to 0003067 |
