search

openrails / test-launchpad-bugs

TEST repository for Launchpad bug migration exploration
0 stars 0 forks source link

[BUG 1463537] Flashing MSTS water (1.0.1.3095, Thousand Islands Railway) #383

Closed twpol closed 3 years ago

twpol commented 9 years ago

Imported from https://bugs.launchpad.net/bugs/1463537

Property Value
Reported by Paul Charland (p-charlie-0)
Date reported Tue, 09 Jun 2015 19:43:19 GMT

I've had reports of the MSTS layered water flashing on several of the routes I have created, the DAR, STR, and the TIR. These all work in MSTS, and until the latest version of OR had been working fine, but now change visible layers just by rotating the camera angle a degree or two horizontally.

twpol commented 9 years ago

Imported from https://bugs.launchpad.net/or/+bug/1463537/comments/1

Property Value
Posted by Paul Charland (p-charlie-0)
Date posted Tue, 09 Jun 2015 19:43:19 GMT
twpol commented 9 years ago

Imported from https://bugs.launchpad.net/or/+bug/1463537/comments/2

Property Value
Posted by Paul Charland (p-charlie-0)
Date posted Tue, 09 Jun 2015 19:44:27 GMT
twpol commented 9 years ago

Imported from https://bugs.launchpad.net/or/+bug/1463537/comments/3

Property Value
Posted by Paul Charland (p-charlie-0)
Date posted Tue, 09 Jun 2015 19:45:10 GMT
twpol commented 9 years ago

Imported from https://bugs.launchpad.net/or/+bug/1463537/comments/4

Property Value
Posted by Paul Charland (p-charlie-0)
Date posted Tue, 09 Jun 2015 19:45:53 GMT
twpol commented 9 years ago

Imported from https://bugs.launchpad.net/or/+bug/1463537/comments/5

Property Value
Posted by James Ross (twpol)
Date posted Wed, 10 Jun 2015 08:11:23 GMT

I notice in the attached environment file that two of the water layers have exactly the same height (0). Even if Open Rails consistently sends these layers to the GPU in the same order (which it might not be doing, though), I could not guarantee a consistent visual result because of the way GPUs interpolate depth.

For this bug, we should check that OR is consistently sending the layers in order. If we are, this is a content problem that we probably can't fix without a dirty hack.

twpol commented 9 years ago

Imported from https://bugs.launchpad.net/or/+bug/1463537/comments/6

Property Value
Posted by Paul Charland (p-charlie-0)
Date posted Wed, 10 Jun 2015 11:41:14 GMT

Hi James,

I see, it's actually set to -.008 and I checked the other routes in question and they are set the same. I think I did this at the time to eliminate the semi-transparent top layer from showing the "wave" action.

Now, if you can't eliminate this at your end I think an easy solution might be changing the watermid.ace to the same ace at the watertop.ace and even if the "flashing" continues it should show the same color and not be seen. I could just post a substitute texture file for people to replace the mid files in the routes.

I can test that here as I see I used the same settings for the Conn River route I'm working one, there should be the same water problem.

Paul :-)

On 10/06/2015 4:11 AM, James Ross wrote:

I notice in the attached environment file that two of the water layers have exactly the same height (0). Even if Open Rails consistently sends these layers to the GPU in the same order (which it might not be doing, though), I could not guarantee a consistent visual result because of the way GPUs interpolate depth.

For this bug, we should check that OR is consistently sending the layers in order. If we are, this is a content problem that we probably can't fix without a dirty hack.

** Summary changed:

  • Flashing MSTS Water
  • Flashing MSTS water (1.0.1.3095, Thousand Islands Railway)
twpol commented 9 years ago

Imported from https://bugs.launchpad.net/or/+bug/1463537/comments/7

Property Value
Posted by James Ross (twpol)
Date posted Wed, 10 Jun 2015 12:38:24 GMT

That -0.008 is for the uvscroll (part of the animation which I don't think OR currently supports). I'm talking about the layer height in world_water_layer_height(). That's set to -0.5, 0 and 0. So the mid and top layers are at exactly the same height.

twpol commented 9 years ago

Imported from https://bugs.launchpad.net/or/+bug/1463537/comments/8

Property Value
Posted by Paul Charland (p-charlie-0)
Date posted Wed, 10 Jun 2015 13:03:00 GMT

Hi James,

I see! I'll tinker with that and see if the problem goes away. I'll let you know. Maybe this is a Paul bug and nothing to do with OR.

Paul :-)

On 10/06/2015 8:38 AM, James Ross wrote:

That -0.008 is for the uvscroll (part of the animation which I don't think OR currently supports). I'm talking about the layer height in world_water_layer_height(). That's set to -0.5, 0 and 0. So the mid and top layers are at exactly the same height.

twpol commented 9 years ago

Imported from https://bugs.launchpad.net/or/+bug/1463537/comments/9

Property Value
Posted by Paul Charland (p-charlie-0)
Date posted Wed, 10 Jun 2015 14:22:12 GMT

Hi James,

I just changed the watermid height to -0.2 on the DAR route and had another look and the water looks the same but does not flash, so this is on me.

I'll go through the enviro files and do this to all of them and to the routes effected, send them to Nick and if this cures it for him I'll post them to trainsim and ET.

Thanks for helping. Can you give me a day to make sure this is the solution... should be able to close the ticket if this works.

Paul :-)

On 10/06/2015 8:38 AM, James Ross wrote:

That -0.008 is for the uvscroll (part of the animation which I don't think OR currently supports). I'm talking about the layer height in world_water_layer_height(). That's set to -0.5, 0 and 0. So the mid and top layers are at exactly the same height.

twpol commented 9 years ago

Imported from https://bugs.launchpad.net/or/+bug/1463537/comments/10

Property Value
Posted by James Ross (twpol)
Date posted Wed, 10 Jun 2015 14:35:27 GMT

Ah, that's good news. If you're happy that this is a suitable fix, we'll resolve this bug as 'invalid' (sounds harsher than it really means).

Happy to help and don't hesitate to file bugs on anything else that's been reported to you for your routes.

twpol commented 9 years ago

Imported from https://bugs.launchpad.net/or/+bug/1463537/comments/11

Property Value
Posted by Paul Charland (p-charlie-0)
Date posted Wed, 10 Jun 2015 15:14:58 GMT

Hi James,

I've adjusted the files for the DAR route and sent them to Nick, as soon as I hear back from him I'll let you know and you can mark the bug as invalid! It's working fine here, I just want someone else to say so!

Paul :-)

On 10/06/2015 10:35 AM, James Ross wrote:

Ah, that's good news. If you're happy that this is a suitable fix, we'll resolve this bug as 'invalid' (sounds harsher than it really means).

Happy to help and don't hesitate to file bugs on anything else that's been reported to you for your routes.

twpol commented 4 years ago

Imported from https://bugs.launchpad.net/or/+bug/1463537/comments/12

Property Value
Posted by Carlo Santucci (carlosanit1)
Date posted Tue, 21 Apr 2020 13:26:39 GMT

Being the problem common to many routes (see e.g. here http://www.elvastower.com/forums/index.php?/topic/33897-flashing-water/ ; same occurs in MEP), and being the problem not present in MSTS, finding a solution for OR without modifying files would be desirable. This is possible, and a patch is proposed for that.