[dennisat]

I've noticed that from X2013 onwards water areas in Dorset Coast that were previously perfectly good now flick from blue (depends on weather) to black upon small movements in viewpoint. Also, other textures with transparencies are shimmering noticeably as you run past them whereas they previously were quite sharp. There have been no unusual messages on the log and I have made no modifications to my graphics setup.

I have reverted to X2012 and the effects do not occur.

Dennis

[copperpen]

I have just come across this on the new Mid East Plus as well. Happens day or night.

[BB25187]

Hi,

It seems that various other textures / 3D are impacted.
On the two pictures below, you can notice some aliasing effect appearing qith revision #2013 on the track texture and on some outline of the Windows of the locomotive.
I don't use the MSTS environment.
The tuning of the graphic card are the same in both cases. I din't succeed in fixing this by changing the settings of my CG.

At revision #2012
http://imageshack.com/a/img841/853/auiy.jpg

Ar revision #2013
http://imageshack.com/a/img841/3994/bjre.jpg

Hope it helps
Regards

[James Ross]

dennisat, on 12 February 2014 - 06:45 AM, said:

I've noticed that from X2013 onwards water areas in Dorset Coast that were previously perfectly good now flick from blue (depends on weather) to black upon small movements in viewpoint. Also, other textures with transparencies are shimmering noticeably as you run past them whereas they previously were quite sharp. There have been no unusual messages on the log and I have made no modifications to my graphics setup.

The water is fixed in X2013. The texture anisotropic filtering problem is not an OR bug as far as I can tell; we changed nothing about that part of the code. The graphics drivers may no longer be able to cope because of something else we changed, but there's little to nothing we can do about that.

[BB25187]

James Ross, on 14 February 2014 - 12:12 PM, said:

The water is fixed in X2013. The texture anisotropic filtering problem is not an OR bug as far as I can tell; we changed nothing about that part of the code. The graphics drivers may no longer be able to cope because of something else we changed, but there's little to nothing we can do about that.

Co occurrence is not causality for sure. It is puzzling however that the point appears at this precise revision...
By the way, I tried to comment the code below, at line 561 of Material.cs, revision #2013

                    rs.AlphaBlendEnable = true;
                    rs.DestinationBlend = Blend.InverseSourceAlpha;
                    rs.SourceBlend = Blend.SourceAlpha;

And I restored the code of revision #2012:

                    rs.AlphaBlendEnable = false;

And then I am back to the nice, clean track I got with revision #2012.

(Edit / update: Juste "replacing rs.AlphaBlendEnable = true;" by "rs.AlphaBlendEnable = false;" is sufficient)

[lineman]

I'm running revision 2021 and I still get the flashing water. Hope it can be fixed easily!

Wow! Ask and ye shall receive. Just downloaded revision 2029 and all is well again!

Can't thank the team enough, you guys are good!

[James Ross]

BB25187, on 14 February 2014 - 01:10 PM, said:

And then I am back to the nice, clean track I got with revision #2012.

(Edit / update: Juste "replacing rs.AlphaBlendEnable = true;" by "rs.AlphaBlendEnable = false;" is sufficient)

Yes, and you've broken the feature that was added. B) My point is, we didn't change any filtering/anisotropic settings so it's entirely down to the graphics driver that it has changed rendering. There's not a lot we can do about that.

[BB25187]

James Ross, on 14 February 2014 - 02:12 PM, said:

Yes, and you've broken the feature that was added. :p My point is, we didn't change any filtering/anisotropic settings so it's entirely down to the graphics driver that it has changed rendering. There's not a lot we can do about that.

Yes, I perfectly got your point.
But since my graphic driver is up to date and I won't debug it, I prefer to disable this new feature which ends up as a regression rather than an improvement on my machine. B)

Edit / Update: BTW, I noticed that even with this change, the most visible part of the new feature introduced at rev #2013 seems to be preserved: (at least) the terrain is blended in the last 50 (?) meters and thus the pop up effect is still reduced. So, I am happy this way! I may post some videos during the week end to illustrate.