3889 replies to this topic

[chepas]

I reverted back to normal (height 35) rubbers after fuzzel said they were redrawn couple of day ago. It was only yesterday I went back to height 24 because they were quite bad. Interested to see what fix did.

 

The opacity clipping hasn't made any difference this one.

 

I've left a part of the backdrop (trees) so you can see what's happening in the hole. You can just see the bald head of the (dynamic) model that gets clipped..

https://www.dropbox....7/opac.png?dl=0

[toxie]

i've just done one more fix, so please re-test with the next build.

[chepas]

i've just done one more fix, so please re-test with the next build.

 

Kind of broke it a bit more It looks like some edge detection now on models that are under the playfield.

 

https://github.com/jsm174/vpinball This one I built does say 2284 though so I may have gone backwards?.

Edited by chepas, 29 October 2015 - 08:36 AM.

[toxie]

the latest one is 2295. i also recommend that you use svn, as this features the 'original', no mirroring involved: http://sourceforge.n...code/HEAD/tree/

[toxie]

i just did one more fix.. at least from my testing this should fix your issue..

[Jafjas]

Ok boys and girls, I'm cautiously optimistic I have found the issue with the tearing problem.

Before I left work I did one more Google search for tearing in Windows, and it took me to an obscure article concerning Vsync on newer GFX cards and turning off Aero in Win7 and later, and that it can cause some strange issues. Well I realized I never turned Aero off on my laptop but I've ALWAYS switched my pins to Windows classic desktop.

As soon as I got home I turned Aero back on and played for 10 or so minutes. I saw no screen tearing at all. Then I got brave and also turned on extreme FXAA and ambient occlusion in VPX (which caused random stutter in the past as noted above), and to my utter delight, the ball still moved smoother than John Travolta on a dance floor.

I'm not sure if this is the cause of anyone else's issue, but it may be worth investigating on your end.


Sent from my iPhone using Tapatalk

Hey dyopp!

 

Like I said, my tearing is gone since I enabled the earo again, props for your findings.

What are your system specs, because everything is running smooth on your system with ambient occlusion on.

Ambient occlusion makes my system stutter en laggy.

But it depends on the table i play.

for example Tommy and the adams family play smooth.

 

Jasper

[gStAv]

I have an i7 but couldn't run AO without stutter on my old GTX 560,
Now with my recently installed GTX 970 I can play with AO without any stutter.
Only had the pleasure to play Cirqus Voltaire as of yet.

[JMG]

wow, give the guy a medal who discovered the aero effect on tearing

although now, i dont get any tearing but occasional stuttering balls on my 650 ti...although not consistently and on all VP (9-X) versions...maybe i should think about upgrading a little bit.

[cyberpez]

rev2290 is up:

 

- CodeViewer Preferences: Added colour choosers, corrected tab stop order.
- add ForceReflection for ball. If set the ball will be reflected on the playfield even it's on a ramp or upper playfield.
- workaround for rubber collisions pushing the ball upwards
- fix dynamic object clipping issue if playfield features a material with opacity enabled
- default table: fix ramp exit being too narrow (ball got stuck)
 

Regarding the ForceReflection setting: This is a script command for the ball. The idea is if the ball enters a wire ramp you can enable the ForceReflection option and the ball gets reflected on the playfield but depending on the rest of the table this should be switched off if the ball leaves the wire ramp. Forcing the reflection on upper playfields or plastic ramps will produce a strange unrealistic look.

 

I was able to get around 20 games in on my "problem" table with this build.  I did not notice any of the strange ball behavior that I have seen in past builds.  Rubbers are at a height of 35 with friction at .1 and elasticity at 1.3.  I should have a bit more time tonight to test but looking better so far.  Thanks guys!!

[dyopp21]

 

Ok boys and girls, I'm cautiously optimistic I have found the issue with the tearing problem.

Before I left work I did one more Google search for tearing in Windows, and it took me to an obscure article concerning Vsync on newer GFX cards and turning off Aero in Win7 and later, and that it can cause some strange issues. Well I realized I never turned Aero off on my laptop but I've ALWAYS switched my pins to Windows classic desktop.

As soon as I got home I turned Aero back on and played for 10 or so minutes. I saw no screen tearing at all. Then I got brave and also turned on extreme FXAA and ambient occlusion in VPX (which caused random stutter in the past as noted above), and to my utter delight, the ball still moved smoother than John Travolta on a dance floor.

I'm not sure if this is the cause of anyone else's issue, but it may be worth investigating on your end.


Sent from my iPhone using Tapatalk

Hey dyopp!

 

Like I said, my tearing is gone since I enabled the earo again, props for your findings.

What are your system specs, because everything is running smooth on your system with ambient occlusion on.

Ambient occlusion makes my system stutter en laggy.

But it depends on the table i play.

for example Tommy and the adams family play smooth.

 

Jasper

 

I have two systems, one with a GTX770 and one with a GTX660.  The 770 is smooth with AO turned on, the 660 however, is not smooth.  I have not tested this thoroughly on a lot of tables, so i hesitate to say the 770 will run every table smoothly with AO turned on.

 

My question is this: is AO really a big game changer?  I can't really notice any difference whether it's off or on.

[Trinity]

Game changer? nah. But I can tell a difference.   Personal preference I think.

[ClarkKent]

Supporting real time shadows would be nicer than AO but until then there is no better solution I think...

[BorgDog]

OK, so question from a table making newb, if 35 is the consensus height for rubbers, why when placing a new rubber object it puts them at 25?  And is 35 the correct height for all years models etc of tables?  I've done a bit a searching and can't really find an answer.  I found a few references on pinside talking about wire guide height and they seem to say 5/8" to center, which if I do my math right would be 29.4 vp units.  Is there a simple guide that lists this standard stuff?  I guess I should have asked this before starting to build tables and now probably need to add it to my list of things to correct.

[toxie]

As always: If you find better defaults for elements that match real life behavior, let us know. So should i change the default to 35 then, now that the rubber collision issue seems to be "fixed"?

 

As for AO: Its no game changer for sure, it just adds some subtle detail that makes the table feel a bit more realistic (if you flip between two images, one with it being on, one with it being off, the AO one looks better for sure, but if you have no direct comparison its hard to tell if its even on). I'm also working on a static variant, which means that people will always get a certain amount of this subtle effect on static elements, no matter if its turned on or off (at no cost during gameplay).

[Knorr]

OK, so question from a table making newb, if 35 is the consensus height for rubbers, why when placing a new rubber object it puts them at 25?  And is 35 the correct height for all years models etc of tables?  I've done a bit a searching and can't really find an answer.  I found a few references on pinside talking about wire guide height and they seem to say 5/8" to center, which if I do my math right would be 29.4 vp units.  Is there a simple guide that lists this standard stuff?  I guess I should have asked this before starting to build tables and now probably need to add it to my list of things to correct.

Its because of the peg primitive. The mold for the rubbers is at height 35...
At least thats why i choosed this height.

[freneticamnesic]

I won't play without AO now. It's great.

[cyberpez]

 

OK, so question from a table making newb, if 35 is the consensus height for rubbers, why when placing a new rubber object it puts them at 25?  And is 35 the correct height for all years models etc of tables?  I've done a bit a searching and can't really find an answer.  I found a few references on pinside talking about wire guide height and they seem to say 5/8" to center, which if I do my math right would be 29.4 vp units.  Is there a simple guide that lists this standard stuff?  I guess I should have asked this before starting to build tables and now probably need to add it to my list of things to correct.

Its because of the peg primitive. The mold for the rubbers is at height 35...
At least thats why i choosed this height.

 

 

 

That was my un-scientific reasoning also.

[fuzzel]

I won't play without AO now. It's great.

I won't play without beer. That is great

[Brandrew]

 

I won't play without AO now. It's great.

I won't play without beer. That is great

 

Off Topic: Please pass me another...

[toxie]

 

i've just done one more fix, so please re-test with the next build.

 

Kind of broke it a bit more It looks like some edge detection now on models that are under the playfield.

 

https://github.com/jsm174/vpinball This one I built does say 2284 though so I may have gone backwards?.

 

 

okay, got it finally figured out..

 

first of all make sure you have the latest revision from svn as i did some more (unrelated, but who knows ) changes.

then you have to set the opacity->active checkbox for the material of the playfield (this was ignored previously, but that was a bug), similar as with all other table elements that want to be using transparency.

then also disable the reflection checkbox for all the objects that will go through the playfield while still being visible because of the partially transparent playfield (otherwise you get weird effects as the reflection does not 'know' about the alpha hole in the playfield).