373 replies to this topic

[htamas]

yes... to use this option you will need to go into the Vpinmame registration and make the DDRaw value a "0" not a "1" like it is as default

 

 

 

 

I've done this for just the roms that I use in VPX .... you of course wont need to do this for EM tables...
your also going to notice your machine main screen might go black or whatever.... this is the because the B2S and rom are loading after the table and pushing the vP player to the background... 
so you will have to Alt+tab to get it to the front.... if your using PBX as the front end.... read back a few pages there is a workaround to fix that 

 

this sounds like alot of work... and it is a big pain in the @$$ ....bbbuuuuttt..... the performance gain I have gotten is worth it ever it was 4x a PITA

 

Well, setting the ddraw to 0 means that the DMD cannot be properly sized... which is a problem on my 2-screen cabinet.

But this indeed made the table work with forced fullscreen. The trouble is that I don't see any visible or measurable performance improvement although I was hoping this will help eliminate the microstutter I get on several (but not all) tables. But even in fullscreen, the ball microstutter is still there

My FPS value is the same 132-133 on pretty much all tables like in windowed fullscreen.

Adding the alt+tab issue on top of this, it appears the forced fullscreen is a no-go because it doesn't help on my rig. Hopefully these issues can be fixed once the beta will become the official new version.

[atarian]

Hi htamas - if you haven't tried already it might be worth trying to run the b2s as an exe and force it to 2 cores. Then force VP to the other 6. If you have only 4 cores try 1 for b2s and 3 for VP. I have always done this and it works well. After switching to true FS as a test I tried not forcing anything to any cores and I got stuttering back.

I am convinced there is a problem for some PCs when not using affinity - whether it's timing or not I have no idea. Another interesting observation is I get better performance with b2s forced to it's own core than I do with VPinmame with no b2s and forcing. My theory is maybe using b2s the Vpinmame processing is tied to the b2s threads, running in separate exe, whereas VPinmame running from VP is running in the VP threads. Again I am no expert but well worth giving the affinity a shot.

For years people said VP only uses one core, but this definitely isn't the case nowadays with the latest versions.

Edited by atarian, 20 May 2016 - 07:22 PM.

[Drybonz]

But even in fullscreen, the ball microstutter is still there

 

So many people are talking about this problem in so many threads, myself included on some tables... I hope we are able to figure out how to fix it.

[hauntfreaks]

 

yes... to use this option you will need to go into the Vpinmame registration and make the DDRaw value a "0" not a "1" like it is as default

 

 

 

 

I've done this for just the roms that I use in VPX .... you of course wont need to do this for EM tables...
your also going to notice your machine main screen might go black or whatever.... this is the because the B2S and rom are loading after the table and pushing the vP player to the background... 
so you will have to Alt+tab to get it to the front.... if your using PBX as the front end.... read back a few pages there is a workaround to fix that 

 

this sounds like alot of work... and it is a big pain in the @$$ ....bbbuuuuttt..... the performance gain I have gotten is worth it ever it was 4x a PITA

 

Well, setting the ddraw to 0 means that the DMD cannot be properly sized... which is a problem on my 2-screen cabinet.

But this indeed made the table work with forced fullscreen. The trouble is that I don't see any visible or measurable performance improvement although I was hoping this will help eliminate the microstutter I get on several (but not all) tables. But even in fullscreen, the ball microstutter is still there

My FPS value is the same 132-133 on pretty much all tables like in windowed fullscreen.

Adding the alt+tab issue on top of this, it appears the forced fullscreen is a no-go because it doesn't help on my rig. Hopefully these issues can be fixed once the beta will become the official new version.

 

 

on those tables with the shutter try this... setting in the table editor

 Capture.PNG   10.13KB   99 downloads

[Drybonz]

on those tables with the shutter try this... setting in the table editor

 

I'm using those settings... There was a guy in another thread talking about having some success with an "alpha ramp" setting, but I couldn't find that in the editor.

Edited by Drybonz, 20 May 2016 - 09:18 PM.

[htamas]

 

 

 

on those tables with the shutter try this... setting in the table editor

Capture.PNG

 

 

Amazing... this fixed the stutter.

Thank you!

[Drybonz]

Oh wait... I see what you are saying.  This table wasn't following my global settings for vsync.  Let me try this out.

[hauntfreaks]

 

on those tables with the shutter try this... setting in the table editor

 

I'm using those settings... There was a guy in another thread talking about having some success with an "alpha ramp" setting, but I couldn't find that in the editor.

 

 apha ramps were removed in VPX

 

 

 

 

 

 

on those tables with the shutter try this... setting in the table editor

Capture.PNG

 

 

Amazing... this fixed the stutter.

Thank you!

 

 

glad to hear... and yeah the down side is the the no DMD resizing, i'm also a 2 sreener

[Drybonz]

Hmm... I still get stutter in Medieval Madness, but I need to check some other tables and make sure they are following my default settings.

yeah the down side is the the no DMD resizing, i'm also a 2 sreener

 

I'm also using two screens with true fullscreen but I can re-size my DMD.

[hauntfreaks]

Hmm... I still get stutter in Medieval Madness, but I need to check some other tables and make sure they are following my default settings.

yeah the down side is the the no DMD resizing, i'm also a 2 sreener

 

I'm also using two screens with true fullscreen but I can re-size my DMD.

 

you only lose the function when the set the DDraw to "0"....

 

and you kinda can resize it, but not properly ... i'll take a pic later

Edited by hauntfreaks, 20 May 2016 - 09:56 PM.

[Drybonz]

you only lose the function when the set the DDraw to "0"....

 

and you kinda can resize it, but properly ... i'll take a pic later

 

 

I have all my tables set to DDraw 0.

[hauntfreaks]

First pic is DDraw set to "1"   second pic set to "0"
"0" puts a large black border around the DMD, no matter how its resized the black border is there when set to "0"
this is what happening on my machine... very weird it isn't the same for everyone....

 

[htamas]

That's exactly how it is on mine too.

[Carny_Priest]

Yes, that's normal behavior for ddraw=0. There are only a few DMD settings that eliminate the black bars. I use h=195 and w=771. I run three screens though. With ddraw off and two screens, then you would have to adjust screenres.txt and display resolutions to work around the DMD size rather than adjust the DMD to fit the available space.


Sent from my iPad using Tapatalk

Edited by Carny_Priest, 20 May 2016 - 11:16 PM.

[Drybonz]

Oh yeah, sorry, I get the black bars around the DMD too... I thought you meant you couldn't resize it at all.

 

I have been sizing it down just a touch like you did in the pic so you can't see the black, but it would be nice to be able to make it a little bigger.

Edited by Drybonz, 21 May 2016 - 01:21 AM.

[sliderpoint]

Same here, it seems to have a maximum size that it stretches to and just keeps adding black border. I've just gone back to fake full screen.


-Mike

[Drybonz]

I noticed tonight that in VPX and VP9 both, if I turn off my second (backglass) monitor, I get a very noticeable improvement in the smoothness of motion of the ball.  I'm not sure why that is.

[jimmyfingers]

Two monitors vs. one does seem to add extra variables to the mix and can affect the ball smoothness.  I recall several years ago while trouble shooting VP 9 gameplay fuidity issues (pre-DX9) that it smoothed things out to only have one monitor on / connected. Several factors may be present, however, there is one major one to look out for with two or more monitors - make absolutely sure that the refresh rates of both screens match.  I used to have a mini-cabinet 2 screen set-up and could overclock the PF to about 75hz, however, if I did not also up the refresh on the backglass monitor to match, I would get significant ball stutter. 

 

I think that there is a reasonable amount of cases being reported about general ball stutter issues that may ultimately have this as at least part of the problem as some T.V.s seem to "like" defaulting to 59hz where computer monitors more at 60.  I think this aspect of verifying / setting the screens exactly the same is extremely important.  I would also do tests further then what Windows / nVidia control panel reports to verify that it's actually working with VP at that setting.  For instance, alternating the primary screen from each of the designated PF and backglass monitors / T.V.s and enabling / playing the default table (removes Pinmame from the equation) with VP's vsync set to 1 - verifying that the frame rate is what it should be (i.e. 60 for each).  Don't do it one monitor at a time as most would already know that changing windows screen configuration does not necessarily keep the same settings per screen as you would have when in an "extended" dual monitor mode.  Hence, the idea to just toggle the "primary" setting and verify that VP is happy / consistent with Windows in what it thinks the refresh rate is.

[ICPjuggla]

I have been using the windowed fullscreen because I'm not offered an option for 1080 true fullscreen in the video options.  I would rather use a true FS because the windowed looses focus and the taskbar pops up over it, etc...

I too have this issue on my cab.

[Drybonz]

make absolutely sure that the refresh rates of both screens match.

 

I think you may be on to something there... my main monitor is 144hz and my backglass is 60hz.  I'll set up a profile to match the HZ and do some testing.  I'll report back and let you know how it went.  Thanks.