2087 replies to this topic

[boiydiego]

running dx9 it gives me a few black lines in t2chrome ,ball is flickering,ball is much faster than in 9.20 or 9.2.1 the speed diffrence of the ball i also see in other table where ball is too fast.is this normal?

Edited by boiydiego, 04 March 2014 - 07:15 PM.

[mukuste]

T2 issue is known. I specifically noted in the release post that flicker is usually due to AA.

 

The ball shouldn't behave differently than in VP9, nothing was changed on the physics in this version. However, someone also mentioned in the VP 9.2.1 beta thread that the ball is much faster in the last builds, but I have no idea what could cause it. (see http://www.vpforums....33#entry257408)

Edited by mukuste, 04 March 2014 - 07:26 PM.

[arngrim]

just set the color on the playfield and plastic images to pure black not gray... takes just a min..

But that dont fix the head from turning into a giant flasher lol...

Untitled.png

just tried ripley believe or not night mod

 

http://www.vpforums....s&showfile=8625

 

lights are off on dx9 at start of the table when they are lit in dx7

 

9.2 Sans titre.png

dx9 Sans titre.png

 

i tried to change the color of the table from grey to black but it didn't change anything.

 

what element exactly do i need to change from grey to black?

[krille81]

Dx9 is my official exe file now. The most tables run fine. Another with small issues. But my fauvoritetable whitewater still have big troubles with the lights.. I hope that will be fixed becouse i like that table. Thank you very much for all the hard work with vp dx9


Sent from my iPhone using Tapatalk

[mukuste]

Dx9 is my official exe file now. The most tables run fine. Another with small issues. But my fauvoritetable whitewater still have big troubles with the lights.. I hope that will be fixed becouse i like that table. Thank you very much for all the hard work with vp dx9

 

I played Whitewater a lot recently - such a fun table! The black insert lights you can fix yourself, just go into Table>Images and set Transparency to pure black for all "pfX" textures.

 

Then the only issue left I think are the two lamps on the upper playfield panel... those will need to be fixed once we have the depth bias parameter.

[krille81]

Thank you!


Sent from my iPhone using Tapatalk

[tipoto]

T2 issue is known. I specifically noted in the release post that flicker is usually due to AA.

Can we have an option to disable AA on the EMreels on a table by table basis? T2 has to be played without AA, except for the ball, otherwise everything is blurred.

[Argo]

I also get some strange grey, oblique lines in "The Simpson Pinball party" and "The Flinstones".

[naboodiver]

Still a couple of bugs.     I am not sure if others are seeing the same but on my table I am seeing the following:

 

Can anyone explain why the ball flickers on Rocky and Bullwinkle?    http://www.vpforums....s&showfile=8782

 

The boogie men on Elvira Party Monster are missing.    http://www.vpforums....s&showfile=9003

[The Loafer]

Found an potential issue (can't call it a bug) with the "Windowed Fullscreen".  I had noticed the only option set for this was the full resolution. The problem is on my desktop PC, I run an eyefinity setup, which is 3 monitors set up via one full resolution of 5968X1080.  This is awesome for games, not so for VP.

 

Is there any way that "windowed fullscreen" can be set up for just one of the 3 monitors?  I suspect perhaps not, the way eyefinity works and the term "fullscreen" makes me think it may not but be aware that when I select "fullscreen" (not windowed fullscreen), no matter what resolution I use, it simply duplicates it to the 3 screens, so it works as expected and I am presented with the desired 1920x1080 resolution. Problem of course is that the VP tech beta doesn't like full screen too much...

 

Hope this helps, not the end of the world for me because 99% of my VP is on my pincab but there are surely others out there who run eyefinity (AMD) or Surround (NVidia) multimonitor set ups.

 

Thanks,

 

Rob

[jimmyfingers]

To return to the bug I discovered while testing some AA aspects and in which I initially referred to a few posts back (http://www.vpforums....=47#entry257169).  Here are two sets of screen captures of a test table with a blue flasher object set at height 60 with a ramp below it at 52.8.  The first pair shows VP9_DX9_test7 vs. VP921 rev930 when the lower ramp has the additive alpha enabled (the flasher object does not).  The second pair shows  VP9_DX9_test7 vs. VP921 rev930 and even when there is no additive alpha enabled on the lower ramp it is still ordered incorrectly..

 

In version VP9_DX9_test7 it does not draw correctly and it orders the ramp below with the additive alpha above that of the flasher object:

 Capture of test table with flasher object height above ramp GI that has additive alpha enabled (In VP9_DX9 incorrectly ordered - displayed).PNG   270.33KB   17 downloads

 

Here is the same table in VP921 rev930:

 Capture of test table with flasher object height above ramp GI that has additive alpha enabled (In VP921 correctly ordered - displayed).PNG   272KB   17 downloads

 

And even without additive alpha enabled on the lower ramp (they both have it off) the order is still incorrect as shown here:

VP9_DX7_test7:

 Capture of test table with flasher object height above ramp GI without additive alpha enabled (In VP9_DX9 incorrectly ordered - displayed).PNG   271.3KB   17 downloads

 

VP921 rev930:

 Capture of test table with flasher object height above ramp GI without additive alpha enabled (In VP921 correctly ordered - displayed).PNG   272.19KB   17 downloads

 

I'm attaching the test table as well to this post.

Attached Files

•  Test Table for Flasher Object Issue with Lower Additive Alpha Ramp Drawing Above in DX9 Builds.vpt   664KB   0 downloads
•  Test Table for Flasher Object Issue with Lower Ramp Drawing Above in DX9 Builds (copy without additive alpha enabled on lower ramp).vpt   664KB   0 downloads

Edited by jimmyfingers, 05 March 2014 - 04:48 AM.

[DJRobX]

Still a couple of bugs.     I am not sure if others are seeing the same but on my table I am seeing the following:

 

Can anyone explain why the ball flickers on Rocky and Bullwinkle?    http://www.vpforums....s&showfile=8782

 

The boogie men on Elvira Party Monster are missing.    http://www.vpforums....s&showfile=9003

 

I reported both of these issues.   I think the problem with Rocky & Bullwinkle is that "Fake anti-aliasing" is turned on in the table options.   Try setting it to 0.   Champion Pub has the same issue. 

 

Not sure why the boogie men are missing.  

Edited by DJRobX, 05 March 2014 - 04:48 AM.

[jimmyfingers]

Actually, further to my last post, even when the higher flasher object has additive alpha enabled and the lower ramp does not, the ordering is still off.  This could be a major issue as there would be lot's of cases, if the flasher object were to be used vs. the old ramp method, where one would have a flasher (with addtivie alpha enabled) but a ramp below it for some other type of alpha based graphic (i.e. flasher object over a transparent lower image where additive alpha is not desired to be on for ramp texture and will not work correctly from these test results).

 

Here are some more screen captures showing this:

In version VP9_DX9_test7 with additive alpha flasher object over non-additive alpha ramp:

 Capture of test table with flasher object with additive alpha enabled above ramp without additive alpha enabled (In VP9_DX9 incorrectly ordered - displayed).PNG   271.49KB   17 downloads

 

In version VP921 rev930 with additive alpha flasher object over non-additive alpha ramp:

 Capture of test table with flasher object with additive alpha enabled above ramp without additive alpha enabled (In VP921 correctly ordered - displayed).PNG   269.65KB   17 downloads

 

EDIT:

Here's actually some better captures using a test pattern instead of the original GI style image I first noticed / tested it with:

In version VP9_DX9_test7 with additive alpha flasher object over non-additive alpha ramp:

 Capture of test table with flasher object with additive alpha enabled above ramp without additive alpha enabled using test pattern (In VP9_DX9 incorrectly ordered - displayed).PNG   259.51KB   17 downloads

 

In version VP921 rev930 with additive alpha flasher object over non-additive alpha ramp:

 Capture of test table with flasher object with additive alpha enabled above ramp without additive alpha enabled using test pattern (In VP921 correctly ordered - displayed).PNG   270.37KB   17 downloads

 

I'm attaching again the updated table with the test pattern as shown in the last two captures:

 

 

Attached Files

•  Test Table for Flasher Object Issue with Lower Ramp Drawing Above in DX9 Builds using test pattern (copy wihtout additive alpha enabled on lower ramp).vpt   676KB   1 downloads

Edited by jimmyfingers, 05 March 2014 - 05:22 AM.

[bodydump]

Actually, further to my last post, even when the higher flasher object has additive alpha enabled and the lower ramp does not, the ordering is still off.  This could be a major issue as there would be lot's of cases, if the flasher object were to be used vs. the old ramp method, where one would have a flasher (with addtivie alpha enabled) but a ramp below it for some other type of alpha based graphic (i.e. flasher object over a transparent lower image where additive alpha is not desired to be on for ramp texture and will not work correctly from these test results).

 

Here are some more screen captures showing this:

In version VP9_DX9_test7 with additive alpha flasher object over non-additive alpha ramp:

Capture of test table with flasher object with additive alpha enabled above ramp without additive alpha enabled (In VP9_DX9 incorrectly ordered - displayed).PNG

 

In version VP921 rev930 with additive alpha flasher object over non-additive alpha ramp:

Capture of test table with flasher object with additive alpha enabled above ramp without additive alpha enabled (In VP921 correctly ordered - displayed).PNG

 

See the previous post for the test table(s).

 

I've been experiencing this issue as well.  I just recently added a lot of flasher objects to my No Good Gofers wip.  All of them are on top (in drawing order and actual height) of many non-additive alpha ramps.  I built it in 921 rev930 and everything looks perfect.  The flashers are on top of everything.  In DX9 test 7 the flashers appear underneath all of the non-additive alpha ramps.

 

Flashers in 921

 

Flashers in DX9 test7

[xzotic]

Should the performance be significantly better than the standard 9.2.0? My initial tests with the DX9 Test7 build is that performance is slightly less than 9.2.0. Not by much but it's definitely not better.

 

Monster Bash shows slight ball stutters every couple of seconds or so whereas 9.2.0 is completely smooth on my system. i7 920 over clocked Nvidia GTX480, win7 32bit.

This was with all AA inc FXAA off in both VP and the Nvidia inspector.

 

A couple of tables had a thin black line across it (on a diagonal) also. Sorry can't recall which ones they were. Not very scientific I know but it looks like there is quite a bit more to do anyway with this so I'll wait and test an updated version another time.

 

I just thought I'd flag the fact that at least on my system performance of the current build is not better and in actual fact ever so slightly worse than 9.2.0.

Edited by xzotic, 05 March 2014 - 05:48 AM.

[chinzman93]

@Mukuste: Really enjoying the latest DX9 build. Great Job! I did have one question, maybe I missed the answer. I was playing around with this tonight on my office machine using what I thought was Test 7 (it was actually Test 6) and running task manager. With all tables on Test 6, I was seeing activity on 2 CPU's (1 about 80 percent used and 1 about 50 percent used) and the third empty. When I noticed that this was Test 6, I switched back to Test 7 and restarted task manager. On Test 7, I noticed load spread evenly across my all 3 of my cores. Just curious if this was an anticipated change? It's great either way.

 

However, since the load is now spread over all the cores (at least on my machine - Win XP with Athlon II Triple Core, 4 GIG and HD 7790), I was wondering if this could be contributing to the ball stutter when dB2S is being used? I have never looked to see how much CPU dB2S uses and whether or not it is single threaded.  Maybe since dB2S no longer has a dedicated core available, people are starting to see stutter. Unfortunately, I don't have it installed on this machine. 

 

Not sure if this is related or not, just thought I would mention it. 

Edited by chinzman93, 05 March 2014 - 06:04 AM.

[mukuste]

 

T2 issue is known. I specifically noted in the release post that flicker is usually due to AA.

Can we have an option to disable AA on the EMreels on a table by table basis? T2 has to be played without AA, except for the ball, otherwise everything is blurred.

 

 

But there already is a per-table AA option.

 

 

 

Found an potential issue (can't call it a bug) with the "Windowed Fullscreen".  I had noticed the only option set for this was the full resolution. The problem is on my desktop PC, I run an eyefinity setup, which is 3 monitors set up via one full resolution of 5968X1080.  This is awesome for games, not so for VP.

 

Is there any way that "windowed fullscreen" can be set up for just one of the 3 monitors?  I suspect perhaps not, the way eyefinity works and the term "fullscreen" makes me think it may not but be aware that when I select "fullscreen" (not windowed fullscreen), no matter what resolution I use, it simply duplicates it to the 3 screens, so it works as expected and I am presented with the desired 1920x1080 resolution. Problem of course is that the VP tech beta doesn't like full screen too much...

 

Hope this helps, not the end of the world for me because 99% of my VP is on my pincab but there are surely others out there who run eyefinity (AMD) or Surround (NVidia) multimonitor set ups.

 

Thanks,

 

Rob

 

Multimonitor support isn't implemented yet. It should be rather simple, but I hope some other dev will lend a hand here since I don't have a spare monitor to test.

 

 

Actually, further to my last post, even when the higher flasher object has additive alpha enabled and the lower ramp does not, the ordering is still off.  This could be a major issue as there would be lot's of cases, if the flasher object were to be used vs. the old ramp method, where one would have a flasher (with addtivie alpha enabled) but a ramp below it for some other type of alpha based graphic (i.e. flasher object over a transparent lower image where additive alpha is not desired to be on for ramp texture and will not work correctly from these test results).

 

[...]

 

Thanks for the clear bug report. This was a simple (but rather subtle) bug, and it's fixed now.

 

 

Should the performance be significantly better than the standard 9.2.0? My initial tests with the DX9 Test7 build is that performance is slightly less than 9.2.0. Not by much but it's definitely not better.

 

Monster Bash shows slight ball stutters every couple of seconds or so whereas 9.2.0 is completely smooth on my system. i7 920 over clocked Nvidia GTX480, win7 32bit.

This was with all AA inc FXAA off in both VP and the Nvidia inspector.

 

A couple of tables had a thin black line across it (on a diagonal) also. Sorry can't recall which ones they were. Not very scientific I know but it looks like there is quite a bit more to do anyway with this so I'll wait and test an updated version another time.

 

I just thought I'd flag the fact that at least on my system performance of the current build is not better and in actual fact ever so slightly worse than 9.2.0.

 

People with your kind of hardware have pretty much universally reported major speedups on MB. (Except that everyone else has been using Win7 64bit, but I don't think it should be an issue.)

Two points:

• Can you give specific FPS numbers from before and after?
• Do you run b2s or a frontend? If so, try disabling them and compare again.

 

 

@Mukuste: Really enjoying the latest DX9 build. Great Job! I did have one question, maybe I missed the answer. I was playing around with this tonight on my office machine using what I thought was Test 7 (it was actually Test 6) and running task manager. With all tables on Test 6, I was seeing activity on 2 CPU's (1 about 80 percent used and 1 about 50 percent used) and the third empty. When I noticed that this was Test 6, I switched back to Test 7 and restarted task manager. On Test 7, I noticed load spread evenly across my all 3 of my cores. Just curious if this was an anticipated change? It's great either way.

 

However, since the load is now spread over all the cores (at least on my machine - Win XP with Athlon II Triple Core, 4 GIG and HD 7790), I was wondering if this could be contributing to the ball stutter when dB2S is being used? I have never looked to see how much CPU dB2S uses and whether or not it is single threaded.  Maybe since dB2S no longer has a dedicated core available, people are starting to see stutter. Unfortunately, I don't have it installed on this machine. 

 

Not sure if this is related or not, just thought I would mention it. 

 

VP should still be (for the most part) single-threaded, just as before. Something is strange if you get load over three cores. Can you check in Task Manager which processes actually have CPU activity and how much?

[BananaBoat]

How are the per table settings accessed?

Sent from my HTC_PN071 using Tapatalk

[ClarkKent]

 

Dx9 is my official exe file now. The most tables run fine. Another with small issues. But my fauvoritetable whitewater still have big troubles with the lights.. I hope that will be fixed becouse i like that table. Thank you very much for all the hard work with vp dx9

 

I played Whitewater a lot recently - such a fun table! The black insert lights you can fix yourself, just go into Table>Images and set Transparency to pure black for all "pfX" textures.

 

Then the only issue left I think are the two lamps on the upper playfield panel... those will need to be fixed once we have the depth bias parameter.

 

Maybe someone can update this wonderful table to the new features? I even reworked the playfield and plastics some time ago but nobody seems to care...

 

https://www.dropbox....al_enhanced.psd

[chinzman93]

@Mukuste: Really enjoying the latest DX9 build. Great Job! I did have one question, maybe I missed the answer. I was playing around with this tonight on my office machine using what I thought was Test 7 (it was actually Test 6) and running task manager. With all tables on Test 6, I was seeing activity on 2 CPU's (1 about 80 percent used and 1 about 50 percent used) and the third empty. When I noticed that this was Test 6, I switched back to Test 7 and restarted task manager. On Test 7, I noticed load spread evenly across my all 3 of my cores. Just curious if this was an anticipated change? It's great either way.

 

However, since the load is now spread over all the cores (at least on my machine - Win XP with Athlon II Triple Core, 4 GIG and HD 7790), I was wondering if this could be contributing to the ball stutter when dB2S is being used? I have never looked to see how much CPU dB2S uses and whether or not it is single threaded.  Maybe since dB2S no longer has a dedicated core available, people are starting to see stutter. Unfortunately, I don't have it installed on this machine. 

 

Not sure if this is related or not, just thought I would mention it. 

Disregard this. After some investigating, the time at which I noticed this was suspiciously close to the logged time of my last scheduled background AV scan....

 

On the bright side, it played very well during the scan..