1488 replies to this topic

[kiwi]

This is a fresh crash, this morning, I was changing the starting angle of the flipper.

 

Crash report VPX rev3592
============
Process: vpinballx3592.exe
Reason: 0xC0000005 - EXCEPTION_ACCESS_VIOLATION at 001B:00470E66
Attempt to read from 0xFFFFFFFF
Thread ID: 0xECC [3788]

Call stack
==========
00470E66  (0x050AC97C 0x00000000 0x00000001 0x00000000)
004C608B  (0x00000000 0x00000000 0x00000000 0x00000000)

Environment
===========
Date/time: 20/1/2019, 09:26:32:156
Number of CPUs: 4
Processor type: 586
System: Windows XP

Memory status
=============
Total Reserved: 69804K (68M) bytes
Total Commited: 245432K (239M) bytes
Total Free: 1781852K (1740M) bytes
Largest Free: 990464K (967M) bytes

Registers
=========
EAX=FFFFFFFF EBX=0019473C ECX=FFFFFFFF EDX=00000199
ESI=7E3AF3C2 EDI=001A5760 EBP=00000170 ESP=00000178 EIP=00000000
FLG=00010246 CS=001B DS=0023 SS=0023 ES=0023 FS=003B GS=0000


Failed to save minidump.

The same here, many crashes when I copy objects from other tables and close the table from which I copied the object,
however, I always save the table before doing so.

 
Today I enabled the reflection elements of the playfield and then I clicked in the nothing, I saw as if there had been a refreshment of the table panel and the checkbox of the reflection has been deselected by itself.

 
Something similar when I changed the parameters of the flipper, but I realized that nothing had changed when I went to re-edit the parameters again.

[sliderpoint]

 

I'm putting up a problem I've seen for quite a while with VP.  Timing of Light Sequences and (with their addition) primitive animations do not play the same on different PC's.  My Cab is an AMD processor and my laptop in an Intel processor.  They both have nvidia graphics cards.  Here are some videos I captured to show the difference.   The cab and laptop vids are to show how much faster my cab shows light sequences than my laptop.  The TimingTest vids show the difference in animation speeds.  That one is also important because it shows that 'timers' do run correctly.  As you can see in the video, the ball hits the target and starts the animation, it also starts a timer to turn on a light 5 seconds later.  The light comes on at the same time on both machines, but animation is way slower on the laptop.

 

Looking to see how this can be solved. 

 

http://gofile.me/2820O/c9SkZOpxG

 

Thanks

-Mike

I suppose you use Win10 on your laptop and what OS is running the cab?

 

 

Win10 on both.

[toxie]

Regarding the strange backdrop lights bug(s): I finally found what is causing all this mysterious behavior and i'm currently searching for an elegant way to fix this, and then we can use these again without getting insane. 

[rothbauerw]

With the playfield mesh hit event, is there a way to read the hit threshold value that triggers the hit event?

[gtxjoe]

Does Activeball.velz work  ?

[rothbauerw]

Ball velocities don't work because it changes by the time you read it in the hit event sub.

[Thalamus]

Might be too early and might even be wrong about it. But, the already mentioned ball through bumpers. Could it be a bug related to Hit threshold ? I've seen it happen on a few tables - my updated Avatar, JPJ' Time Machine and I know that there is one more, but, can't recall which. They all originally was set to a value of 2. Temporary fix - 0,1.

[fuzzel]

With the playfield mesh hit event, is there a way to read the hit threshold value that triggers the hit event?

No there is not way to read it. And I wouldn't use it at the moment because it seems a bit buggy I think the calculation for the threshold goes through the roof if a ball is rolling on a primitive, even a hit threshold of 500 or 1000 of a playfield_mesh gives you unexpected hit events.

 

 

Update: found the bug and will be fixed in the next version. And you will be able to read the HitThreshold from a primitive and/or from a target in it's hit function:

 

sub prim_hit

   msgbox prim.HitThreshold

end sub

Edited by fuzzel, 22 January 2019 - 08:21 PM.

[rothbauerw]

Thanks fuzzel.

[toxie]

Regarding the strange backdrop lights bug(s): I finally found what is causing all this mysterious behavior and i'm currently searching for an elegant way to fix this, and then we can use these again without getting insane. 

 

Next build will have fixes for (hopefully) all backdrop light issues, both for bulb and 'classic' lights.. (and it will also disable transmit for all backdrop lights)

 

Please test a lot! (and also check 'normal' table lights, just to be sure that i did not mess something up by accident for these)

[fuzzel]

rev3608 is up:

 

- separate backdrop decals from main material code (faster and cleaner)

- fix some regressions (introduced with 10.5) with backdrop lights (most prominent: bulb lights had wrong falloff, the 'classic' lights wrong image mapping)

- add support to read out the current hit threshold if the ball hits a Primitive or a HitTarget

- fix hit threshold for special playfield_mesh 

- update BASS library to 2.4.14

- use DX9 builtin sRGB conversion for sprites

- use matching sRGB input function for sprites/BG/etc

- use matching sRGB output function, now that the DX9 builtin functionality for sRGB texture reads is used

- disable software-path mipmap generation with correct sRGB-aware filtering again, as its simply way too slow

[toxie]

Explanation on all the sRGB changes: performance of loading a table and during gameplay should now be roughly the same as before the last revision! (if not, then please let me know)

In addition, the background and decals/lights on the backdrop (i.e. in desktop/DT mode) should be sRGB correct, too, now.

 

And then, as mentioned above, please test lights on the backdrop (and in general, just to be sure), if these are now again how they should be. Plus: They should now also work correctly on Intel/builtin graphics boards.

[xenonph]

Its working for me!!

As you can see the lower left light is now lighting up!!

Now I need to adjust for this change as I think I brightened all the DT lights quite a bit before, to try and get more light out of them.

Thank you Toxie and Fuzzel for all the updates!!!

Your work is very appreciated!!

Carry on.

[Aubrel]

I just tried the new beta and it doesn't work at all for me : I can't launch any table

During "Initializing visuals..." I get "InitPin3D Error code: 80004005" => "Problem occured during Player init" => "Fatal error D3DERR_INVALIDCALL...." and crash

 

PS : previous betas were working great... (2screens cabinet setup, win7 x64, GTX1050ti,... nothing special...)

Edited by Aubrel, 26 January 2019 - 06:07 PM.

[lukpcn]

I’ve installed this 3608 revision over 3592 and I have two errors as well and vpx quits on every table.

first it shows error “InitPin3D Error code: 80004005” and right after there is this one: “Problem occured during Player init”

works like a charm at 3592 rev.

I’ve just overwritten the files… should I do anything more ?

[BorgDog]

I’ve installed this 3608 revision over 3592 and I have two errors as well and vpx quits on every table.

first it shows error “InitPin3D Error code: 80004005” and right after there is this one: “Problem occured during Player init”

works like a charm at 3592 rev.

I’ve just overwritten the files… should I do anything more ?

 

just tried it, I get that too.

[jpsalas]

I have tried many tables and they all work fine on my computer.

Single screen, Windows 10, I7, GTX 1070, not the latest drivers for the GTX, 417.35 (I'll update to the latest driver to see if it still works)

 

Edit: Updated t the latest GTX drivers, 417.71 and still all the tables working fine.

Edited by jpsalas, 27 January 2019 - 05:42 AM.

[lukpcn]

Updated to 417.71 drivers, still no luck.. InitPin3D Error again and again

[Thalamus]

Not tried it on my cab yet. But, updated on my old desktop computer and had no issues. Will try the cab later today. Both these machines have rather old nvidia drivers.

[Aubrel]

Nvidia 399.07 here and same probleme.

so not driver version related...

 

I also tried to set back the defaut nvidia 3D settings, to set screen color to RGB instead of YCbCr444 but still the same issue...

 

BTW my 2nd screen is plugged to my motherboard using AMD Radeon R7 APU. It could be related?

Edited by Aubrel, 27 January 2019 - 09:08 AM.