Jump to content



Photo

The VPM alpha/beta thread


  • Please log in to reply
1790 replies to this topic

#1101 Mrtr32

Mrtr32

    Enthusiast

  • Members
  • PipPipPip
  • 251 posts

  • Flag: Canada

  • Favorite Pinball: Funhouse

Posted 14 December 2020 - 06:04 PM

For me after tested the 3 dll,

 

For the Black Hole table, I have the same weird distortion when ending the game and I have no sound effect when the ball come to lower playfield

 

I don't know wish build I have so I don't have distortion and sound effect lost

 

 

Thank you toxie!



#1102 wiesshund

wiesshund

    VPF Legend

  • Members
  • PipPipPipPipPipPipPip
  • 11,859 posts

  • Flag: United States of America

  • Favorite Pinball: How many can i have?

Posted 14 December 2020 - 06:55 PM

For me after tested the 3 dll,

 

For the Black Hole table, I have the same weird distortion when ending the game and I have no sound effect when the ball come to lower playfield

 

I don't know wish build I have so I don't have distortion and sound effect lost

 

 

Thank you toxie!

file creation/modification times

 

10:14 no forced inline

10:20 floatclocks no 5203

10:24 floatclocks no 5202


If you feel the need to empty your wallet in my direction, i don't have any way to receive it anyways

Spend it on Hookers and Blow


#1103 Thalamus

Thalamus

    Pinball Wizard

  • Platinum Supporter
  • 4,984 posts

  • Flag: Norway

  • Favorite Pinball: GOT, Alien Star, LOTR, TOM

Posted 14 December 2020 - 07:06 PM

I guess my test is run differently. Test table being Whirlwind and this is what I'm seeing.

 

Load the table with any of the new dll 100% clean and it seems ok. Stop the game, press start and you will have distortion and "echo" of some of the sound, mainly side effect and not main tune.

 

Obviously, at some time I fiddled around with vpinmame.dll in april. Have a "backup" - named 20200426 (probably the date I made a copy to try a new one). Running the same test, play a couple of balls, exit and press play without 100% exit of VPX, not being able to reproduce the distortion and echo.

 

Rom used : whirl_l3

 

I've jumped a few times between the good and the newest and this seem to be reproducible.

 

I'll see if I can get some builds from Carny' thread to when this starts to happen. Very hard to say, but, might it be that the sound from the old is even slightly clearer ? Don't put to much into that though. Like yourself, I don't trust my hearing all the time. Tinnitus :\


From now on. I won't help anyone here at VPF. Please ask Noah why that is.


#1104 Thalamus

Thalamus

    Pinball Wizard

  • Platinum Supporter
  • 4,984 posts

  • Flag: Norway

  • Favorite Pinball: GOT, Alien Star, LOTR, TOM

Posted 14 December 2020 - 07:43 PM

Update : There is a clear distortion going on since 5225, at least on my machine. Again !! If you start table, stop, start without exit vpx.

 

5013 - ok
5076 - ok
5134 - ok
5185 - ok
5197 - ok
5199 - ok
5225 - problem


Edited by Thalamus, 14 December 2020 - 07:45 PM.

From now on. I won't help anyone here at VPF. Please ask Noah why that is.


#1105 wiesshund

wiesshund

    VPF Legend

  • Members
  • PipPipPipPipPipPipPip
  • 11,859 posts

  • Flag: United States of America

  • Favorite Pinball: How many can i have?

Posted 14 December 2020 - 08:03 PM

Load the table with any of the new dll 100% clean and it seems ok. Stop the game, press start and you will have distortion and "echo" of some of the sound, mainly side effect and not main tune.

 

 

 

Have a feeling that stems from the fact that during table exit, vpinmame is not actually exiting
only when you terminate VPX does it really exit

In the process, on these later beta builds, something must be...

Well, i am not sure, perhaps duplicating itself, or increasing itself?
When the table is relaunched

 

Like, and this is just an example to explain what i mean, if vpinmame was originally initialized during the 1st table launch with an amplification value
of X, and since it did not actually exit (DLL was still active) when exiting the table, instead of re-initializing from scratch, it did present amplification value + the value it would have initialized with.

Again, just trying to explain what i mean, not suggesting that it is literally that value or something, just that maybe, instead of resetting the entire process
it is adding to or adjusting over the top of what is already initialized.

 

Like thalamus, i am finding initial launch is good
but exit the table and relaunch it a number of times.

 

Granted, for a cabinet user in a cabinet environment, that would be a non issue as they always exit VPX fully between tables
but for a table author, they may launch and exit a table a dozen times.
And a desktop player may not even use a front end and just play everything from VPX, never exiting between plays.

 

One other thing i have noticed outside of VPinMAME 3.2, in the beta versions.
Pressing F3 to reload PinMAME, seems to be more apt to just leave you with a non functional existence of PinMAME
Possibly for same reason? Not actually fully exiting and reinitializing itself?

 

Example
Open firepower table, wait for ROM to finish booting
Now hit F3
Wait for ROM to finish rebooting
Now try to actually start the table.
ROM will not begin play, have to exit VPX and reopen

Maybe different symptoms of same root cause?


If you feel the need to empty your wallet in my direction, i don't have any way to receive it anyways

Spend it on Hookers and Blow


#1106 toxie

toxie

    VPF Veteran

  • VP Dev Team
  • PipPipPipPipPipPip
  • 5,734 posts
  • Location:berlin, germany

  • Flag: Germany

  • Favorite Pinball: AFM

Posted 14 December 2020 - 08:57 PM

Update : There is a clear distortion going on since 5225, at least on my machine. Again !! If you start table, stop, start without exit vpx.

 

5013 - ok
5076 - ok
5134 - ok
5185 - ok
5197 - ok
5199 - ok
5225 - problem

could you please test the 3 builds i posted? cause one of these may have caused this.



#1107 toxie

toxie

    VPF Veteran

  • VP Dev Team
  • PipPipPipPipPipPip
  • 5,734 posts
  • Location:berlin, germany

  • Flag: Germany

  • Favorite Pinball: AFM

Posted 14 December 2020 - 09:13 PM


One other thing i have noticed outside of VPinMAME 3.2, in the beta versions.
Pressing F3 to reload PinMAME, seems to be more apt to just leave you with a non functional existence of PinMAME
Possibly for same reason? Not actually fully exiting and reinitializing itself?

F3 will 'just' reset the machine itself, so like what would happen if one does the same on a real machine. Some machines will not properly reset due to bugs in the emulation.  :/



#1108 Thalamus

Thalamus

    Pinball Wizard

  • Platinum Supporter
  • 4,984 posts

  • Flag: Norway

  • Favorite Pinball: GOT, Alien Star, LOTR, TOM

Posted 14 December 2020 - 09:14 PM

I did toxie - read my first reply. None of those helped.


From now on. I won't help anyone here at VPF. Please ask Noah why that is.


#1109 toxie

toxie

    VPF Veteran

  • VP Dev Team
  • PipPipPipPipPipPip
  • 5,734 posts
  • Location:berlin, germany

  • Flag: Germany

  • Favorite Pinball: AFM

Posted 14 December 2020 - 09:15 PM

Ah, okay, misinterpreted that. Wiesshund, same for you??



#1110 toxie

toxie

    VPF Veteran

  • VP Dev Team
  • PipPipPipPipPipPip
  • 5,734 posts
  • Location:berlin, germany

  • Flag: Germany

  • Favorite Pinball: AFM

Posted 14 December 2020 - 09:58 PM

Some new builds, this time with the exact revs like commited (I suspect rev 5214 now).

Attached Files



#1111 wiesshund

wiesshund

    VPF Legend

  • Members
  • PipPipPipPipPipPipPip
  • 11,859 posts

  • Flag: United States of America

  • Favorite Pinball: How many can i have?

Posted 14 December 2020 - 10:14 PM

Ah, okay, misinterpreted that. Wiesshund, same for you??

 

Yes.

Same as thalamus, if clean exit of VPX, seems ok
If VPX has been ran once, on a PinMAME table
ANY PinMAME table
then subsequent launches result in bad audio, even if it is a different table it does seem.
I did about 20 relaunches of firepower, which i know is a silly amount generally, and it actually got worse at the end
audio began to clip out entirely on the peaks.

exit VPX cleanly and relaunch and all was right again.

 

Will try the 2nd pack of test builds in a bit

 

Again, this may very well go un-noticed by a typical cabinet user
it would be an author or desktop user that runs from vpx directly that would most likely notice it
(or someone going through the list of tables to test)

 

EDIT

Wanted to clarify what i meant by clip out entirely.
You know what audio does in mame on a CPU that is just too slow to keep the audio emulation running?
How the audio will become terribly choppy and cut out all together when it can not keep up?
Ok take that, and add a high level of distortion to it.


Edited by wiesshund, 14 December 2020 - 10:20 PM.

If you feel the need to empty your wallet in my direction, i don't have any way to receive it anyways

Spend it on Hookers and Blow


#1112 Mrtr32

Mrtr32

    Enthusiast

  • Members
  • PipPipPip
  • 251 posts

  • Flag: Canada

  • Favorite Pinball: Funhouse

Posted 14 December 2020 - 10:54 PM

toxie,
 
No luck with all the dlls in testbuilds2, the Black Hole Sound issue is present
 
This is the last Working dll
 
 

Edited by Mrtr32, 15 December 2020 - 12:03 AM.


#1113 Thalamus

Thalamus

    Pinball Wizard

  • Platinum Supporter
  • 4,984 posts

  • Flag: Norway

  • Favorite Pinball: GOT, Alien Star, LOTR, TOM

Posted 14 December 2020 - 11:09 PM

You can figure out that yourself, Just run md5sum or some other checksum program and compare it to the releases that you have to your disposal.


From now on. I won't help anyone here at VPF. Please ask Noah why that is.


#1114 toxie

toxie

    VPF Veteran

  • VP Dev Team
  • PipPipPipPipPipPip
  • 5,734 posts
  • Location:berlin, germany

  • Flag: Germany

  • Favorite Pinball: AFM

Posted 15 December 2020 - 07:23 AM

 

toxie,
 
No luck with all the dlls in testbuilds2, the Black Hole Sound issue is present
 
This is the last Working dll
 
 

 

The Black Hole issue is a known one. Due to the change of Votrax emulation, all games that use this sound chip improved, BUT Black Hole, which now has exactly the issues you pointed out.  :/

So this is independent of the issue we're investigating now. But thanks for helping!

 

Will wait for your tests, Thalamus and Wiesshund then. Thanks!!



#1115 Thalamus

Thalamus

    Pinball Wizard

  • Platinum Supporter
  • 4,984 posts

  • Flag: Norway

  • Favorite Pinball: GOT, Alien Star, LOTR, TOM

Posted 15 December 2020 - 08:56 AM

Ups. Didn't see that you created new test builds - Ok, for me at least, it seems to start from 5214, 5213 is OK, at least for me.

 

use very high precision timer resolution (if possible 0.5ms)

also kick out stoneold timing methods (QueryPerformanceXXX is always supported nowadays)
and thus also disable the then useless rdtsc and high priority options

 

Seems to be my issue.


Edited by Thalamus, 15 December 2020 - 09:01 AM.

From now on. I won't help anyone here at VPF. Please ask Noah why that is.


#1116 wiesshund

wiesshund

    VPF Legend

  • Members
  • PipPipPipPipPipPipPip
  • 11,859 posts

  • Flag: United States of America

  • Favorite Pinball: How many can i have?

Posted 15 December 2020 - 09:48 AM

The Black Hole issue is a known one. Due to the change of Votrax emulation, all games that use this sound chip improved, BUT Black Hole, which now has exactly the issues you pointed out.  :/

 

 

So this is independent of the issue we're investigating now. But thanks for helping!

 

Will wait for your tests, Thalamus and Wiesshund then. Thanks!!

 

 

Black Hole is it's own blackhole LOL
It is perfect all the way through, on the 1st game, it is when it plays the game over sound, that it kills everything for subsequent games
on the votrax chip.

 

But it does that in vpm 3.2 even i am pretty sure, but will retest it.

 

As for current
 

VPinMAME5223.dll:

3rd restart of table, without exiting vpx
audio began to break up and distort
not all of the audio, "fire power" seemed normal
but the hit targets on right side, and the bonus count down when losing the ball were breaking up and clipping

 

4th restart
Not as bad, but still present and would have little bursts of worse audio breakup than the 3rd restart, but the majority was better

From there it was kind of random as to how bad it was going to be

 

VPinMAME5222.dll:
Pretty much the same symptoms, no noticeable difference
I used in ear headphones, good ones, so that i could hear the ROM audio very closely
and i am almost sure that i heard the tiniest hint of distortion, but with out the audio breaking up, on the 2nd restart
not enough for anyone to pay much attention to, but i am sure it was there.

 

VPinMAME5214.dll:

I kept the ball out long enough on the 2nd restart, to run the bonus lights up a few times, because i was sure i heard a bit of raspyness

when the table started on the 2nd relaunch
Then i let the ball hit the drain (Well i didn't let it, it just did)
And there was definite non normal distortion going on, just not to the point of the audio cutting out

By the 3rd restart and it did the same as the previous ones

The others were the same, sos did not seem to be a point in listing them all redundantly

 

Now if you kill VPX, which you have to do to swap pinmame dlls anyway, and get it's process unloaded
then all is fine, and you can sit and run that table as many games as you want.
It is only when exiting back to the editor, and then relaunching.

You can even close the table, reopen it, and the audio will still be messed up.
You can open a different table, and it will just inherit the issue already in progress.

 

This is what happens after you quit and relaunch the table about 20 or so times

 

by then Pinmame begins to have some issues beyond audio though, watch the lights.
And listen to the bonus level sound drop pitch for no reason.

 

Not that i have any clue, but i feel like Pinmame is not getting a clean chance at life when VPX is not fully exited
Like every time you quit and relaunch, instead of getting a nice clean reset and restart, it is just getting more info shoved at it
when it has kind of already established itself a certain way, and adding to or corrupting when it set up on the first launch.

If that makes any sense?


Edited by wiesshund, 15 December 2020 - 09:52 AM.

If you feel the need to empty your wallet in my direction, i don't have any way to receive it anyways

Spend it on Hookers and Blow


#1117 toxie

toxie

    VPF Veteran

  • VP Dev Team
  • PipPipPipPipPipPip
  • 5,734 posts
  • Location:berlin, germany

  • Flag: Germany

  • Favorite Pinball: AFM

Posted 15 December 2020 - 12:01 PM

Okay, so 5214 seems to be it then. Thanks a lot!

Will remove the part that i think is causing this and then do a new build.



#1118 toxie

toxie

    VPF Veteran

  • VP Dev Team
  • PipPipPipPipPipPip
  • 5,734 posts
  • Location:berlin, germany

  • Flag: Germany

  • Favorite Pinball: AFM

Posted 15 December 2020 - 12:30 PM

So, please verify the theory with this build here. Thanks!

Attached Files



#1119 Thalamus

Thalamus

    Pinball Wizard

  • Platinum Supporter
  • 4,984 posts

  • Flag: Norway

  • Favorite Pinball: GOT, Alien Star, LOTR, TOM

Posted 15 December 2020 - 04:18 PM

Yes. This build seems to have resolved the issue for me at least. Awesome, thank you Toxie !


From now on. I won't help anyone here at VPF. Please ask Noah why that is.


#1120 wiesshund

wiesshund

    VPF Legend

  • Members
  • PipPipPipPipPipPipPip
  • 11,859 posts

  • Flag: United States of America

  • Favorite Pinball: How many can i have?

Posted 15 December 2020 - 06:13 PM

So, please verify the theory with this build here. Thanks!

so 5224 i guess revision?
what was the issue?

 

have not got to test yet, but guessing i will get same results as thalamus
since we have seen the same thing so far


If you feel the need to empty your wallet in my direction, i don't have any way to receive it anyways

Spend it on Hookers and Blow