1790 replies to this topic

[ClarkKent]

I installed IJ, T2 and TZ and noticed that some sounds are extremely low and others have normal loudness. E.g. the main theme in IJ can be barely heard but others are loud as expected, mechanical sounds are always loud. Weird.

[toxie]

There is still some issue with the gain.txt handling, as i don't know in which units this is specified. As a workaround you can just delete all .txt files in the directories.

[ClarkKent]

Thanx for the info! Maybe you can contact the PinSound designers because of the units?

[ClarkKent]

I renamed all .txt files to .txt.bak (ReNamer is a good program!. But I did not notice any difference. Music is still very quiet. Speech in T2 can almost not be heard at all...

[toxie]

.txt.bak will still be detected..

[ClarkKent]

really!? would it help to rename them to gain.bak?

[ClarkKent]

I just tried even to delete the file gain.txt but it does not help at all. And the reason is simple - some samples are really really low (e.g. the main title music from IJ) and others are normal loud. I think this could only be solved if the value in the gain.txt is being recognized. As far as I investigated it seems the higher the value the louder the sample is played back. But I asked an user that made some extraordinary work for PinSound - let's see if he can help...

[ClarkKent]

I found something very very interesting and I think it should help! I seems you have to also use the ducking setting file for full compatibility. Can you consider this for the next build?
 
About the gain:
 
dB actually means nothing. It's always how it sounds in the mix. Hence why preamps usually don't show a dB setting on the pots. You shouldn't need to boost anything in Audacity. That's what you use the PinsoundStudio for. It has knobs for that. Boosting a sound by 6dB's could easily mean it's clipping now. Not so good. You can also use the Gain.txt. 11 means it's boosted by 1. 10 is the standart setting. You can go up to 20.
 
About the ducking settings:
 
To be fair the ducking settings are not documented as Pinsound wanted to fix these settings. I begged them to keep them open for changes and they were kind enough to leave that option in. 
The ducking setting is indeed in a config.ini file in the root folder.
It looks like this: 
ducking_jingle = 10
ducking_sfx = 80
ducking_voice = 65
This means that a jingle will duck the music by 90%, the sfx will duck the music by 20% and the voice will duck the music by 35%. Setting the voice ducking to 50 for example will make the ducking way more obvious and the voice shouls be easier to understand - but that's also a process that goes hand in hand with the volume and can be a lot of finetuning to get right. 

Edited by ClarkKent, 13 June 2015 - 09:05 AM.

[toxie]

I know about all of that already. The gain.txt is also recognized, but as said, somehow it doesn't seem to match the PinSound units (yet).

Same goes for ducking.txt. I already have experimental code for this, but nothing that is stable enough.

 

I have to go back to VP10 land for a while to fix some issues there, but will come back to this later-on (thats why its still called experimental version ).

Also i'm working with the PinSound guys directly on a better solution long-term.

[The Loafer]

Can any VPM guru look at the Atari generation 2 emulation and/or the implementation of the atari2.vbs and/or how the dipswitch's are setup? I noticed that both Hercules and superman pinball tables have the same issue: although the state is saved ok (ie: your last score is saved when you exit VP and reload the table again), it seems impossible to win credits for beating any high score you preset, no matter what the dipswitch's are set to. As well, there appears to be no way to get the multiple credits if you beat the highest score.

Long time superman fan here who is getting frustrated with this bug and when I tested the Hercules vp10 table, I noticed the same issue!

Thanks in advance,

Rob

[toxie]

Phew.. Try contacting gaston about it, he doesn't read the forum that much i think..

[Gaston]

Hi Rob,

 

the replay works fine here. Make sure you don't have the free game dip switch set (no credits will be awarded then of course), and that the NVRAM starts out clean, meaning you will have to reset the scores to beat first before setting them using the right coin chute contact!

 

Cheers,

Gerrit "Gaston" Volkenborn

 

PS: I'm a Superman owner myself!

Edited by Gaston, 16 June 2015 - 06:52 AM.

[The Loafer]

Hi Gaston!

That is odd, I have tried with a clean nvram (I deleted it), set the scores manually and It is certainly not set on free play as I must add credits to play. Same thing happens on Hercules except I think that had had preset scores, I didn't need to adjust it.

I will adjust this, thanks so much for looking into it, super appreciated!

Gasp!: you own a superman? Lucky guy!
Hi,

One more question: mould you post a pic of your dip switch settings to ensure there's nothing else interfering? I can then delete the nvram and start from scratch.

Rob

[mjr]

And another thing which happens very rare but from time to time (mentioned on the other place also) on STTNG and Indy (Williams) after a ball drained it set´s the sound to a very low level when the next ball is released. When i play this ball with the lower sound and loose it with the release of the next ball all is ok and the sound becomes it old level. 

 

Sorry for taking so long to get to this - I've been obsessively working on another project and couldn't rip myself away.  I finally looked into this and tracked down the problem; just checked the fix into the source control system.  

 

(The root cause was a really indirect side effect of the DCS speedup code I added for those games.  The speedup was throwing off the MAME virtual time keeping, which in turn threw off the virtual interrupt handling.  The main game ROM in those games sends a reset command to the sound board at the start of every ball, and then sends a "set volume" command shortly thereafter to restore the sound level setting - necessary because the default volume in the sound board after booting is very quiet.  The interrupt timing glitch I introduced caused those "set volume" commands to be randomly lost some of the time.  That's why it got so quiet:  without the "set volume" command, you got the default reset volume.  The sequence would repeat at the start of the next ball, so that would usually fix it when that new set volume command got through properly.  The fix I just checked in should take care of it by cooperating properly with the MAME time keeping.)

[Talantyyr]

Great! Thanks for that fix!

 

 

I was experiencing exactly the same today and was about to make a thread for it cause i didn't know why the sound was so quiet. Strangely, when i opened my coin door and pressed the volume+ button the dmd said 30 (maximum) and the sound was working again as before.

Edited by Talantyyr, 18 June 2015 - 06:08 PM.

[mjr]

Great! Thanks for that fix!

 

I was experiencing exactly the same today and was about to make a thread for it cause i didn't know why the sound was so quiet. Strangely, when i opened my coin door and pressed the volume+ button the dmd said 30 (maximum) and the sound was working again as before.

 

Not strange at all, actually - it's consistent with the bug.  There are two CPUs in the system: the 6809 that runs the game mechanics, and a DSP chip that runs the sound board.  The two communicate via a little internal network.  The actual physical volume level that you hear on the speakers is controlled by the DSP chip, but the 6809 controls the user interface (the coin door buttons) and the NVRAM that remembers the setting.  When the machine boots up, the 6809 sends the DSP a "set volume" command on the network to restore the saved setting.  Likewise, whenever you change the volume via the coin door buttons, the 6809 sends the DSP another "set volume" command to tell it the new setting.  

 

The bug that was causing the sudden volume drop was the result of those command messages getting lost randomly, maybe about 1 time in 5.  The reason it happened at the start of a new ball is that the game ROM on the 6809 resets the DSP chip at the start of every ball.  That makes the DSP revert to its boot-up volume level, which is really quiet.  Right after sending the reset signal, the 6809 sends a "set volume" command to the DSP to set the volume back the way it was.  You don't normally notice this at all because it happens very quickly (a fraction of a second).  Whenever the bug occurred randomly, the "set volume" got lost, so the DSP stayed at its reset volume.   Which brings us to your observation:  if you open the coin door and press a volume button, that sends yet another "set volume" command from the 6809 to the DSP, restoring the volume to its proper level.  The 6809 never forgot the right volume level - it was just that the original message it tried to send the DSP got lost.  Anyway, with the fix, the messages should never get lost in the first place.

[toxie]

The first post has been updated with the new build(s)..

 

It includes the volume fix by mjr, plus support for GNR 3.00 Dutch, plus some optimizations by mjr to the whitestar II/stern emulation (there is even much more awesome stuff to come from him though )

 

EDIT: oh, and plus improved speech emulation on Jeutel and Nuova Bell..

Edited by toxie, 19 June 2015 - 05:22 AM.

[freneticamnesic]

So mjr is some kind of white knight? he comes in with this amazing plunger solution, then he creates a brand new VP plunger, and now he's fixing and improving a bunch of VPM stuff? 

Thank you mjr, your contributions are very appreciated!

[mjr]

Thanks, but my contributions are puny in the scheme of this huge project.  Toxie and Fuzzel have been doing the heavy lifting that makes the whole thing possible.  It's great that they've been so welcoming!

[ClarkKent]

Anything new about the gain.txt volume control of PinSound?