3889 replies to this topic

[fuzzel]

Ok did a quick look. The changes should be correct.

 

@hauntedfreaks: what did you exactly do and on which table? To play a sample on the backglass device you have to tag a sample in the sound manager "Toggle BG Out". Doing so will close the sound manager and when you reopen it you will see a "*BG*" in front of the import path. Then save the table, close it and reload the table again because all samples will be loaded to the direct sound device while loading the entire table.

[ClarkKent]

I still think there is something wrong with VP or the new VBS scripts. It happens often that suddenly the table does not react normal - and after quitting to Windows everything behaves as I hold the mouse button clicked or double clicked (if I close a window, it's opened again immediately). If I'm not wrong this did not happen before the new VBS changes.

 

And in Kingpin it does not react to my plunger after the first ball (sometimes only on the first one) - but pressing Enter on the keyboard shoots out the ball every time. Weird. And changing the kicker in the Hotel to 0,7 the ball vanishes even faster than before - no animation at all. But with the regular setting 0,3 it's stuck at the kicker.

Edited by ClarkKent, 08 June 2015 - 03:52 PM.

[Talantyyr]

Hey clark, have you checked against an older version of VP and Kingpin? Just to make sure it's VP and not your cab computer

[ClarkKent]

Unfortunately I do not have older versions of VPX and PinMAME anymore - is there an archive with all versions?

[Outhere]

https://www.dropbox....ev1907.zip?dl=0

https://www.dropbox....ev1920.zip?dl=0

https://www.dropbox....ev1949.zip?dl=0

https://www.dropbox....ev1953.zip?dl=0

https://www.dropbox....ev1975.zip?dl=0

https://www.dropbox....ev1988.zip?dl=0

https://www.dropbox....ev1993.zip?dl=0

https://www.dropbox....ev2003.zip?dl=0

Edited by Outhere, 08 June 2015 - 05:27 PM.

[Trinity]

Nice!

[hauntfreaks]

Ok did a quick look. The changes should be correct.

 

@hauntedfreaks: what did you exactly do and on which table? To play a sample on the backglass device you have to tag a sample in the sound manager "Toggle BG Out". Doing so will close the sound manager and when you reopen it you will see a "*BG*" in front of the import path. Then save the table, close it and reload the table again because all samples will be loaded to the direct sound device while loading the entire table.

 

thanks for the reply... yes in vp9 I have done that for a couple tables, moving some sounds to the BG.... VPX all the sounds are coming from the BG all the mech sounds, chimes, music....etc. none of them are coming from within the cab like in VP9.... and my audio options in vp9 and in vp10 are set exactly the same....  i'm wondering is something reversed in vpx?.... can someone else vouch this is happening in there cab also? 

[ClarkKent]

I found out two things:

1st - the one with the kicker was "only" the problem because of the changed value from 0,3 to 0,7. Some tables did not react at all or react 10 seconds later on shooting into the kickers - with the new value 0,7 it works. Only in the latest Kingpin table I had to set it to 0,6 because with 0,7 there is no animation of the ball before falling into the hole of Hotel Lex. All VPX tables have to be changed I presume, but that was written in the release anyway.

 

The 2nd one is performance. I noticed a drop in fps after some time of playing. It's interesting that it did not get better quitting VPX and starting it again. But restarting Windows helped. In the moment I noticed the drop in fps I tried several tables - I suddenly saw fps drops to 70 or 80 fps in all VPX tables instead of the constant 120 fps before. It's even more weird that if I use vsync with 60 fps (I set Kingpin to that value for testing) it went down to 30 or less. But, and this is the but - with PM5 tables I always got 120 fps at the same time before resetting Windows. I will watch this for some time and report back if this happens again. Nevertheless I would like to say that this NEVER happened before a VPX beta until the new kickers were included (I think with version 2012 or so).

 

I think it's definitely not because of PinMAME as I tried 2.12 unified and it was the same or didn't see any difference in performance or reaction. I'm using the latest beta now again because of the missing light show in Kingpin. And I also think there is no problem with the scripts as at the moment I can plunge the ball every time at the moment (I hope it will stay this way).

Edited by ClarkKent, 08 June 2015 - 07:47 PM.

[fuzzel]

Thanks ClarkKent. For KingPin keep in mind that there actually two Hotel Lex kickers on top of each other (sw51 is the kicker on the playfield and sw51b is the second below playfield) I suggest to set hit accuracy to 0.6 to sw51 and check Legacy on the sw51b.

 

As for the perfomance: yes there is a still a problem with the collision system and the newest changes. Toxie is working on a fix. I hope once fixed that the absorbing ball issue is also fixed. It seems that the whole collision system has problems since the early versions of VP10.

[ClarkKent]

After writing my last message I ran Kingpin again and let it run without a game for some time. Suddenly the framerate jumped from 60 to 30 to 60 and so on. After some trial and error I switched off the lampshade and it got better, mostly 60 fps. But also while playing performance could go down to 25 to 30 fps - maybe because of the collision system problem you mentioned. But this was not a problem with older versions of VPX (I remember Kingpin ran good with version 2012, no performance problems). Did you change something there, too?

Does anybody still have version 2012 and 2023? Would like to test them again.

[hauntfreaks]

as a comparison, I ran Kingpin for 25-30mins... I was getting an average 200-221 framerate.... running it in build 2044

[leeweatherbee]

VPX_beta_rev2012:https://app.box.com/...lera9ks13nwdtfg

 

VPX_beta_rev2023   https://app.box.com/...agwjbs9hwu0lhlh

[hauntfreaks]

never mind....lol

Edited by hauntfreaks, 08 June 2015 - 08:42 PM.

[hauntfreaks]

I also set the both the lux kickers to 0.5 and getting the animation swirl on near misses....

 

[fuzzel]

After writing my last message I ran Kingpin again and let it run without a game for some time. Suddenly the framerate jumped from 60 to 30 to 60 and so on. After some trial and error I switched off the lampshade and it got better, mostly 60 fps. But also while playing performance could go down to 25 to 30 fps - maybe because of the collision system problem you mentioned. But this was not a problem with older versions of VPX (I remember Kingpin ran good with version 2012, no performance problems). Did you change something there, too?

Does anybody still have version 2012 and 2023? Would like to test them again.

No no changes to the rendering engine only changes to collision and physics handling.

[fuzzel]

The rule of thumb is still valid: kickers which are used as triggers and not visible -> legacy setting.

The hit accuracy for visible kickers where you want a swirl animation can vary depending on the table and the area of the kicker but if the accuracy is too low (like 0.3 mostly is) you will notice that the ball goes into the kicker but fires the hit event much later because the ball's velocity is very low and will sink very slowly into the kicker. The hit accuracy is a factor for the z height of the ball:

Let's say you have a ball with a radius of 25 so the midpoint of the ball is at height 25 (if it's on the playfield) and a kicker on the playfield (height is 0) if you set the hit accuracy to 0.7 the kicker will fire the hit event when the the ball's midpoint is lower/equal 17.5. So 0.7 is just a start value and depending on the playfield's friction and/or surrounding elements the accuracy can be lower but if it's 0.3 the ball's midpoint must reach 7.5 which is very low. You normaly have a higher friction at the kicker bevel (well depending of the kicker type) but the friction will slow down the ball. I have to apply a bit more gravity to the ball in such situation but at the moment the ball will sink slowly into the cup until it's deep enough. Therefore it's ok to play around with the accuracy but don't use too low values

 

Ok did a quick look. The changes should be correct.

 

@hauntedfreaks: what did you exactly do and on which table? To play a sample on the backglass device you have to tag a sample in the sound manager "Toggle BG Out". Doing so will close the sound manager and when you reopen it you will see a "*BG*" in front of the import path. Then save the table, close it and reload the table again because all samples will be loaded to the direct sound device while loading the entire table.

 

thanks for the reply... yes in vp9 I have done that for a couple tables, moving some sounds to the BG.... VPX all the sounds are coming from the BG all the mech sounds, chimes, music....etc. none of them are coming from within the cab like in VP9.... and my audio options in vp9 and in vp10 are set exactly the same....  i'm wondering is something reversed in vpx?.... can someone else vouch this is happening in there cab also? 

 

I can try that on my cab later. If you still have a revision before 1940 could you try it with the default table if it happens there too? If not it's ok I can test it here also but it would speed up searching

[KieferSkunk]

I'm just curious: Does this "hit accuracy" value really mean what you think it means?  It sounds more like that's a "switch height" setting - the point at which the kicker's switch is triggered by the ball resting on it, and as described, it implies that the switch offers no resistance to gravity.  I would think this would be very inexpensive to simulate.  I would also think that if you set this value too high, you could trigger the "hit" event and still have the ball fall out of the kicker again on its own - this does happen in real machines, and it could be good to account for that if you wanted to, say, reproduce a design flaw.

 

But in any event, I'd be interested to know why a ball necessarily "falls very slowly" into a kicker.  That sounds more like a physics bug than a desired behavior.  The physics simulator could, of course, simulate a case where a ball rolls around the rim of the kicker and "eases" into the kicker - in that case, a slow hit event would be understandable and expected.  But if the ball falls straight into it, I would think this would take no more than a few frames of animation time at most.

[fuzzel]

Hehe yes it is a physics bug. As I said I have to apply a bit more gravity in this situation. But as long as we have overall collision/physics problems at the moment I stop hacking around on the kickers and improve them later.

The hit accuracy at the moment isn't 100% correct because the velocity isn't considered here. If you set the accuracy to 0.99 the event will be fired and the ball is grabbed by the kicker very early. It shouldn't be much difference to a legacy kicker then. The old kicker hadn't any physics behavior at all. There was only a hit cylinder to detect the collision and if hit the ball was grabbed like a magnet would do. So it's much better now but not perfect if you want to simulate design flaws

Edited by fuzzel, 08 June 2015 - 09:50 PM.

[hauntfreaks]

I can try that on my cab later. If you still have a revision before 1940 could you try it with the default table if it happens there too? If not it's ok I can test it here also but it would speed up searching

 

 

ok.. "my" audio issue is resolved... I went into the envidia digital audio driver it had had only one of the digital outs allocated (BG audio) I enabled both of them  ... and problem solved   sorry for wasting anyones time   resume to your previous scheduled program... 

[cyberpez]

The new kicker on a wall seems to be working better.  I did have to crank the hit accuracy up to .95.  At .9 it took a good 5-10 seconds to catch the ball.  With lower setting it would sit in the kicker and slowly sink until it fell through the wall (15-20 seconds)