65 replies to this topic

[JAM0]

OK Jimmy, I'll test out the original table to see if it behaves that way during castle multiball and report back to you sometime today.

[RipleYYY]

not sure if i undertsand well, but about the castle multiball, you have to lock balls (1 or 2 or more if combined with others multiball) in the castle again to enable for few seconds the big jackpot X1 X2 X3 in the left saucer (near the bumpers)

[OuchTilt]

Just played a couple of hours on this, looks fantastic and no impact on speed. Well done Mr Fingers.

Thanks also to JP for a great table.

[JAM0]

not sure if i undertsand well, but about the castle multiball, you have to lock balls (1 or 2 or more if combined with others multiball) in the castle again to enable for few seconds the big jackpot X1 X2 X3 in the left saucer (near the bumpers)

Yeah I think you are right, I just played JP's original version and it appeared to do the same thing. Sorry for the false alarm Jimmy.

[Dezwin]

First - thank you so much for your work on this mod!

I'm finding that I also have to reduce my max textures to 1024 to get this table to load.

What does this mean? Does having a higher texture setting require more RAM
on the video card? Or is it more complex than that?

Has it been explained somewhere in these forums exactly what this setting does, and what affect it has on visual quality?

My video card has 2 Gigs RAM, so if it is only RAM related, I'm surprised it won't load.
(I am using an EVGA GTX560 2G video card for the play field)

Since this appears to be a "global" setting - will I be losing quality on other tables by leaving it at 1024?

[JAM0]

There might be something else going on for you as that card should have double of what it needs to run this at unlimited textures. How much ram do you have on the computer itself?

Yes, bumping it down to 1024 will effect the resolution on all your tables that have higher resolution than that.

[JAM0]

Jimmy, I do think there is something amiss with castle multiball. Just played it again and was in castle multiball and shot left ramp 3 times. I was left with 0 balls but the castle miltiball light was still on meaning I should have at least two balls to play. It did a ball search and then added a ball to the shooter lane (this happened 2 times). Then that ball drained and I was left with 0 balls again when I should have had one. This time the game froze as it didn't give me a ball in the shooter lane. What I think is happening is on JP's original, it in some circumstances diverts a ball from the left ramp into the plunger lane whereas yours, in the same circumstance does not but rather locks it. It is most easily noticed when you only have castle multiball alone running and you continually shoot the left ramp.

[jhoward1082]

I threw this same exact card into my cab just to test if bumping up from 1gb to 2gb of video memory would make a difference and still ran into the same issues. I tried multiple drivers as well. I have 4gb of system memory, running Windows 7 64-bit edition. It's interesting that some people seem to be able to run the table flawlessly.

First - thank you so much for your work on this mod!

I'm finding that I also have to reduce my max textures to 1024 to get this table to load.

What does this mean? Does having a higher texture setting require more RAM
on the video card? Or is it more complex than that?

Has it been explained somewhere in these forums exactly what this setting does, and what affect it has on visual quality?

My video card has 2 Gigs RAM, so if it is only RAM related, I'm surprised it won't load.
(I am using an EVGA GTX560 2G video card for the play field)

Since this appears to be a "global" setting - will I be losing quality on other tables by leaving it at 1024?

[jimmyfingers]

Jimmy, I do think there is something amiss with castle multiball. Just played it again and was in castle multiball and shot left ramp 3 times. I was left with 0 balls but the castle miltiball light was still on meaning I should have at least two balls to play. It did a ball search and then added a ball to the shooter lane (this happened 2 times). Then that ball drained and I was left with 0 balls again when I should have had one. This time the game froze as it didn't give me a ball in the shooter lane. What I think is happening is on JP's original, it in some circumstances diverts a ball from the left ramp into the plunger lane whereas yours, in the same circumstance does not but rather locks it. It is most easily noticed when you only have castle multiball alone running and you continually shoot the left ramp.

I think I know what it is and it certainly doesn't seem right to keep locking down to 0 balls then do a ball search. Thanks for staying determined on this issue, as although no one else reported it, i do believe that you have found a legitimate problem and it's likely caused by a light layer ramp being slightly higher than the ramp with the switch and it is affecting the ball trigger. However, if that was the case it I thought it would be missing all sorts of other things too from failing to trigger but maybe the other more common events are controlled by the switch further up and around (after the apex) and this one before specifically for situations like the lock where it has to make a change before the other more "active" switch.

I'll take a look into it and hopefully can get it hammered down and and update out as soon as possible (i.e. tonight or tomorrow).

[unclewilly]

If your light layers aren't collidable they shouldn't effect the triggers

[jimmyfingers]

I threw this same exact card into my cab just to test if bumping up from 1gb to 2gb of video memory would make a difference and still ran into the same issues. I tried multiple drivers as well. I have 4gb of system memory, running Windows 7 64-bit edition. It's interesting that some people seem to be able to run the table flawlessly.

First - thank you so much for your work on this mod!

I'm finding that I also have to reduce my max textures to 1024 to get this table to load.

What does this mean? Does having a higher texture setting require more RAM
on the video card? Or is it more complex than that?

Has it been explained somewhere in these forums exactly what this setting does, and what affect it has on visual quality?

My video card has 2 Gigs RAM, so if it is only RAM related, I'm surprised it won't load.
(I am using an EVGA GTX560 2G video card for the play field)

Since this appears to be a "global" setting - will I be losing quality on other tables by leaving it at 1024?

I too have the EVGA GTX560 (but Ti and 1G) and I do have problems with intermittent loading, but, it's intermittent and will work on average 2 out of 3 times. That alone says to me there's an "X" factor that is something just beyond memory for the card (or the computer itself) or at least the sequence in which things are loaded upon playing / rendering the table. I actually am surprised by the amount of trouble I had with this card on my main system and the two GI8 mods that more people didn't have a problem so I'm also inclined to believe that the 560s are adding something else into the mix as there seems to be a couple other posts about them being a bit quirky.

All I can say is to do what I do and just try a few times to load but if it's stuburn, change the resolution down to something like 1600 x 900, run it, then turn it back up, or up the alpha scale, in general mess around with changed in the video section, restart just to be safe, and try again.

To more permanantly resolve, export some plastic textures and resize to 1024 (I would use the "middle" level layers like 1, 3 and 5). If that doesn't work then you could export some of the middle PF light levels as well and resize to 1024.

I've left the table at high detail to attempt to get the most as I did not see a point of having it hard coded down to 1024 for the main release and have it be downgraded for the people that would have been able to run it fine. I even left it at a stage that can be a bit of a pain to me as I prefer the try again technique and getting it too load and look as good as possible then reducing a bunch of textures forcibly to 1024 (also becuase of the generally good feedback from the SS GI8 leading me to beleive I was in the minority of users - which I still think I / we are). Mainly because the option now exists for people to choose 1024 if they're having problems and the people who aren't can leave at 2048 or unlimited and enjoy the detail available to the table / lighting.

Good luck with it and let me know what happens when you try to change or lower the resolution then up it back up or simply how many times you've tried in a row. By the way, I also use Windows 7 but 32 bit and am at driver level 270.61

[unclewilly]

What kind of trigger is it. I've had issues with custom shaped triggers not registering.

[jimmyfingers]

If your light layers aren't collidable they shouldn't effect the triggers

Yah I know, but one ramp I did as a base was still collidable and slightly higher than the ramp for which the trigger was on (I did this as part of the carving up of the ramp to keep the lit section separate and effictiently running being only a set of square alpha ramp objects (but had to rejoin at the top). I also was trying to work around the ball graphical glitch that can be seen when a ball is rolling on an alpha ramp.

What I don't get though is the the trigger in question is actually set to a ramp that itself is a non-collidable ramp. I thought that would make it not able to be triggered? Regardless, of that nuance, it can't be the issue on it's own because that's the same as the original 1.04 release and I can totally see that my new ramp base object must be glazing slightly over the trigger in question. Still confused by that fact that the trigger is surfaced on the a non-collidable ramp

Edited by jimmyfingers, 19 January 2012 - 01:24 AM.

[jimmyfingers]

I just tested the table and it's not that switch as it detects it fine plus the part I thought I glazed over was done so in with another identical shapped / defined ramp in the original version. So, I don't know at this point and my original thought that I had found a ramp higher and collidable is not valid as the same situation exists in the "real" version. I don't get it still though with that trigger being on a non-collidable. Seems like that could still be relevant.

I'm going to have to try and re-produce the castle multi-ball to see the issue JAM0 is talking about and make sure I can even reproduce that as I really don't see at this point where the problem could be or why / how it could differ between this and the base version.

[jimmyfingers]

Jimmy, I do think there is something amiss with castle multiball. Just played it again and was in castle multiball and shot left ramp 3 times. I was left with 0 balls but the castle miltiball light was still on meaning I should have at least two balls to play. It did a ball search and then added a ball to the shooter lane (this happened 2 times). Then that ball drained and I was left with 0 balls again when I should have had one. This time the game froze as it didn't give me a ball in the shooter lane. What I think is happening is on JP's original, it in some circumstances diverts a ball from the left ramp into the plunger lane whereas yours, in the same circumstance does not but rather locks it. It is most easily noticed when you only have castle multiball alone running and you continually shoot the left ramp.

JAM0, I have tested this with some blockers and some debug switch hits - it does not give me any problem on my set-up. Blocking off the myst hole and the coffin ramp / lock (as well as the outlanes and the middle) I tested and got to castle multi-ball several times. Then after I'm in castle multi-ball only First ball in same lock after, introduced jackpot, 2nd gave me double jackpot, 3rd, just released it out after about second back to the trough. In fact it hung on to it so little and the timers working fine for couting down the jackpots from 20 that it was quite difficult at times to get 2 or 3 balls in there at once. Once the timers count down then they all came out regardless (either 2 from double jackpot mode or one from single). If I was at single and then got to double the counter went back up to 20 and I had a 20 count to get the 3rd one up there. On a couple occasions I got it up there 2 or 3 times in that period even as tough as it was in the window and never saw the issue.

Aftter testing several times with my current set-up I further re-downloaded the table, and the ROM and over-wrote my test area - still worked fine. I also wiped the NVRAM and re-applied the table download and rom - still fine.

Long and short, it is not a problem on this side and nothing I can do can reproduce what you're referring to. So, please try and re-download (do not change any settings at all) and wipe your NVRAM (save your original if you want first for high scores and what not) then try again. Put some blockers up over the areas I mention and if you want to get the switches hit quicker use the debug window and enter in the lines "vpmTimer.PulseSw 86", "vpmTimer.PulseSw 87", and "vpmTimer.PulseSw 88" (one per line). Go to one line at a time and hit enter at far right, then go back to the table and you'll see the corresponding light hit (86 and 88 are the edge yellows and 87 is the middle that you need to do last). You can get to castle multi-ball quicker that way but will still have to do the shots, which it sounds like your good at.

In any case, It's not a problem here and works fine on my system in the scenario you describe so must be something else in combination on your end. Maybe a change in physics could affect the way a kicker or something works but fresh download of the table and ROM, and NVRAM wipe saw no issues and ball lock in castle-multi ball behaved correctly and released every time the 3rd ball went in while the countdown was still on (when the count down was over they all came out). The trigger also earlier in question worked unwaveringly and produced the sound I assigned to it every single time the ball went over it.

[JAM0]

Thanks Jimmy for testing that. I'm thinking what possibly happened is that I got a ball stuck somewhere during multiball and I didn't notice.

One question: how do you wipe nvram?

[bladexdsl]

One question: how do you wipe nvram?

open vpinmame\nvram folder eg: C:\vpinmame\nvram

find the file drac_l1 and delete it.

Edited by bladexdsl, 19 January 2012 - 09:01 AM.

[JAM0]

thanks bladexdsl.

I just played the game again without changing anything and it worked perfectly. I think what happened the last time is a ball got stuck somewhere unbeknownst to me. That seems like that would account for the symptoms I had.

[jhoward1082]

Thanks for your help Jimmy. I ended up just taking the plunge and threw XP 32-bit on my cab and the table runs flawlessly now every time. A couple of other tables that would sporadically give me issues now run perfectly as well. Since you're running Windows 7 64-bit and have sporadic issues as well I'm willing to bet that (and possibly some combination of nvidia geforce 400 and 500 series cards/drivers) are the culprit.

I threw this same exact card into my cab just to test if bumping up from 1gb to 2gb of video memory would make a difference and still ran into the same issues. I tried multiple drivers as well. I have 4gb of system memory, running Windows 7 64-bit edition. It's interesting that some people seem to be able to run the table flawlessly.

First - thank you so much for your work on this mod!

I'm finding that I also have to reduce my max textures to 1024 to get this table to load.

What does this mean? Does having a higher texture setting require more RAM
on the video card? Or is it more complex than that?

Has it been explained somewhere in these forums exactly what this setting does, and what affect it has on visual quality?

My video card has 2 Gigs RAM, so if it is only RAM related, I'm surprised it won't load.
(I am using an EVGA GTX560 2G video card for the play field)

Since this appears to be a "global" setting - will I be losing quality on other tables by leaving it at 1024?

I too have the EVGA GTX560 (but Ti and 1G) and I do have problems with intermittent loading, but, it's intermittent and will work on average 2 out of 3 times. That alone says to me there's an "X" factor that is something just beyond memory for the card (or the computer itself) or at least the sequence in which things are loaded upon playing / rendering the table. I actually am surprised by the amount of trouble I had with this card on my main system and the two GI8 mods that more people didn't have a problem so I'm also inclined to believe that the 560s are adding something else into the mix as there seems to be a couple other posts about them being a bit quirky.

All I can say is to do what I do and just try a few times to load but if it's stuburn, change the resolution down to something like 1600 x 900, run it, then turn it back up, or up the alpha scale, in general mess around with changed in the video section, restart just to be safe, and try again.

To more permanantly resolve, export some plastic textures and resize to 1024 (I would use the "middle" level layers like 1, 3 and 5). If that doesn't work then you could export some of the middle PF light levels as well and resize to 1024.

I've left the table at high detail to attempt to get the most as I did not see a point of having it hard coded down to 1024 for the main release and have it be downgraded for the people that would have been able to run it fine. I even left it at a stage that can be a bit of a pain to me as I prefer the try again technique and getting it too load and look as good as possible then reducing a bunch of textures forcibly to 1024 (also becuase of the generally good feedback from the SS GI8 leading me to beleive I was in the minority of users - which I still think I / we are). Mainly because the option now exists for people to choose 1024 if they're having problems and the people who aren't can leave at 2048 or unlimited and enjoy the detail available to the table / lighting.

Good luck with it and let me know what happens when you try to change or lower the resolution then up it back up or simply how many times you've tried in a row. By the way, I also use Windows 7 but 32 bit and am at driver level 270.61

[JAM0]

Glad to hear it jhoward!