118 replies to this topic

[jimmyfingers]

Noah, you're recent updates to Gorgar and Firepower I think are some of the best playing / physics settings for VP tables (at least VP9 which is the only version that I've used). After assessing both tables to see what it was vs. others to seemingly give a more realistic playing speed and help give the ball some weight it seems that your higher friction (.0040) was a key part. It also seems that keeping the gravity high and the slope not too high helped to still allow the ball to gain some accelleration when rolling down the table vs. being a bit too floaty. The other last aspect I noticed on both these tables was the setting for "MaxBallSpeed" in the script and on each was set at about 45 for each. Correct me if I'm wrong, but I thought that setting was the same as the dampening friction variable? In any case on the tables I've adjusted to be closer to these two of yours, I now set that string in the script just in case it somehow is different. But I am confused with you saying that you prefer no dampening speed when this value (at least as MaxBallSpeed) seems to be one of the key factors in these last two of your table updates and one in which seems to help a great deal in making these tables play as nice / realistic as they do.

While we are on the topic of physics, I had thought of starting a topic for a couple months now on one thing that I've spent hours on trying to tweak and that I still see missing in VP vs. real life in how the ball does seem to lack some weight. While studying real pinball / videos it seems that in real life the ball can move quite slowly, yet when it reaches it's apex or head straight down the table, the rate of acceleration can be quite high and visually perceivable but at the same time can take a fairly long time going up the table to slow down, come to a "stop, then head back down the table with accelerating speed. In VP, this seems tough to reproduce as if set to try and mimic the slow decrease in speed, then the increase back down the table is not fast enough and seems floaty, yet when the setting are manipulated to give a decent rate down the table, the ball then transitions too quickly from and upward motion to a stop to a downward motiion.

The other thing that seemingly isn't as easy to set up to mimic real life is the lateral motion and the amount of energy the ball seems to retain and carry with it after hitting an object. Outside of setting elasticity levels higher on all rubbers and colidable objects, it just seems that the ball mass itself is not high enough or any different (or different enough) from the other objects / wall masses where as in real life the steel ball vs. the rubbers have much different densities and react accordingly. I see this momentum aspect on the difference in how much more lateral a ball travels and keeps it's energy on a real life table (especially watching EM tables). It's vary tough to get this lateral feel in VP and setting higher elasticity levels help but higher slope and gravity counteract it as they draw the ball down which is a dilema as we need those higher so they don't leave the ball floaty.

Cupid asked way back, after the ball sizing option was added, whether a ball weight option should be included. I full heartedly think this would be a fantastic option and one of the last missing links in being able to set-up VP tables to very close to real-life. Being able to set the ball density / weight differenlty I think would give us a necessary extra element to work with along with the other settings and help get over, or at least mitigate, some of these obersvations I (and others) have had regarding ball weight / momentum and all the things we are trying to do to mimic this. Being able to set it differently then other objects, which is what I beleive was proposed by cupid, would be ideal as I think in real life it is the mass / density of the stell ball vs. the woods, rails, and especially the rubbers that gives it it's momentum and complete feel of what it does in contrast to the objects it hits.

Lastly, on these observations, ball weight, and lateral movement goals with still maintaining slow speed (higher friction) and decent downward acceleration, I have experimented with and become as happy as I've been with the physics by taking Noah's VP Firepower and Gorgar settings as a base and modifying slightly in this manner: Increasing gravity higher, lowering slope, and increasing elasticity. I've been setting the gravity at about 1.75 or 1.775 and decreasing the slope like on Gorgar from 5.8 to about 5 with a slightly less friction number .375 (sometimes till using .004 when gravity is up around 1.775). This higher gravity / lower slope seems to help the ball still travel laterally somewhat better as the slope is lower but when the ball is travelling down / starting to travel down does still have some decent acceleration - the higher friction is helping from moving too fast and also aiding in a slower transition from up to down. Gravity and slope are definitely not the same thing but the difference can be subtle and hard to see without lot's of tests just basing it visually on things. The last caveat to all these aspects to help the ball feel more weighty and with higher mass and energy that it maintains after hitting an object is that the elasticity of most playfield collidable objects has to be increased. I've seen previous topics that have similar concepts on VP tables tending to be a bit too fast compared to real life where they potentially should be slower but with more "bounce" and ability to accelerate down. I have seen with the settings I used above that modifying the top of sling shots and other mainly upward facing rubbers to about .8 is good (sometimes even higher), and most other rubbers between .7 and .8. Also changing woods and metal rails to about .45 to .55 elasticity.

This is a good topic for sure but I feel some of the tweaks, at least for myself, go much more further than just the flipper settings and get more complicated and difficult to easily / consicely relay in how to modify /edit all the necessary elements of a VP table to acheive especially as it varies greatly from table to table. Overall, all information from flipper settings to slope / gravity / friction to elasticity to general aspects all can definity be used to get some of the VP tables quite close to playing like their real life version. Thanks to all who add what they can and for their experiences with settings that help these recreations become more true-to-life.

Edited by jimmyfingers, 21 December 2011 - 04:38 PM.

[chas]

The dampening speed affects the ball velocity, not objects on the playfield.>>>>>>>>>>>
it did dampen objects like pops.
it did seem to dampen the entire table gameplay.
thats how it worked on my single core PC.
did not seem to work for just 1 object.
click on it.
no dampen option.
there is no option in the editor for dampening a certain object.
the only option i found for dampening is in physics with table slope and GDL etc.......

>>>>>>>>>This is interesting. I never looked into this setting. You explained well what it does, but you didn't give any numbers. By lowering it, what number/setting do you choose to use there? Then, by how much do you increase strength? I want to know more about this cuz it sounds like it just might cure some of the remaining issues I have with VP flippers.>>>>>>>>>>
ok,
i did spend a little time on this.
EX;
with strength on 5, speed on 0.075 plaw 2.
if you bump down the plaw to 0 it will be equal to about 0.065/0.0675 with plaw on 2.
it seemed to vary a little on tables.
it looks like the plaw will multiplay 1-2x.
in the user guide plaw of 4 is maximum.
plaw 4 did not always seem to work.
i leave the plaw on 2.
this is how it worked for me on my desktop single core.
maybe it will be different or have better results on a computer with a better system.

Edited by chas, 21 December 2011 - 07:17 PM.

[analog999]

Here's my question to anyone familiar with the VP source - whose calculations are these physics settings based off? There are too many settings to "compensate" for this problem or that problem... there shouldn't be any compensation necessary.

The idea that all these tables vary so much just blows me away. I know they are commercial, but I have played much more realistic tables, like Dino Madness & Crazy Castle. They run on my 1ghz android phone in full 3D, and they have better physics than both VP & FP. Also, the Pinball Hall of Fame series is very very good, on a wii. Most of us here are running computers that can emulate the wii! Something is wrong with this whole picture...

There should be no stuttering in any of these games, and the physics should be easier for table creators to manage. I think the whole thing needs a makeover, a new code branch to experiment with some new ideas. Of course, it needs to support existing tables.

Who else agrees with these statements?

Edited by analog999, 21 December 2011 - 08:37 PM.

[rosve]

Here's my question to anyone familiar with the VP source - whose calculations are these physics settings based off? There are too many settings to "compensate" for this problem or that problem... there shouldn't be any compensation necessary.

Maybe the whole things needs to be revamped with NVidia PhysX (if available for nvidia cards) or some other open source physics engine.

Well, I'm working for a company that is making one of the most accurate physics simulation software there is. We have more than 100 parameters to play with to tune the behaviour of an articulated robot arm. In my opinion VP has too few parameters.

Edited by rosve, 21 December 2011 - 08:17 PM.

[Bob5453]

There should be no stuttering in any of these games, and the physics should be easier for table creators to manage. I think the whole thing needs a makeover, a new code branch to experiment with some new ideas. Of course, it needs to support existing tables.

Who else agrees with these statements?

We all agree. The source code is available. Try to have this fixed for us by christmas. There seems to be no one actively working on it now, so knock yourself out. Link to source code below. Seriously, contact Destruk if anyone thinks they can add something worthwhile to Visual Pinball. Suggestions are great, but there is no one listening to your suggestions.

http://vpinball.svn....trunk/?view=log

[analog999]

That's crazy. There doesn't need to be 100 parameters for a pinball simulation. These are the settings I think you should have:

Flipper Mech (williams, DE/stern/sega, bally, & versions of each to support various coil stops)
Coil type (have a model for each coil type already preset, ie FL-11629 or FL-116230)
Flipper recoil speed (ie how fast flipper drops back to its original position, ie is there a spring)
Flipper type (standard williams, lightning flipper, stern, etc etc)
Flipper Rubber type (soft, medium or hard)
Table slope (1 setting, not a min and a max)

After those settings had been nailed down, it would create a standard for the table designers, and they could use those settings to create their tables and forget about the rest of the work. Yes, maybe the internal programming would have a lot of parameters (not a ton would be necessary, IMO), but those would not be exposed to the table designers. That would be just like in real life! Steve Richie picks out the options above, he doesn't determine the little nuances of each flipper physics setting on the table he's creating.

[rosve]

That's crazy. There doesn't need to be 100 parameters for a pinball simulation. These are the settings I think you should have:

Flipper Mech (williams, DE/stern/sega, bally, & versions of each to support various coil stops)
Coil type (have a model for each coil type already preset, ie FL-11629 or FL-116230)
Flipper recoil speed (ie how fast flipper drops back to its original position, ie is there a spring)
Flipper type (standard williams, lightning flipper, stern, etc etc)
Flipper Rubber type (soft, medium or hard)
Table slope (1 setting, not a min and a max)

After those settings had been nailed down, it would create a standard for the table designers, and they could use those settings to create their tables and forget about the rest of the work. Yes, maybe the internal programming would have a lot of parameters (not a ton would be necessary, IMO), but those would not be exposed to the table designers. That would be just like in real life! Steve Richie picks out the options above, he doesn't determine the little nuances of each flipper physics setting on the table he's creating.

"Steve Richie picks out the options above, he doesn't determine the little nuances of each flipper physics setting on the table he's creating."
Steve doesn't have to emulate the hardware he is using. I can easily think of at least 5 parameters just to describe how a solenoid works.

[analog999]

We all agree. The source code is available. Try to have this fixed for us by christmas. There seems to be no one actively working on it now, so knock yourself out. Link to source code below. Seriously, contact Destruk if anyone thinks they can add something worthwhile to Visual Pinball. Suggestions are great, but there is no one listening to your suggestions.

http://vpinball.svn....trunk/?view=log

Yes I am aware of the open source. It's about time for me - I am capable. Right now business is slow, and my business has to be my top priority when it comes to developing new stuff. I'm lucky that I get to work in front of a computer and have some time throughout the day to read the forums (lately). But, there is a burning desire for me to do this.

Btw, anyone know why isn't 9.1.2 up on SF?

"Steve Richie picks out the options above, he doesn't determine the little nuances of each flipper physics setting on the table he's creating."
Steve doesn't have to emulate the hardware he is using. I can easily think of at least 5 parameters just to describe how a solenoid works.

Obviously. That was my point. He doesn't have to worry about those nuances, the mechanical engineers do. Table designers here shouldn't either. The mechanical engineers of VP are the source code editors, not the table designers.

And, guess what, there aren't a zillion different types of flipper mechs. So once the design and parameters are down for one, your'e done and everyone can use that mech type in their table if they choose.

Edited by analog999, 21 December 2011 - 08:56 PM.

[Bob5453]

Btw, anyone know why isn't 9.1.2 up on SF?

http://www.vpforums....s...st&p=135034

[analog999]

I feel like I just took this thread off track. I'm sorry for that. The original idea of the thread was to post settings for tables to improve existing ones. Yes it would be great to improve the engine, but for the moment, we work with what we've got. I just messed with settings for Flash Gordon (the 80's one) yesterday. I'll post those soon.

As for jimmyfingers' comments, I think what you're saying is important, but the biggest flaw that most tables have are the flippers & the general table settings, IMO. That's why I'm posting those results. If you have changes to tables that you want to note, I'll be happy to add them to the first post.

[rosve]

We all agree. The source code is available. Try to have this fixed for us by christmas. There seems to be no one actively working on it now, so knock yourself out. Link to source code below. Seriously, contact Destruk if anyone thinks they can add something worthwhile to Visual Pinball. Suggestions are great, but there is no one listening to your suggestions.

http://vpinball.svn....trunk/?view=log

Yes I am aware of the open source. It's about time for me - I am capable. Right now business is slow, and my business has to be my top priority when it comes to developing new stuff. I'm lucky that I get to work in front of a computer and have some time throughout the day to read the forums (lately). But, there is a burning desire for me to do this.

Btw, anyone know why isn't 9.1.2 up on SF?

"Steve Richie picks out the options above, he doesn't determine the little nuances of each flipper physics setting on the table he's creating."
Steve doesn't have to emulate the hardware he is using. I can easily think of at least 5 parameters just to describe how a solenoid works.

Obviously. That was my point. He doesn't have to worry about those nuances, the mechanical engineers do. Table designers here shouldn't either. The mechanical engineers of VP are the source code editors, not the table designers.

And, guess what, there aren't a zillion different types of flipper mechs. So once the design and parameters are down for one, your'e done and everyone can use that mech type in their table if they choose.

I totally agree with you on this. It would be great to have different types of flippers, bumpers ... tuned and ready to use,

[JAM0]

I agree with you analog999. I am surprised at the stutter a lot of us experience despite our decent (or better) computers, as well as the difficulty in getting the physics of a table the way we want it. Thanks for bringing all this up and I am excited that you are looking into the source code!

[analog999]

I agree with you analog999. I am surprised at the stutter a lot of us experience despite our decent (or better) computers, as well as the difficulty in getting the physics of a table the way we want it. Thanks for bringing all this up and I am excited that you are looking into the source code!

LOL don't get too excited. It took a team of people to get VP where it is today. I'm sure those people put a lot of time into it. I'll be lucky to have time this next year (self employed, new baby on the way, etc etc), but I will try to make some kind of progress in the physics department.

Edited by analog999, 21 December 2011 - 10:53 PM.

[DaveyDarko]

I love No Good Gophers, but the physics seem awfully "floaty" if you know what I mean.
Does anyone have some settings to improve the physics?

[chas]

As for jimmyfingers' comments, I think what you're saying is important, but the biggest flaw that most tables have are the flippers & the general table settings, IMO. That's why I'm posting those results. If you have changes to tables that you want to note, I'll be happy to add them to the first post. >>>>>>>>>>>>>>>

i feel the best way to do it is to scale the flippers to size and align the rails and flippers.
aligned the rails to the flippers with start angles 120/240.
end angle 65/295.
played much better.
thats just my opinion.

when i use shift and the arrows to move the flippers there are 2 ways to do it.
hold shift down and use the arrows will move the flippers about 1/4inch.
press shift, release and use the arrows the flipper will move about half and keep repeating to move a little at a time.
is there another way to move the flippers with a little less movement for more accurate positioning with the shift and arrows?

Edited by chas, 24 December 2011 - 08:16 AM.

[chas]

The original idea of the thread was to post settings for tables to improve existing ones>>>>>>

i tried it your way out of the DL.
did you try my suggestion for Alien Star?
align the flippers and rails straigth similar to the IPDb.org screen shot.
scale the flippers to the correct size and length.
strength on 5, speed0.075, elasticity 0.45, plaw2, OB 0.....
start angle 120/240 the flippers and rails align similar to the IPDB screen shot..
end angle 65/295.

after aligning flippers and rails i start with strat angle 65/295, OB 0.
some tables end angle 65/295 maybe too much swing.
bump the endangle down to 67.5/292.5.
if still to too much bump end angle down to 70/290.
then i use oblique if needed.

Edited by chas, 29 December 2011 - 07:46 AM.

[Slydog43]

I love No Good Gophers, but the physics seem awfully "floaty" if you know what I mean.
Does anyone have some settings to improve the physics?

I too think its much to floaty, great game just not emulated well.

[analog999]

The original idea of the thread was to post settings for tables to improve existing ones>>>>>>

i tried it your way out of the DL.
did you try my suggestion for Alien Star?
align the flippers and rails straigth similar to the IPDb.org screen shot.
scale the flippers to the correct size and length.
strength on 5, speed0.075, elasticity 0.45, plaw2, OB 0.....
start angle 120/240 the flippers and rails align similar to the IPDB screen shot..
end angle 65/295.

after aligning flippers and rails i start with strat angle 65/295, OB 0.
some tables end angle 65/295 maybe too much swing.
bump the endangle down to 67.5/292.5.
if still to too much bump end angle down to 70/290.
then i use oblique if needed.

I agree these are better settings, if not just to fix the problem where the flippers and those rails are completely in the wrong spot. But, the table is still lacking - the power on a backhand is not strong enough. Definitely an improvement though. I'll post your results.

By the way, you'll have to change the apron once you move the flippers and rails to the correct spot, otherwise the ball will get stuck.

Edited by analog999, 29 December 2011 - 07:24 PM.

[Bob5453]

I made this demo for you guys, so you can actually see what happens when you try to make shots with VP flippers. It's not really a pretty site. Maybe one of you guys can find the perfect flipper settings and share them with me.

http://www.vpforums....topic=17515&hl=

[analog999]

No Fear added.