13 replies to this topic

[jimmyfingers]

This is a Hybrid of my Firepower BMPR Mod from NFs / URs original Firepower table but blended with PhysMODv5 in attempts to get the best of both and make up a little for some nuances still with PhysMODv5 (Mod of a Mod).  In general, BMPR Mod tables (at least my versions of BMPR) are good PhysMODv5 staring points because, since the first PhysMOD versions, it's been required to modify table components and have higher elasticity and bumper / slingshot strength settings for example.  This has been something that I've been having to do in my BMPR Mods anyway because at the core of my modifications was increasing table friction to help the ball look weightier but using the actual BMPR routine to help compensate for the friction increases (and lower slope for lateral ball motion improvement).  Also, by doing these elasticity increases already with my BMPR mods, I had added / needed my dampening routines to the rubber objects groups for not having the ball bounce as hard / fast at higher collision speeds and these same routines are helpful in PhysMODv5 tables for similar reasons (although the new spin and friction models helps with this somewhat now too).

 

The flippers with PhysMODv5 are much improved with aim and interaction but backhands from the base / heel area of the flippers are a bit weak and have some difficulty representing the shots / angles that are attainable on comparable real tables / flipper setups (when keeping the flipper geometry - start / end angles - consistent with their real world counterparts).  This is an aspect that seems to have been mentioned by a few other people as well in the main PhysMOD topic / VP development topic.  These backhand / base shots are tougher when the ball is moving through the in-lane towards the flipper (especially with some speed) but also are experienced somewhat with the ball cradled and then shot from close to the base / heel.  In efforts to improve the backhand issue slightly the flippers are a little higher set for strength settings and I've crated a revised "FlipperDampener" routine and trigger size to assist with taking some speed off the ball but not doing so in the small area close to the corresponding flipper bases. However, unlike the previous routine(s), there is no alteration of trajectory disproportionately from the VelX and VelY values (i.e. does not mess with the nice aim of the PhysMOD flippers).

 

The BMPR routine used is based from the revised one implemented with the Centaur Table Fuzzel and I did.  It uses a faster timer but smaller values for the various augmenting equations, which themselves have been altered slightly in their formula (for side action, up table, and down table - MX, MYU, and MYD respectively).  It's mostly used in this table to allow for a lower table slope value that helps for more lateral ball movement that still seems not quite as realistic with PhysMOD alone as in the real world, especially for older generation SS and EM tables, despite the improvement from ball spin helping to carry the ball further than with native VP.  Would be nice to see if some of the future / proposed mass settings that were discussed ever can be implemented and the affects on the lateral motion / momentum of the ball.  Ultimately too, the way that rubbers interact with the ball on a real table (absorbing and compressing) would have some bearing on the angles / force of ball bounces and lateral game play aspects that maybe too some day can be incorporated into VP's physics

 

Other "helper" physics routines have been removed where possible and the original BMPR Firepower table had some for the left loop and top middle area that now are no longer needed thanks the the new physics of the PhysMOD v5  The rubber dampen routines I used have been modified to actually do the opposite at very low speeds and that is, with a currently crude implementation,.the ball actually gets a bit of a bounce boost for low speeds off of the rubbers.  This is to help mimic again the real life way that a ball can bounce frequently but minimally off of the top of a lane guide / rubber post but while not having to drop the PhysMOD table friction settings too low.  I have seen some tables released with very low PhysMODv5 friction setting and the result of any friction setting much below about .07 or .06 is that the ball is sliding much more on the table than actually rolling with the effects of ball spin on the PF itself becoming essentially nothing (this is easily observed on any table with very low PhysMOD friction settings by using the F11 key as you will see the ball move / pushed across the table but not have the green dots depict any rolling behaviour).

 

This tight relation with friction and ball spin / slide is another large part of the reason that the BMPR hybrid is helping out.  However, on this table I feel it's still a little "sticky" / tiny bit slow but not out of the realm of what someone's real Firepower table may be set to play as or in general for what a table of this generation would play like.

 

I've kept all the visuals identical as with the original BMPR Mod release except for enabling the built-in ball reflections and increasing the ball size to match better the table dimensions that were used (keeping it close so that better comparisons can be drawn from the two versions - and even the original from NF / UR).  Increasing the ball size can also have the appearance of slowing down game play / adding weight, which can help the look and feel at times but in this case may contribute to a little bit of the slower play.  I've come to think that it's not so much the ball size increase that seems to give it the look of more weight / less speed (even on VP9x tables) but more to do with the fact that a larger dimension / oversized table, which are the ones needing the ball size changes, requires the ball to travel further and as such while keeping the same space available on one's display effectively means the ball is not going as fast per inch of screen (a theory I'm still contemplating).  In any case, using the ability to increase the ball size seems to result in a heavier looking ball / slightly slower play.

 

Lastly, the big changes were in huge amounts of time spent trying to find the best blend of flipper settings for tricks and realism (not having the ball fly around the table from a flick or with a post pass that looks like it could rip the post off the table).  These refined flipper tricks and interaction were the subject of a video post I made last week that can be found here:

http://www.vpforums....firepower +bmpr

 

If anyone trying this table is not able to make some of the same flipper plays as in that video, especially the tap passes, then you may need to look at assessing / improving / overclocking your USB polling rate if your input devices are USB based.  I wrote another post on that topic here:

http://www.vpforums....showtopic=29207

 

It seems better to not post PhysMODv5 tables in the actual table release sections so, following a similar method, here is the link for a public download location:

https://dl.dropboxus...MPR Hybrid).vpt

 

Enjoy and I hope people have a similar experience as I did playing this table that for me produced as close to the "real thing" as I've witnessed in VP to date.  Thanks too for the VP development team and the advancements they are facilitating from physics to graphics to ease of use.

Edited by jimmyfingers, 27 September 2014 - 09:45 PM.

[freneticamnesic]

Thanks! Plays awesome, excited to give it extensive testing

 

A single nudge using keyboard keys tilts the game, though

[javier1515]

Thanks JF plays amazing.

[jimmyfingers]

Thanks! Plays awesome, excited to give it extensive testing

 

A single nudge using keyboard keys tilts the game, though

Find this line:

vpmNudge.TiltSwitch=1:vpmNudge.Sensitivity=11':vpmNudge.TiltObj=Array(LeftSlingshot,RightSlingshot,Sw21,Sw22,Sw23)

 

And adjust the .Sensitivity value to something like 2 or 3

 

Also, you / others might want to unrem the ':vpmNudge.TiltObj=Array(LeftSlingshot,RightSlingshot,Sw21,Sw22,Sw23) line, however, it was remmed out in the original from NF / UR which is why I left that particular element that same way.

Edited by jimmyfingers, 27 September 2014 - 11:18 PM.

[freneticamnesic]

thanks

[unclewilly]

Cant wait to try it out jimmy. 8 more days and ill be able to unpack my dev pc and get my cabinet set up

[blashyrk]

I'll test this later tonight and please say you'll mod more tables with the hybrid physics, thanks for your outstanding work

[bolt]

Very good. Thank you JF.

[The Loafer]

Hi Jimmy.

 

Overall, very impressive work, especially how the ball interacts with everything, feels very real.  I own a real firepower and you've done some great work.  My firepower has had a flipper rebuild a couple of years ago, so the flippers are very strong and accurate, and this is one area where I do feel like there is room for improvement its in how the flippers feel slower in your table than the real thing, where I will be a little late on shots compared to the real table, and compared to other physmod 5 tables (though yours is not alone, I've felt these slower flippers in other physmod 5 tables as well and have mentioned this before). I used to see this issue as "lag" but I'm not so certain since this doesn't seem to be the case in most other phys5 tables and even some higher CPU intensive tables don't have this issue.  Just so its clear, it's not a bad flipper setting, just seems like I have to compensate more than I should.  IMHO anyway.

 

The ball speed seems to be pretty good overall, my firepower pin seems to be a bit faster in the lower parts of the table but perhaps I simply have it set at a higher slope so something I can experiment with your table.

[jimmyfingers]

Hi Jimmy.

 

Overall, very impressive work, especially how the ball interacts with everything, feels very real.  I own a real firepower and you've done some great work.  My firepower has had a flipper rebuild a couple of years ago, so the flippers are very strong and accurate, and this is one area where I do feel like there is room for improvement its in how the flippers feel slower in your table than the real thing, where I will be a little late on shots compared to the real table, and compared to other physmod 5 tables (though yours is not alone, I've felt these slower flippers in other physmod 5 tables as well and have mentioned this before). I used to see this issue as "lag" but I'm not so certain since this doesn't seem to be the case in most other phys5 tables and even some higher CPU intensive tables don't have this issue.  Just so its clear, it's not a bad flipper setting, just seems like I have to compensate more than I should.  IMHO anyway.

 

The ball speed seems to be pretty good overall, my firepower pin seems to be a bit faster in the lower parts of the table but perhaps I simply have it set at a higher slope so something I can experiment with your table.

I think these are reasonable areas of critique and decent observations.  Some of the speed in general feels a little slower than I would like, but this is where I was saying in my first / post where the friction setting is tricky to reduce much more as than the ball is visibly sliding too much in PMv5 (green dots / F11 verifies this).  I too though would like to see it be able to be a little less sticky feeling but maintain the ball to table spin characteristics.  For a small change / improvement and with about the min. friction to use before table spin / roll becomes too affected, try .065 with the slope increase to 4.55 or 4.6.  The speed also seems to be on the slower side for tables that are over sized and which I was discussing in my first post as to my theory.  Whatever the case, it seems that once one adjusts the ball to 53 or 54 it appears slower / with more weight and in the previous versions of VP this was welcomed, by me at least, where as now with PMv5 and it's new physics, this resulting slow down due to table dimensions / ball size increase can put it a bit too much on the slower feeling side.  I've witnessed / confirmed this while working on this table and porting all my physics settings to another one to try out and the results were quite different and puzzled me at first until I noticed the table size issues and that if copied to a similar over sized table, the speed of play felt relatively the same - all other settings held constant - yet copying / setting physics / BMPR the same on another smaller / correctly sized tabled, that the ball felt much quicker - quicker then even it seems it really should for only at max a 10% difference in pinball size.  

 

If the flippers feel a little weak, then find the flipper dampener routine in the script and lower the level from where it is now (it's somewhat exponential so with it being at 4 now you might need to try 2 or 1).  However, it sounds like the flippers for you are less about strength and more about responsiveness.  I think part of what you're feeling is the use of the "Coil Ramp Up" values mainly plus some potential aspects of your system.  The differences from PMv5 table to table that you're observing will likely have some bearing on the utilization of this parameter.  I found that once I got to the precision level needed for the tricks combined with having to use some of this "Coil Ramp Up" parameter to achieve these tricks, that my system settings / input lag from USB devices plus the "max prerendered frames" in VP's video settings made a very big difference.  I have "max prerendered frames" set to 1 and as far as input device lage, I can now get down to as low as 6 or 7 ms registering for both a press and a release from my mouse buttons, which I use now as my flippers, after USB overclocking on my X-Arcade Tankstick.  This made a huge difference in my ability to play the table optimally and is why I also wanted to take the time to mention it at length in another post before I released this table link as I do feel it could be key in reducing what you are noticing.   Take a look at this article if you haven't already and without embarking on the overclocking aspect, I'd be curious what values your getting from the keyboardtest.exe program as a start on your cabinet (if it's anything above 24 to 32 ms, you're going to lose a few frames there as well - 16 to 8 is good to great).  Could you post your results from that if you feel comfortable enough installing that small program (should be safe enough to uninstall afterwards and I haven't noticed any ill-effect from either my development station or my mini-cabinet but I do understand reluctance with installing software - even if small - on ones cab).  Here is the link again:

http://www.vpforums....207#entry282893

 

I still think too that some fine tuning of the new physics in general and for the flippers would help as well as there is a tough balance to get flipper tricks as an option and "Coil Ramp Up" seems to be necessary for some yet add this somewhat weaker / laggier feel to flipper hits / timing.   As a test too I'm wondering if you could change the flippers "Coil Ramp Up" setting down to 0 and see how you feel then about being later on the shots then expected and as experienced on your real Firepower.  That will wipe out the ability to do tap passes thought and likely rolling tap passes, however, I'm curious for comparison and for problem isolation as to what you experience and feel the difference is with this one setting changed, keeping everything else the same.  

 

You're obviously a good resource for these comparisons having a real machine and part of the reason I picked Firepower with the new hybrid physics / PMv5 tweaking was that that was my first BMPR table and seemingly another good one to break into hopefully a new realm of improved physics again.  I will say too though that the PMv5 at this point can not do enough on it's own yet even though it is a vast improvement over the previous VP physics (I personally have difficulty go back after playing / modding some tables with it).  Hopefully some day the PMv5 / VP10 physics code can be optimized as well as other physics things improved like the interaction / compression of rubbers, and maybe some continuous collision models for the rubbers / flippers as well that might help with the more real world aspect of the flipper coming in contact and staying in contact briefly before the end result / final trajectory, whereas with VP it all seems to come down to a momentary, more geometrically static, collision and I think this has a bearing on some of what we are seeing with the flippers (missing a bit of a hit and push although it is grabbing / biting better now with the new friction settings option on the flippers set fairly high).

Edited by jimmyfingers, 01 October 2014 - 12:36 AM.

[The Loafer]

Yeah, I saw and read your previous post about the flipper lag. I haven't looking into it for my pincab yet but my reasoning (perhaps flawed) is that I don't see this lag issue on most other tables, that's why I personally felt it was more flipper speed than actual lag.  Could very much be wrong, I'm no expert.  I will try and look into this more this weekend (pretty busy real life at the moment), it would be cool to burn away some of the mystery around flipper lag, etc. and if whatever I report back can help others, it would be my pleasure.

 

You provided great information in your posts on this issue Jimmy.  I can't speak for others but if the response is a little low, from MY perspective anyway and I suspect others its a lot to digest and I'll admit some of it is beyond me.  I imagine others may feel the same.  But this type of info can lead us to better enjoyment, so I certainly appreciate the effort you are going through to help improve flipper response.

[dark]

It's probably the best display of game play physics I've seen in a VP recreation to date.

[dboyrecords]

Played many games and LOVE it! Soldier on!

[dboyrecords]

Bump. Played a couple games last night, really love it! Any developments/new test tables?