88 replies to this topic

[jimmyfingers]

@BorgDog and Atarian - One thing I should mention that is very important to keep in mind while trying to tune older / EM tables is that PhysMOD and in turn VPX physics in general does seem to play more realistically / benefit the more modern tables that play faster and would have a real life slope of around 6 to 7 (around a 6 setting in VP).  I do actually still use BMPR and since PhysMODv5 (as my Firepower hybrid demo table showed) and still for VPX, however, I barely have to use any significant settings for modern tables (can have minimal parameters for only slight augmentation of ball speed), yet find I still very much like / need BMPR to reasonable degree as the table slope get's lower in order to reproduce a decent / comparable slow speed as with the real machines, yet allow for perceived acceleration down table of the ball and trying to still capture the large lateral ball movement that always catches my eye on real machines and for me has always been something that didn't seem fully simulated in any version of VP (and why I originally created BMPR but even with I still can't get the lateral movement - never have really - that I see in real life on EMs).  With adequate table friction and since PhysMOD physics some lateral motion has been improved natively with the ball spin helping the ball along on some vectors / trajectories and further helped (or masked at least) with the less lateral ball action and movement of modern / steep slope tables on which I find that I can be relatively content even with BMPR totally disabled, yet have to use it with increasing amount as I drop down in table slop / older generation tables. 

 

So far one of the tables I've noticed being able to play nicely "out of the box" without tuning or tweaking and get a reasonable feel of realism is JP's TOTAN.  Proper dimensions is very important also so that the ball can be kept at it's native 25 radius / 50 VP unit setting (way too much to explain as to why this affects physics for how I know / have verified this and involves source code aspects regardless of other counteractions such as the ball mass settings - many hours spent).  But even without taking my word, geometrically speaking and for the most accurate recreations, best practice is to start with a properly sized table for more similar in game interactions as to the real life table.  After a base of a correct dimensioned table, look at the guidelines document for a good starting point for physics defaults for objects that Fuzzel and I put together / he posted (JP's TOTAN is essentially built from that to my understanding and I think is a good example and has had positive feedback including some from previous TOTAN owners).

 

But to clarify for BorgDog (EMs) and atarian's question, getting EMs / early SS tables to play like real life with VPX is more difficult then for modern tables and may need some other supplemental routines like BMPR to get as similar a feeling as with the more modern table recreations.  I think when PhysMOD was developed that the focus / test tables used from Mukuste was on modern pinball machines / slopes and the variation / divergence on the physics as the slope dropped was not as tested / noticed for tables that in real life often only have about a 4 - 5 degree slope as with EMs and early SS tables that play very laterally and slow.  It seems still getting lateral ball movement is indeed difficult and too low of a slope still does not really achieve it while also having the downward acceleration affected too much and visually leaving the down table ball motion not looking correct.  

 

One last thing though that does help to give more lateral movement is to ensure high elasticity settings for rubbers.  I use very high settings (not over 1 though) but pretty close to it for rubber "bands" I call them for the rubbers that are suspended between two posts (separate wall objects from visual new rubber objects and from pegs as also walls).  The problem with high elasticity at times, however, can be a fast moving ball doesn't compress the rubber or hop off the table and lose some of it's X/Y speed as with real life so the ball can get too fast looking at times and in turn need another supplemental routine to help with the dampening effect for higher speed collisions but keeping lower speed ones nicely bouncy.  One can try and use the built-in "Elasticity Falloff" if sticking with pure rubber objects, however, I believe there are some nuances with the underlying coding that diminishes the ball speed (that I also initially saw in my manual dampening scripted routines before updating them) that adversely affects the ball trajectory when applying non-linear formulas to alter the speed of the Ball's X and Y velocity when the speed of X and Y at time of the collision are significantly different - mostly straight up table shot (I eventually changed mine to take the higher of the two speed augmentations / reducing values and apply evening to both vector's which gave better results). I still haven't verified in the source code the elasticity falloff coding / formula to confirm if this is the case but visually I have seen it look like how my first routines did before updating and is reasonable to theorize based on the general starting point of applying a non-linear formula to the dampening.

Edited by jimmyfingers, 08 January 2016 - 01:44 AM.

[zany]

I love to see all this talking about the physics. VP/Physmod5/VPX has always and still is by far the best pinball physics out there. But digging into details like this will VPX even better, and really close to the real thing...so keep on and discuss this further!  
It would be nice to have a reference table, to discuss around, like JP's TOTAN....a table i also have the opportunity to play in real life....as seen in the photo!

On table that looks really nice when it comes to physics, is Groni's AFM, and i really look forward to try it out.


 

Attached Files

•  12087988_10153683970469766_7513588963889536006_n.jpg   77.81KB   15 downloads

Edited by zany, 11 January 2016 - 06:43 PM.

[BorgDog]

jimmy, thanks for the detailed response once again, I am afraid however you have gone way over my head, as firstly I'm pretty new to this hobby and have no idea what BMPR is, but did a little reading on your posts on the Firepower table and alas am still pretty lost on what exactly it means or how I would possibly go about implementing it.  

 

Secondly I'm not really sure what exactly the "Elasticity Falloff" does for me.

 

And does not the slope in VPX match real world slope?  you mention a "real life slope of around 6 to 7 (around a 6 setting in VP)"  so the 4-5 real life of the older tables would be??

 

thanks again.

[atarian]

Many thanks Jimmy for the detailed info.  I have also mentioned before myself that JPs TOTAN felt particularly good in terms of physics to me - and I can compare it to a real table I have played locally.  There are others too, but that one particularly stands out.  I am a big fan of EMs so anything to simulate those more accurately is interesting to hear - hopefully that info will help BorgDog.  If you ever plan to release a demo of a VPX table using BPMR to simulate EMs that would definitely be of great interest to see.

 

I have dabbled with the Physics settings myself and can say it a task in itself. I haven't had much sucess and tend to make things worse, but plan the weekend to play with the values suggested in the txt file of this thread as a way to learn some more.

 

@zany - a perfect way to compare the physics with that setup.  You just need to get an AFM to the left of your virtual cab for comparison purposes of course 

Edited by atarian, 08 January 2016 - 06:53 AM.

[Ben Logan]

I think I've played most VP10 beta tables, and final release tables. What an exciting era for VP!

 

I encourage you guys to check out Congo for VPX. Best physics I've come across so far. Kind of an underrated table, imo. The theme isn't the most compelling, but the table is really nicely tuned. Very realistic gameplay. Ball movement is more predictable than with some VPX tables, but still seems alive, weighty, and exciting.

[StevOz]

Never liked BPMR, feel it is one personal interpretation on pinball physics, not necessarily correct, whilst adding extra scripting overhead, that can and do lead to performance issues. Using the available parameters within VP can provide as good as if not better physics settings without any extra overheads.

Edited by StevOz, 11 January 2016 - 09:18 AM.

[jimmyfingers]

Never liked BPMR, feel it is one personal interpretation on pinball physics, not necessarily correct, whilst adding extra scripting overhead, that can and do lead to performance issues. Using the available parameters within VP can provide as good as if not better physics settings without any extra overheads.

Thanks for steering this conversation into a poll about personal opinions on BMPR, but this really isn't on topic nor helpful to the people looking for tips and information on tuning VPX physics. But before returning to the main point of this thread, the topic of BMPR only was raised by me as a side aspect to what I've noticed with what seems to be more difficulty in tuning lower sloped tables with PhysMOd and VPX and the main reason I even dared raise discussion about lower slope tables was to relay my experience / observations to at least try and help BorgDog and others from not necessarily pulling their hair out as to why the older generation tables seemed more difficult to tune with VPX (and hence I mentioned what I still do / use to help with that phenomenon being BMPR).  It wasn't a promotional piece for BMPR and I don't really care if people use it or not.

 

I do care if people make claims about it, being stated in the supposed form of fact rather than opinion, that they actually get their facts right and don't spread misinformation.  You say the routine "can and (does) lead to performance issues". This is entirely untrue and unsubstantiated. BMPR adds ZERO performance impact in either FPS or stutter to any table it's used on and you could have verified this yourself easily if you spent merely a minute or two loading up a BMPR table and playing for a few seconds with both BMPR enabled and disabled (it was built with a simple single parameter for disabling / enabling and one could also disable the underlying timer as well).  Timer based scripting / sub routines that have no direct visual component (no texture / lighting manipulations or even text box updates to an extent) and are more computational can run quite extensively and elaborately without any measurable impact in VP.  You'd have seen from the FPS quantitatively that performance impact is indistinguishable in either case and your comments in this aspect for BMPR at least are baseless.  Time spent understanding something a bit more before bashing it would seem reasonable - by the way it's BMPR (Ball Momentum Physics Routine) not "BPMR".

 

One of the things that kills me the most with some "responses" on this forum is, after spending huge amounts of my time developing something like BMPR or tuning / creating a table with it and also going to lengths trying to describe things in any area in effort to try and help others, to then only have someone make baseless and / or unverified claims without even spending a mere few minutes of their own actual time before doing so.  That type of disparity in time / effort and validity is one of the things that turns me most off of even trying to contribute here or in any other areas of VP.  I've also spent a reasonable amount of time simply just in this topic with posts above to try and help people understand some important considerations for VPX tuning and from the "likes" on the last couple posts appears that people do appreciate that, but is disheartening to see a two liner comment float in that only focuses on the negative and bashes a side topic / option like BMPR - still something I developed and invested time out of my life to provide to the VP community, whether as needed nowadays or not.  

 

Everyone is entitled to their opinion (stated facts should still be actual facts) and I am fine with not everyone liking BMPR.  However, you make my total effort on BMPR seem extra shitty with the general tone of your post and especially with the "one personal interpretation on pinball physics, not necessarily correct" comment.  I never said it was "correct" or the perfect pinball physics addition for VP9 nor was it my own personal view of what pinball should be, but provided a sub routine that I did feel at the very least helped give people options.  There's enough people on here that liked it, wanted me to help them with their tables, and used it themselves to clearly show it was not merely "one personal" view on it seemingly helping VP visuals / ball motion at the time.  Try the VP9 version of Centaur that I made with fuzzel and tuned using BMPR and show me a better playing version of that table using the default VP9 physics.  Without / before ball spin, VP9 was always going to be limited and everything done on a computer for pinball in general is still "faking" it in some way or another.  I created the routine with a ton of configuration and internally documented parameters to fundamentally provide a adjustable tool that could be used to help further tune the VP9 tables to one's tastes and how one felt it may potentially get closer to the visuals and ball movement on a real table.  The complexity of it was more in getting the ball tracking up and running correctly than it was from adding the BMPR code itself. After playing most VP9 tables then going and playing a real table for the first time in years (back 4 years ago now) it was astonishing how much faster VP9 played and how much more lateral travel and gradualness the ball had in real life especially older SS and EM tables and why I started developing the routine when I came home that same evening.  VP9 did appear to have short comings in lateral ball movement and transition from up table to down table while trying to keep other physics settings balanced.  The most that was personal about BMPR from my angle was the recommendations, when others employed it, for things like setting the table friction and lowering the slope overall and most importantly that all table objects needed to be updated with higher elasticity settings to best / appropriately use it.  

 

The biggest component and complexity or tediousness for both BMPR and now with VPX is the same general thing, one needs to spend the time / effort and go around to every single collidable object on the table (or collections of similar objects) and appropriately adjust the physics to get more realistic play out of a table (with BMPR and PhysMOD one had to do that for at least elasticity, with VPX you now have to do that for elasticity and friction).  

 

jimmy, thanks for the detailed response once again, I am afraid however you have gone way over my head, as firstly I'm pretty new to this hobby and have no idea what BMPR is, but did a little reading on your posts on the Firepower table and alas am still pretty lost on what exactly it means or how I would possibly go about implementing it.  

 

Secondly I'm not really sure what exactly the "Elasticity Falloff" does for me.

 

And does not the slope in VPX match real world slope?  you mention a "real life slope of around 6 to 7 (around a 6 setting in VP)"  so the 4-5 real life of the older tables would be??

 

thanks again.

 

Sorry BorgDog as I missed your original post / questions from a few days back. 

 

As far as BMPR, for obvious reasons I'm not going to get into that much further here but if you search on the forum you can find some detailed posts and demonstration videos of the general concept at least for how it applied to VP9.  If you want to also discuss further than PM me and we'll take it off-line, however, you may want to wait to see one of my VPX / BMPR hybrid tables first to see if you like the effect before investing time on it.  I plan at some point, hopefully somewhat soon, to get two VPX versions of early SS tables uploaded but have had other VP type time allocation aside from table development take precedence up until recently (as well as now further losing a bit of motivation to bother prioritizing the time polishing off and posting any VPX table that might have BMPR as part of it). 

 

Elasticity Falloff is a setting originally on just flippers but now also on rubber objects.  It is a setting that diminishes the effect of elasticity / ball collision speeds proportional and increasing with the speed of the ball at time of impact.  It is not currently available on wall objects.  It is a good basic and native option for allowing high elasticity settings on flippers or rubber objects (for keeping good bounce on low speed collisions) but allowing for higher speed collisions to be cushioned / reduced / tamed and helps prevent the ball flying around the table at high speeds while using higher elasticity settings.  This falloff setting helps as without VP being able to currently model the compression that occurs with the rubber and also helps, it seems, to mimic the loss of energy from other aspects of high speed collisions on real tables (a bit of air ball, vertical trajectory diminishing X/ Y speeds that also doesn't occur the same in VP). I mentioned in a post above that I believe that there's a chance in how it's implemented could effect the trajectory but that's only based on some visual observations only and I still have to look back at the source code to confirm (that's party of why I personally use a modified external "dampening" routine and one that I also add a little bounce at very low speeds to help keep the friction on a table higher).

 

Here is a link to some descriptions about VP10/X's new physics that started with PhysMOD.  It's my understanding that the page was created initially by Mukuste himself (developer of the initial PhysMOD / updated physics for VP).  The Elasticity Falloff description is under the flipper options:

https://github.com/c...ki/VP10-Physics

 

 

I understand that my comments about 6-7 degree slope for real world modern pinball tables being set in VPX to to a slope value of 6 can seem odd and I would say that ideally they should match but I don't think they quite do.  Or at least it seems safer and more within range of decent playability to go for the lower end of that range but using 6 at the highest.  Most of the modern pinball recreations I personally feel play nice natively in VPX are using a 5.5 or 5 slope with still a gravity of 1 and friction still relatively high / close to recommended values (again see that link above).  The other key thing is that we are able in VP to set / manipulate friction and gravity unlike the real world for pinball (a little for friction I suppose with polishing or how the PF is managed but not at all to the same extent we can set it in VP).  So with these other settings adjusted, slope becomes less directly relatable and different to say the least and more subjective to how other things are set.  Problem again with lower slopes is that if setting for an EM around 4 (or what may seem "logical" for already described with VP from above to 3 or so), the table does seem to start to not quite look right for down table ball acceleration so this value of slope to me at least in VPX does seem to differ from the exact real world counterpart.  

[BorgDog]

Thanks for answering my questions jimmy, looks like I have more reading to do.  

[atarian]

I also just want to add Jimmy that I really appreciate your work on physics analysis in general and the detailed explanations. I don't pretend to understand the details particularly in regard to how tables are made however I do get the 'giste' of what you are describing having been update a basic understanding of real world physics.

I am also a very keen on EMs and early SS tables such as Centaur and find the BPMR made a world of difference to me personally. I love the VPX physics but without any criticism directed at the authors of the VPX EMs they don't feel 'right' compared to the moderner table physics some of which are spot on

I tried tweaking some EMs but for me it was disastrous mainly due to my abilities. I was pleased to read BorgDog was trying to capture a better EM feel. I hope you feel motivated to continue to experiment with a VPX / BPMR hybrid. You never know but successful findings might identify a setting change that may be implementable in VPX. Who knows until it is tried out.

Finally to add the first time I tried an early BPMR I thought it didn't feel right at first. However I had the chance to play a few real pins including EMs some time later, and revisited and found that they were muched improved with BPMR. I had become so conditioned to the Virtual world physics at the time I found any change unusual. If I am using the right term it is the lateral movement / momentum that could be better in VPX for EMs.

[toxie]

@BorgDog: could you share the test table for testing please?

[BorgDog]

@BorgDog: could you share the test table for testing please?

No problem.  https://dl.dropboxus... test table.vpx

 

left flipper kicks out balls at the top, right flipper drops the target/releasing the balls.  right flipper also resets the target and gets rid of the ball at the bottom.

I also just want to add Jimmy that I really appreciate your work on physics analysis in general and the detailed explanations. I don't pretend to understand the details particularly in regard to how tables are made however I do get the 'giste' of what you are describing having been update a basic understanding of real world physics.

I am also a very keen on EMs and early SS tables such as Centaur and find the BPMR made a world of difference to me personally. I love the VPX physics but without any criticism directed at the authors of the VPX EMs they don't feel 'right' compared to the moderner table physics some of which are spot on

I tried tweaking some EMs but for me it was disastrous mainly due to my abilities. I was pleased to read BorgDog was trying to capture a better EM feel. I hope you feel motivated to continue to experiment with a VPX / BPMR hybrid. You never know but successful findings might identify a setting change that may be implementable in VPX. Who knows until it is tried out.

Finally to add the first time I tried an early BPMR I thought it didn't feel right at first. However I had the chance to play a few real pins including EMs some time later, and revisited and found that they were muched improved with BPMR. I had become so conditioned to the Virtual world physics at the time I found any change unusual. If I am using the right term it is the lateral movement / momentum that could be better in VPX for EMs.

 

I'm also currently in the market to add a real em pin to my house, then I can compare the two side by side, and that will work better once I build my full size vp cab also. I agree on the lateral movement issue, get close to enough of that and the table is described as "floaty".

[jdark]

got a weird behavior of the ball. it got stuck on the right wall. look at the screenshot, the ball sticks to the wall and wont move. It is the default table when using new table from menu.

 

 

 

Attached Files

•  stuck.png   938.79KB   20 downloads

[fuzzel]

This is bug in VP10.0. If you raise a rubber a bit higher than the ball the internal hit shape of the rubber "punches" the ball into the playfield. Depending on the ball's momentum and spin this can cause various strange effects. I've already changed that and will be fixed in VP10.1

[jipeji16]

Thank you Fuzzel for this interresting discussion !!! Physics is not easy to set up ! Here i've found some bases !!

Thanks to everybody

[handy1957]

Hello all, forgive me if this is the wrong place to ask this but I just downloaded VPX3 and also downloaded two tables.

 

Diamond_Lady_VPX_1.0.1

Black_Knight_1.2.0.vpx

 

And both tables exhibit a uneven ball roll to the left of the table. 

Like the table is unbalanced.

 

Has anybody experienced this before and if so, how can I fix it.

[Thalamus]

Pretty sure you have a joystick or something attached. Disconnect and/or calibrate.

[BobAlbright]

I have found that adjusting dead zone can help with that

[handy1957]

Ahh, I will give that a try.  Thank you.

 

I did not realize that a joy stick would affect that. 

 

I had just reloaded windows in the last few weeks and I had not even touched the

calibrations for it.

 

Edit:  Ok so I unplugged the joystick and rebooted and tried again with the same issue.

The ball when falling down veers to the left.  Most of the time going into the far left lane.

 

Any other suggestions?

Edited by handy1957, 25 October 2017 - 02:04 AM.

[BobAlbright]

So did you try adjusting the dead zone ( in preferences)?

[gtxjoe]

Unplug your steering wheel too?