Register    Login    Forum    FAQ    PinSimDB.org

Board index » Future Pinball » BAM corner




Post new topic Reply to topic  [ 87 posts ]  Go to page Previous  1, 2, 3, 4, 5, 6 ... 9  Next
Author Message
 Post subject: Re: BAM - FP Physics Tweaks. [vip]
 Post Posted: Wed Nov 02, 2016 11:46 pm 
Offline

Joined: Thu Aug 16, 2012 11:12 pm
Posts: 2781
Location: Arkansas, USA
Thanks Rafal,

My initial impressions are good. On my initial testing, I couldn't tell any difference in the physics between the BAM.dll dated 9/8/16 and your newest one dated 11/2/16. I'll have to test it a bit more to make sure.

I'll also have to check and see how it saves changes that you make online. What happens when you make a change when you are using the "default.xml" file located in the BAM folder or an XML file that is contained in a ZIP file? Does it save the changes (it could be difficult in a ZIP file)? I'm not sure saving changes to the "default.xml" would be a good thing since the changes would affect many tables and not just the one that you are changing, but I'll have to think about it.

Wow! This should save me a lot of time! I used to have to save changes in the XML and then test them by playing the table and then close the table ... and repeat ... and repeat... I spent a lot of time waiting to load the table.

George


Top 
 Profile  
 
 Post subject: Re: BAM - FP Physics Tweaks. [vip]
 Post Posted: Thu Nov 03, 2016 12:08 am 
Offline

Joined: Thu Aug 16, 2012 11:12 pm
Posts: 2781
Location: Arkansas, USA
ravarcade wrote:
malifica wrote:
This is correct. There are a few parameters that require FP to be restarted to load the new values.

It should be fixed now.
In previous BAM versions, BAMl load physics settings from "tableName.XML" only when FP load table. So, if you loaded table, changed tableName.XML and pressed F5, when BAM was using old settings.
It is fixed now. Every time You press F5 new physics setting should be loaded from "tableName.XML". So for testing, you don't need to restart FP.


It is nice to have the F5 to reload the XML file but you still need to restart FP after you play each table because of FP's memory leak. Or has this changed? I thought it was one of the reasons for using PinballX, Hyperpin or Miownkhan's fix that is posted here:

viewtopic.php?f=55&t=5653

ravarcade wrote:
GeorgeH wrote:
... I verified that increasing the "Damping" in the XML file makes the ball more lively....

Now my "damping" in BAM menu should work exactly same way as "damping" in XML.


It has always seemed to me that "Damping" always worked backwards from the way it should.

George


Top 
 Profile  
 
 Post subject: Re: BAM - FP Physics Tweaks. [vip]
 Post Posted: Thu Nov 03, 2016 3:34 am 
Offline

Joined: Mon Mar 25, 2013 4:02 pm
Posts: 1034
malifica wrote:
On another note, is it possible to pull xball.speed from a hit with an optical target? If so, I may be able to do a realistic ball rolling sound across the whole of the table. Or if we can poll the ball status without an interaction with an object, we could just set up a simple script to play the correct sound based on x and y and velocity. That way it would not need any additional objects on the platforms...

See this demo:
http://www.ravarcade.pl/files/ballTrackingDemo.fpt
You can trace ball position and last hit-speeed.

GeorgeH wrote:
It has always seemed to me that "Damping" always worked backwards from the way it should.

With last update, "damping" in "Physics Tweak" menu is same as "damping" in XML.
In my opinion that "damping" is important part of ball. I only think that it should not be caled "damping". Only name is misleading... little bit.

I did not fix FP memory leaks. So after few "F5" it will crash, but atleast it will every time reload physics setting from "tableName.XML". In my tests, i created "damping-test.fpt" and "damping-test.xml". I have to frequently change physics params and compare it with own math. So quick reload helped me.

_________________
http://www.ravarcade.pl
Better Arcade Mode
current BAM version: v1.5-289, released: May 10, 2020


Top 
 Profile  
 
 Post subject: Re: BAM - FP Physics Tweaks. [vip]
 Post Posted: Thu Nov 03, 2016 11:59 am 
Offline

Joined: Thu Aug 16, 2012 11:12 pm
Posts: 2781
Location: Arkansas, USA
ravarcade wrote:
GeorgeH wrote:
It has always seemed to me that "Damping" always worked backwards from the way it should.

With last update, "damping" in "Physics Tweak" menu is same as "damping" in XML.
In my opinion that "damping" is important part of ball. I only think that it should not be caled "damping". Only name is misleading... little bit.

I did not fix FP memory leaks. So after few "F5" it will crash, but atleast it will every time reload physics setting from "tableName.XML". In my tests, i created "damping-test.fpt" and "damping-test.xml". I have to frequently change physics params and compare it with own math. So quick reload helped me.


I am good with it.

Thanks!

George


Top 
 Profile  
 
 Post subject: Re: BAM - FP Physics Tweaks. [vip]
 Post Posted: Thu Nov 03, 2016 12:04 pm 
Offline

Joined: Sat Oct 15, 2016 11:16 pm
Posts: 32
Rafal,

I too think 'damping' of the nature that is being applied to the ball object is an important element when the ball mass and gravity are much less than actual, for instance within the original physics settings of FP. This 'damping' would influence and assist with "realistic" ball/object interaction.

Edit: Removed the damping comment since I can't get a consistent result with the values.

I think once we can finally set in stone what the ball mass and gravity values should be to properly mirror the real world object, we can go a long way into finalizing all the objects and materials contained within the XML file.

I believe I have seen other physics calibrations that have stated the ball should be around 70 to 80 mass and the gravity in the 4800 to 5800 range, but I have no personal results that would verify this other than anecdotal observation of my playing around with table settings.

I suppose I can set up a table using your demo of the ball coordinates and speed with a create ball at the top of the table and a target a specified distance away and measure the time.

Do we know if the millimeter measurement represented in the editor is true?

I get a feeling the slope degree of the table is not actual just from casually fiddling with it, and is more of a representative value that is FP specific.


Thanks again for your knowledge and help.

Edit: This is frustrating. I cannot for the life of me get the ball to behave consistently.


Last edited by malifica on Thu Nov 03, 2016 5:30 pm, edited 2 times in total.

Top 
 Profile  
 
 Post subject: Re: BAM - FP Physics Tweaks. [vip]
 Post Posted: Thu Nov 03, 2016 12:39 pm 
Offline

Joined: Sat Oct 15, 2016 11:16 pm
Posts: 32
Rafal

The ball tracking demo is AMAZING!

I am imagining a table consisting of only holograms, decals and script. Playing a sound or moving an object with this is trivial... and awesome. But I'm thinking of full video and ball interaction...

Wow wow wow... It will take us all some time to truly grasp what this can let table authors do.

You could have a table object seem to intelligently pursue the ball, etc.

I'm guessing lighting effects and such on the ball are possible...


Top 
 Profile  
 
 Post subject: Re: BAM - FP Physics Tweaks. [vip]
 Post Posted: Thu Nov 03, 2016 6:21 pm 
Offline

Joined: Mon Mar 25, 2013 4:02 pm
Posts: 1034
malifica wrote:
Edit: This is frustrating. I cannot for the life of me get the ball to behave consistently.

This is "normal" thing. All physics engines adds some "random" values. There is no way to remove it.

I have one more experience with random results on CPU Float Point math.
In modern compilers you will find switch called "floatin point model". In Visual Studio it has 3 values: Precise, Strict, Fast. For Precise & Fast model some errors are acceptable.
For example when you compare 2 calculated values to make decisions and results are almost same you may end with different decisions from same input data.
In past i have to write optimisation/scheduling program. I lost week searching for error, because 1 for 20 runs i get different schedules/results for same input data.
Error was not in my program, but it was FPU math error.
Solution was to switch "floating point model" to "strict", but in that model whole math was 2 times slower.
We must accept it.

EDIT:
If you use Kicker for testing, check XML. Set impulseRandomness & vectorRandomness to 0.
<kicker impulse="800.0" vukImpulse="1300.0" impulseRandomness="0" vectorRandomness="0"></kicker>


---------------------------------
BAM.dll Update.
Added Flipper Mass and Omega.

For testing i have to left some "debug messages". Every time you press or release flipper in BAM menu you will see added text line like this:
"Flipper On [1/2] Omega: 33"
I left it, to test if i detect all flippers and only flippers. So you may check if there is more flippers detected.
There is also second reason. When you press Shift key, Flipper is activated with Omega value from XML. When you release Shift key, Flipper is activated with Omega value = 18.0 (or -18.0). I don't know if that value is hard-codded in FP or it is calculated.
So if You see different value, please let me know.

_________________
http://www.ravarcade.pl
Better Arcade Mode
current BAM version: v1.5-289, released: May 10, 2020


Top 
 Profile  
 
 Post subject: Re: BAM - FP Physics Tweaks. [vip]
 Post Posted: Thu Nov 03, 2016 11:34 pm 
Offline

Joined: Thu Aug 16, 2012 11:12 pm
Posts: 2781
Location: Arkansas, USA
To my thinking, some degree of randomness is a good thing although it can be excessive. In the real world, I suppose physics is perfectly predictable but pinballs interact with imperfect objects on the real world tables which tend to introduce a degree of randomness. I think everyone has experienced "do loops" on simulated pinball tables where the same sequence is repeated over and over again. I don't think that is very desirable.

George


Top 
 Profile  
 
 Post subject: Re: BAM - FP Physics Tweaks. [vip]
 Post Posted: Fri Nov 04, 2016 10:02 am 
Offline

Joined: Sat Oct 15, 2016 11:16 pm
Posts: 32
Oh I completely agree about having a random factor within the calculation, otherwise the experience becomes rote.

In my ball mass/gravity/damping testing...

What I have found is that I should have realized that there was a reason the material values were set by the original FP team as they were. Some of these original values seem reversed/non-intuitive when looked at as part of a group.

So in my "better judgement" I had changed all these values to what "made sense" based on the newton documentation and limited observation. Hindsight indicates this was not wise.

As Rafal has shown with some of the code snippets, not everything that is occurring within the simulation is linked directly back to an available documented newton function, and therefore I am observing some behavior that is, for me anyway, somewhat unexpected.

Anyway, I think I now have a better understanding (looking at you, friction) of what is occurring under the hood of FP now, and can better predict what the ball is going to do since resetting all of the materials (not objects) to their original FP values.

Thanks again all for your input, and I am finding there is a vast trove of information (Gimli, blue, George, beta-tester, etc) buried here within the forums.


Top 
 Profile  
 
 Post subject: Re: BAM - FP Physics Tweaks. [vip]
 Post Posted: Fri Nov 04, 2016 12:11 pm 
Offline

Joined: Thu Aug 16, 2012 11:12 pm
Posts: 2781
Location: Arkansas, USA
Where do you get newton documentation? Everything I have learned about physics on FP is by pure experimentation.

You know it might be interesting if each of us creates custom physics for the same table just to see what we each come up with.

George


Top 
 Profile  
 
Display posts from previous:  Sort by  
 
Post new topic Reply to topic  [ 87 posts ]  Go to page Previous  1, 2, 3, 4, 5, 6 ... 9  Next

Board index » Future Pinball » BAM corner


Who is online

Users browsing this forum: Janiecrymn and 44 guests

 
 

 
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Jump to:  
cron