2585 replies to this topic

[jbg4208]

Just to throw in some more information on this for you Dozer.

 

I have two machines,

One is - win7 Intel i5 2400 3.10 ghz 8gb ram and GTX650 and getting a frame rate of 195 steady on Judge dredd v5  - Smooth play.

 

Now the other is - Intel core 2 duo E8400 3.0 Ghz 16GB ram and Also a GTX650. Getting frame rates of 55 at max, on Judge Dredd v5. Most of the time its 25-35

 

This might be an issue of processor speeds/type causing the slow downs in VP10. 

[toxie]

i can hopefully look into this soon. currently its all a bit busy in real life, plus the rest of the time its also actually summer in germany.  

but i would suspect that its the physics again, as it seems to be CPU dependent.

[PilzTom]

looking out of the window and the forecast it seems summer is over... ;-)

[Dozer316]

Thanks Toxie.

 

Let's all wait to hear what the devs come up with.  Maybe there is something internal to VP that is causing the issue.

 

Until then, I'm going to release one more update in the next day or two which addresses some friction issues so the game plays a little slower around the top orbits.

 

If you're happy with the current gameplay you won't need it but if you find the game a bit quick the update will fix it for you.

Just to throw in some more information on this for you Dozer.

 

I have two machines,

One is - win7 Intel i5 2400 3.10 ghz 8gb ram and GTX650 and getting a frame rate of 195 steady on Judge dredd v5  - Smooth play.

 

Now the other is - Intel core 2 duo E8400 3.0 Ghz 16GB ram and Also a GTX650. Getting frame rates of 55 at max, on Judge Dredd v5. Most of the time its 25-35

 

This might be an issue of processor speeds/type causing the slow downs in VP10. 

Thanks for the info.

[toxie]

wouldn't be the first time that the physics trigger a special case that is not properly handled somewhere and makes the engine do a loooooot of work in some unexpected place.

[toxie]

hmmm.. cannot spot the slowdown on my test-systems.. :/

so i need more info from the guys that see the slowdown..

 

especially jbg4208: could you try spotting any kind of difference in your setups? like in the video settings, resolution, FS vs Desktop, etcetc.

 

 

EDIT: maybe its also the 'fault' of the three static balls? could the guys seeing the slowdown just try disabling the creation of these?

Edited by toxie, 01 September 2015 - 07:29 PM.

[Kaan]

I've disabled the static balls and it helped quite a bit, now I get 30+ fps instead of 4~. Load_trough timer affects the fps a lot too, I get more than 60 fps if I disable that one.

 

Regarding the differences in setup, In my case I've found out that 4-5 fps issue only happens on my 2.2 c2d and not on 3.2 c2d. They are very similar cpus and if I underclock the faster one to 2.2 the same issues happen. Hope this will help.

Edited by Kaan, 01 September 2015 - 08:16 PM.

[jbg4208]

hmmm.. cannot spot the slowdown on my test-systems.. :/
so i need more info from the guys that see the slowdown..
 
especially jbg4208: could you try spotting any kind of difference in your setups? like in the video settings, resolution, FS vs Desktop, etcetc.
 
 
EDIT: maybe its also the 'fault' of the three static balls? could the guys seeing the slowdown just try disabling the creation of these?

That is what i have been trying to find for the last week, difference in my setups. Video drivers, Resolutions and FS. They were both setup just the same. I did try DT mode on the slower one the result was the same.

I finally noticed that the motherboard and processors were different. Thats when i made the assumption that it could be a processor issue.

Also notice on the faster rig that pinmame is version 2.5. The slower rig is 2.6. But dont think that would cause an issue.

Ill try the static ball thing when I get home later.

EDIT: removed the static balls and got about 50 more FPS. So it does help. Now getting a consistent 80-85

Edited by jbg4208, 01 September 2015 - 09:24 PM.

[evilantal]

I didn't see an improvement with the new version unfortunately, haven't tried the static ball thing yet though.

I even tried setting textures to 1024 and setting the object accuracy slider halfway, but that didn't help either.

 

I don't get the extreme 5fps that some were experiencing, though. It's just a constant stuttery thing with the ball movement.

[toxie]

so i can only guess that the static balls and the ones in the real trough on slow CPUs get into some kind of weird state which they cannot leave again, causing a lot of computations for the following physics cycles. maybe somebody with the problems could try varying the way the balls are released (height, other speed, angle)?

[Kaan]

I've tried doing some changes but it is beyond me, I don't really understand what to do. If someone can point me in the right direction I'd be happy to test.

 

Tried changing kickers from legacy to normal and some,I think, speed settings in the script but couldn't see any difference.

Edited by Kaan, 01 September 2015 - 10:41 PM.

[Dozer316]

Here's something else I noticed after these static balls (captive and trough) discoveries.

 

Try this - 

 

Revert the table to the state it was when you downloaded or get it again so we're all starting from the same point.

 

On the left hand side of the table where the timers are, turn off the collision timer.

 

This gets me another 100FPS on my dev system.

 

Maybe the collision routine is chewing up cycles on the static balls that only becomes an issue on slower CPU's.

[jpsalas]

That's my old collision routines, VP10 has now built in ball to ball collision detection. You simply need a sub like this:

 

Sub OnBallBallCollision(ball1, ball2, velocity)
    PlaySound("fx_collide"), 0, Csng(velocity) ^2 / 2000, Pan(ball1), 0, Pitch(ball1), 0, 0
End Sub

 

This was in my latest Papa Smurf table. Take a look at that table for the newer rolling sound sub. Also the rolling sound timer interval should be 10 or higher, there is no need for a faster interval.

[jbg4208]

Here's something else I noticed after these static balls (captive and trough) discoveries.

 

Try this - 

 

Revert the table to the state it was when you downloaded or get it again so we're all starting from the same point.

 

On the left hand side of the table where the timers are, turn off the collision timer.

 

This gets me another 100FPS on my dev system.

 

Maybe the collision routine is chewing up cycles on the static balls that only becomes an issue on slower CPU's.

 

Wow, That got the FPS up to 135 max and steady.

[Dozer316]

An update is coming shortly with the new collision sub.

[Pinhead45]

You guys rock!! Disabling the collision timer bumps me up to 108 fps. JD is now fluid not choppy and I can play it.

Edited by Pinhead45, 02 September 2015 - 04:42 AM.

[Dozer316]

Ok, everyone try this version - Version 7

 

https://mega.nz/#!bA...DvmqV7-lBEkbfXI

 

Updated to new rolling ball sound and collision routines thanks to JP. (Increased my frame rate by 120)

Slowed down top orbits with increased wall friction.

Updated some lamp inserts with correct fade timing.

Added 1 crime scene insert array I completely missed in the previous versions.

 

Please report performance when you can  -  Thanks to Toxie and JP Salas for pointing me in the right direction. 

[hauntfreaks]

I'm getting 202 on my rig with JD7

[Dozer316]

I'm getting 202 on my rig with JD7

How is to play at that framerate?

[hauntfreaks]

perfectly smooth.... even during multiball it only dropped the framerate down 1 or 2 frames...   no shutter, no slow downs