4027 replies to this topic

[jpsalas]

 

 

 

A magnet is a trigger with some clever programming/scripting. But it uses properties that a primitive do not have, like being able to change the direction of a ball. But nothing stops you put a primitive on top of a magnet

 

yes

but my objective is for example that an arm magnetizes a ball, then deposits it elsewhere by rotation of the arm. without going through an invisible ramp trick

 

 

Well, that won't work. You will need to use one of the many tricks that has been used before, like ramps, kicker, or a trigger. But how to do this don't belong in this thread, so if you need more help open a new thread

[wiesshund]

 

 

 If it will help just let me know what you would like tested with VPX or VPVR and I'll be happy to do so.

 

 

 

Well, preferably both?
As it may help determine if an issue is inherited from vr or cause all by itself
but it probably also helps if someone with AMD gpu can test also?

cause people with nvidia gpu's are telling me, nope works fine here
so maybe some of the issues are gpu/driver specific

 

And how the hell do you screenshot vpxvr?
I try, and everything just sees right through the player display as if it wasnt there
even tried taking it out of fullscreen mode

[Cliffy]

Backdrop score displays and lamps fixed in 351. Looks like FXAA working again now too. See JP's Jokerz! pic attached- nice and sharp! Thanks again Vincent

[studlygoorite]

 

 

 

 If it will help just let me know what you would like tested with VPX or VPVR and I'll be happy to do so.

 

 

 

Well, preferably both?
As it may help determine if an issue is inherited from vr or cause all by itself
but it probably also helps if someone with AMD gpu can test also?

cause people with nvidia gpu's are telling me, nope works fine here
so maybe some of the issues are gpu/driver specific

 

And how the hell do you screenshot vpxvr?
I try, and everything just sees right through the player display as if it wasnt there
even tried taking it out of fullscreen mode

 

I click "screenshot" in the headset, but if you enable "VR Preview" in the VPVR settings what you see in the headset will show on the desktop and then "Windows Key + Print Screen" works.....probably not for non VR mode though. Also, the VR Preview no longer fills up the entire desktop but only part of it....since maybe 10 or 15 builds back. I see two new builds today for VPVR, off to check them out.

Edited by studlygoorite, 04 October 2022 - 11:19 PM.

[wiesshund]

Ah see i have no VR headset, so i have to set VR mode to disabled

[vbousquet]

Beside the x86 version not always starting up, are there other bugs to report in the latest version (v354) ?

[wiesshund]

Beside the x86 version not always starting up, are there other bugs to report in the latest version (v354) ?

 

I posted this before the forum database mishap ate the post

but it may be invalid now in latest revision
I will have to test against it

Only partial post recovered from someone quoting me
the other half with vpx test is lost to the ether

 

 

Now, i did a quick test of VPXVR latest build

this is with VR set to disabled and just running like a normal VPX table

 

Here is what i noticed

 

1) FSAA appears to be nonfunctional in it (i did try moving MSAA slider in VR settings to no avail)
i think that is carrying over to vpx during the merge.

2) Using the post renderer enhancements like bilateral CAS seems to crash VPXVR
complains that it could not find the shader

3) Screenspace reflections are almost nonexistent, and do in fact ravage FPS something fierce
turning them off tripled the FPS (i do realize a vr client will probably turn out less fps than non vr, different renderer and other different things going on, but triple fps is an odd loss)

 

4) wire ramps have a chewed metal look (maybe they dont in VR view, i am just going by standard screen view)

 

5) the texture rendering seems to be getting inherited from vr? not quite sure how to describe it, effect is very pronounced in the vr version though
as if someone began taking a softening brush to them i guess?

 

6) there is some kind of very odd texture artifacting going on, i cant get a screen shot of vpxvr to show you, i tried and it screenshot my desktop beneath it

but if you open jp's slimer, so we are looking at same table

Library, middle left of table
the top of the library is gone, and the walls are partly transparent, and you can see Slimers pivot mechanism, which should be hidden.
Not only place this is happening, but it is a super good example

2 inches up from drain
Metal wire going left to right, which should not be visible as it is below the playfield

top right, back metal ballguide wall
has a pronounced pixel twinkling / wandering effect going on that is present elsewhere, but this is a good spot to see it
resembled 'snow' on an analog tv when no signal is present.

 

I know this is VP10 and not VPXVR beta thread, but i am thinking that maybe some of this will at least help you in locating things that are causing issues, that are crossing over.

 

If i could figure out how the hell to screenshot vr, i'd get you some pics

reason post may still have some validity is, vpx was inheriting some of those issues

Edited by wiesshund, 05 October 2022 - 12:22 PM.

[wiesshund]

Ok just a quick test

1st rev 306 as i was testing against this version
note wire ramp rendering, and texture clarity, note the red flasher texture on right side
and note anti aliasing

Note presence of vivid screenspace reflections (and 0 performance loss)
(allow for degradation of screen shots obviously)

 

Ok now rev 354
Note the now improved wire ramp rendering
Note the FSAA is now working 100% best as i can tell
Note improved rendering of texture on the red flasher over previous revisions
Note that screenspace reflections now present and vivid, and back to no loss of performance

 

Now, rev 354, with Bilinear CAS sharpening enabled (at minimal performance loss)

Here is last 354 with bilateral cas on
in non jpg-ized tif

 

https://photos.app.g...D9eJvaManxkryUA

[wiesshund]

Here are the following in non compressed TIF images (so no jpg artifacting garbage)

 

AC\DC Luci 

http://photos.app.go...56m13rByrpAPGB8

 

Avengers LE 

http://photos.app.go...zoDMXVtM68fqkd8

 

 

Heavy Metal  

http://photos.app.go...sa2bxDccFgpMsM9

 

So quick testing
visually, and performance wise, i think we are back on track?

 

let's see what others find out?

Edited by wiesshund, 05 October 2022 - 01:53 PM.

[Cliffy]

yeah all my posts were lost but no matter now as 354 seems to be back to a solid base, Vincent. As wiesshund said everything looking sharp and no more missing or blown up elements, transparencies seem correct, details crisp and clear. I myself have had NO problems with vpx starting up. It starts fine every time and I couldn't get it to crash during repeated editing - play test - edit- test more,  sessions. Solid as a rock.

Windows 10, GTX 1060 6gb, vpinmame-sc-3.5-187, b2s 1.3.12

[jpsalas]

Yes, with rev 354 almost everything is fine again I can still see the lines at the sides of the ramps, in JP's Ironman, and I guess they are part of the screen reflections which you wrote they are not still 100%.

 

 

 

 

Attached Files

•  ramp-borders.png   647.41KB   2 downloads

[wiesshund]

Yes, with rev 354 almost everything is fine again I can still see the lines at the sides of the ramps, in JP's Ironman, and I guess they are part of the screen reflections which you wrote they are not still 100%.

 

 

 

 

 

JP
I noticed you had ramps butted up against ramps
and i also noticed that the vertices did not line up, in some places there is a gap in others the ramps are overlapping

 

I turned these to non visible and the lines you see went away

I tried playing with them, and where i could get them to line up perfect, the effect changed
but i could not get them to line up properly through the bends etc
ramps are a pain

[jpsalas]

 

Yes, with rev 354 almost everything is fine again I can still see the lines at the sides of the ramps, in JP's Ironman, and I guess they are part of the screen reflections which you wrote they are not still 100%.

 

 

 

 

 

JP
I noticed you had ramps butted up against ramps
and i also noticed that the vertices did not line up, in some places there is a gap in others the ramps are overlapping

 

I turned these to non visible and the lines you see went away
 

I tried playing with them, and where i could get them to line up perfect, the effect changed
but i could not get them to line up properly through the bends etc
ramps are a pain

 

 

It is not that, since the ramps have been looking great up to including rev 289. Here you have two images in PNG, which you can compare, and to me 289 looks nice and 354 still have something not working, and there are artifacts at the side of the ramps and of some 3D objects, and the images are blurred. You will see that other objects are also blurred like the dome flashers. I guess it is still a work in progress, but the display in 354 is not as good as in 289  

 

 289vs354.ZIP   13.77MB   10 downloads

[wiesshund]

It is not that, since the ramps have been looking great up to including rev 289. Here you have two images in PNG, which you can compare, and to me 289 looks nice and 354 still have something not working, and there are artifacts at the side of the ramps and of some 3D objects, and the images are blurred. You will see that other objects are also blurred like the dome flashers. I guess it is still a work in progress, but the display in 354 is not as good as in 289
 
289vs354.ZIP

 
The lines are happening where the 2 ramps are overlapping each other
If i move the ramps so they are not touching
No lines
I am not sure why it makes lines now
and they are not lines per se, the ramps sort of cease to render where they are overlapping each other
i am not sure if it is doubling the transparency?
but if you bastardize the POV so that the camera has your face right on top of one of the problem areas
it is as if the ramp is not there

Here look at this pic
See yellow wire showing through as if the ramp was not even there?
That is the lines


 
Here i slid over more so we can see more from the side
notice how it is as if the whole side of that ramp was erased at the meeting of wall to floor?

It seems to do that any place you have a ramp intersecting a ramp
Not sure if it does that to just the wall area or the ramp in general



 

turning off the top ramp layers though

 
So i tried this
set visible material to rubber red on all flat ramp pieces
left ramp has reflections enabled
 
right ramp has reflections disabled

 

Ok so no see through on either
isnt reflections

 

Ok so turned reflections back on
right ramp set material on all pieces to
Metal Red Transparent, no invisible spaces

 

Ok so i set back to material metal red for tops
and got back the invisibility potion

 

So i looked at material metal red
I set opacity to active and thickness to 0

Viola! no invisibility

 

Now WHY the invisibility now?
I do not know
But now we know how to reproduce it and what is actually causing it from a user/editor standpoint

 

No Lines (invisible parts of ramp)

 

With Lines (invisibility of parts of ramp)

So when an alpha meets an alpha coming through the renderer...

[jpsalas]

Thanks wiesshund, so I guess is back to the drawing board for me

[Thalamus]

I know that is not for everyone. But, it is a very good idea to take any table with ramps and/or being a bit advanced, export the whole deal as obj and just look at it through a 3D program like blender. I've countless times found things that is very hard to see from the 2D perspective of VP. Great visual tool on its own.

[wiesshund]

Thanks wiesshund, so I guess is back to the drawing board for me

 

Not necessarily
But now we know WHAT aspect of things is creating the invisible portions

from there it is a matter of determining if the change causing the behavior is intentional or unintentional

If intentional, then not entirely back to the drawing board
just a quick edit of the material being used for the offending part

[jpsalas]

I have done some more tests, and I still find 3 things that are different from rev up to (including) 289 and rev 354:

 

In rev 354 I still can see these 3 visual bugs:

- strange artifacts/light at the sides of static objects, as seen in the Iron Man table, at the sides of the top ramp borders. And as wiesshund tested, the glitch disappears if you make those ramps dynamic.

- static objects are more blurred in 354 than in 289 and older versions, check the playfield for example, or the bumpers, or the figures.

- the static objects are moved down, for example check the lights which light up part of the playfield. You can see that clearly if you load both images in photoshop and change from one to the other, then you will see that some images are moved a few pixels down.

 

The images are from my Lord of the Rings I'm updating now, and you can check those visual bugs in them:

 289vs354-part2.zip   14.97MB   1 downloads 

Edited by jpsalas, 06 October 2022 - 04:05 AM.

[Cliffy]

Just to add an example to JP's findings, I also see this edge "lighting" effect on VPW's 1.0.2 Diner. rev 289 looks good, 354 looks good but notice the left edge of the apron and the lighter parts of the backdrop showing through the misaligned right siderail (which I spent hours trying to align and just could not, ugh

These both with brute force 4x ssaa enabled

  

Edited by Cliffy, 06 October 2022 - 04:29 PM.

[fripounet]

suddenly display bug.

missing 2/3 of the pf ???!!!!  rev 354