2087 replies to this topic

[htamas]

Can't believe what I'm seeing, actually.

Awesome.

[mukuste]

holy shit!

 

somebody stop mukuste!

 

Don't stop me now

I'm having such a good time!

I'm having a ball

Edited by mukuste, 05 February 2014 - 01:26 AM.

[ICPjuggla]

Wow, awesome stuff! :-)

[bent98]

Hell yea

[The Loafer]

holy shit!

 

somebody stop mukuste!

 

Somebody stop patito!! Grrrrr LOL

[LoadedWeapon]

wow that is just incredible! does it play or just render? wonder how the fps is in dx9

[The Loafer]

I am thinking exactly what everyone is thinking:  Will this allow authors to push the boundaries of what VP can do, now that they won't be bound by the anchor that we call "poor DirectX 7 legacy driver support".  Although my understanding is not great, I'm pretty sure DirectX 9's much improved texture support and higher threshold will probably tables like UW's Monster Bash to really shine.

 

 

Can't wait until we see where this eventually leads.  The pics are fantastic.  Here we are, another year past this dinosaur program called Visual Pinball and yet again, someone comes up with something new that keeps VP fresh, with great improvements that will ensure our beloved hobby keeps growing.

 

Thanks for your efforts so far sir!  Keep the wips coming!

 

Rob

[Ark Malmeida]

wow that is just incredible! does it play or just render? wonder how the fps is in dx9

That's what I'm wondering.  Hopefully with dx9 VP will be able to use the full power of our graphics cards in newer OS.

 

Absolutely amazing work!

[arngrim]

Just WOW

[mukuste]

wow that is just incredible! does it play or just render? wonder how the fps is in dx9

 

It's fully playable; I didn't touch any of the other subsystems (sound, input, scripting, I/O etc), so those Just Work. Just the framerate is still very bad because I didn't put any effort into optimization so far, but there's a lot of low-hanging fruit there.

[Slydog43]

omg, Wow I have thought that everytime I read this thread.  What would be your opinion on actually getting this to run smoothly?  better than 50/50?  Keep up the incredible work, Thanks from all of the VP community

[fuzzel]

We have to change the rendering from the FFP (Fixed Function Pipeline old DX7 based rendering) to a pixel/vertex shader based rendering. By using this we can parallelize the rendering process.

Edited by fuzzel, 05 February 2014 - 12:40 PM.

[toxie]

Plus optimizing all the draw calls/vertex formats/state changes/material changes/texture changes.

 

So far Mukuste did all the 'ugly' work by porting a lot of things to the new interfaces and getting rid of the old region update code by brute force rendering all the elements instead. But getting this fast, 'modern' (even DX9 allows for a lot of things that are not GPU friendly) and bugfree at the same time will be another major challenge in the end.

[bmiki75]

Okay, those minor niggles have been fixed as well, I'd say we are getting pretty darn close here:

 

2014-02-05 03_07_59-Visual Pinball Player.jpg

I can't belive!!!!! Fantastic !!!!!

[bent98]

Plus optimizing all the draw calls/vertex formats/state changes/material changes/texture changes.

 

So far Mukuste did all the 'ugly' work by porting a lot of things to the new interfaces and getting rid of the old region update code by brute force rendering all the elements instead. But getting this fast, 'modern' (even DX9 allows for a lot of things that are not GPU friendly) and bugfree at the same time will be another major challenge in the end.

 

I am hoping you guys go for max performance and not worry so much about capability. In the end if people need to recreate tables for DX9 Platform so be it. Ofcourse it would be great to have existing tables just work in the new version but we also need to move things forward and shouldn't be limited by legacy support.

[kruge99]

Just a humble suggestion of naming this migration to DX9.  We might want to think a little bit ahead to the future, and save VP 10.x for an actual DX10 release.  Perhaps we could name the DX9 version something else?  I'm not thinking VP 9.3 or anything like that, but maybe we could take a queue from some of the MAC and Windows beta naming conventions?  Something like VP (name of a rather significant bridge or causeway) for example?  (we've only just begun heading down this road to DX9 and the light at the tunnel can now almost be seen rather than being a distant hope)  Since DX9 could potentially/eventually be a bridge to an OpenGL or DX10 renderer as I mentioned above.

 

 

Best Regards,

Todd.

[mukuste]

 

Plus optimizing all the draw calls/vertex formats/state changes/material changes/texture changes.

 

So far Mukuste did all the 'ugly' work by porting a lot of things to the new interfaces and getting rid of the old region update code by brute force rendering all the elements instead. But getting this fast, 'modern' (even DX9 allows for a lot of things that are not GPU friendly) and bugfree at the same time will be another major challenge in the end.

 

I am hoping you guys go for max performance and not worry so much about capability. In the end if people need to recreate tables for DX9 Platform so be it. Ofcourse it would be great to have existing tables just work in the new version but we also need to move things forward and shouldn't be limited by legacy support.

 

 

So far it seems that most elements can be ported quite easily without breaking compatibility. The obvious exceptions are cases where renderer bugs were exploited to achieve special effects; digging through some old threads, it seems that ramps together with Stereo 3D and/or tweaking the RO/RU flags sometimes were used to that effect? Is that still happening? Did the advent of flashers phase that out completely?

 

If someone could give me a brief overview of the "state of the art" of lighting in VP tables, that would be extremely useful. It seems to be a convoluted topic where people have invented many clever hacks.

 

If something comes up where we absolutely have to choose between compatibility and performance, then I agree that we should give priority to the latter.

 

 

Just a humble suggestion of naming this migration to DX9.  We might want to think a little bit ahead to the future, and save VP 10.x for an actual DX10 release.  Perhaps we could name the DX9 version something else?  I'm not thinking VP 9.3 or anything like that, but maybe we could take a queue from some of the MAC and Windows beta naming conventions?  Something like VP (name of a rather significant bridge or causeway) for example?  (we've only just begun heading down this road to DX9 and the light at the tunnel can now almost be seen rather than being a distant hope)  Since DX9 could potentially/eventually be a bridge to an OpenGL or DX10 renderer as I mentioned above.

 

 

Best Regards,

Todd.

 

There won't be a DX10 release, DX10 is basically obsolete since DX11 came out. Anyway, the features in DX9 will keep us happy for a LONG time, so I don't think we need to port to another graphics API anytime soon. So I still think VP10 is a good name. But if you want to suggest a catchy "VP: Subtitle", I'd love to hear it Perhaps "Visual Pinball: The DX9 Edition: Electric Boogaloo"

[fuzzel]

The lighting in VP9 is realized with lightmaps and images. Since VP9.2.0 you can use lightmaps but normaly if you want to realize light you have to draw a lit version of the area of the table you want to be lit OR you use an alpha ramp with an alpha flare map. Correct lighting isn't possible in VP today because a lot of things are prerendered. I would suggest to not support VP9 tables in VP10(DX9) anymore. Instead VP10 should load VP9 tables but playing them shouldn't be a priority. We should clean up the player/editor and remove settings which came in to simulate effects we can do with shaders now.

[DJRobX]

I am hoping you guys go for max performance and not worry so much about capability. In the end if people need to recreate tables for DX9 Platform so be it. Ofcourse it would be great to have existing tables just work in the new version but we also need to move things forward and shouldn't be limited by legacy support.

 

 

You don't necessarily have to sacrifice compatibility.   I would, however, expect to eventually see new "native" implementations of things (particularly around lighting) that blow away the old hacks that people were doing to achieve effects that hardware can now handle effortlessly.   

[arngrim]

+1 for fuzzel, worst case we can have one executable for vp9 dx7 and another for vp10 dx9, it is not a problem for the frontend to have a call to a specific emulator version