- This topic has 153 replies, 31 voices, and was last updated 3 years ago by
.
-
Posts
-
Yes, please, bring on any kind of patches/merge requests as Thalamus pointed out. Most of that code wasn’t added by me and i also don’t have the proper setup to test this with, so any help fixing such things is appreciated!
You need to login in order to like this post: click here
Nice to know I am not crazy :). I’ll find a use for the external sound card. I was able to reproduce the vbs fix directly in vpx, but its a much larger range than I expected there once all the math was done. I poked a few extra things during quick work breaks (my home covid office is about 2 feet from the pincab) and didn’t see any improvements. DirectSound3D is rather old and what tutorials I can find basically end up saying you just have to fiddle with it until you get something you like. Maybe I will have more luck tonight.
Interesting. Any chance you have Future Pinball (with pinevent pupssf) or FX3 setup with PUP SSF? That is a completely different code-base than vpx for ssf so could help troubleshoot and also Im curious if my implementation of 7.1 SSF does similar.
You need to login in order to like this post: click here
Not yet. Its early days on my pincab. Looking very much forward to doing so. But maybe I should just to see…. hmmm. We’ll see how I do today.
You need to login in order to like this post: click here
I didn’t have much luck last night, progress was elusive. The way sounds are moving around the table defy logic. Every time I learned something new and thought it was relevant it just made things worse. The vpx math comes out perfect to what I would expect. By the way https://www.geogebra.org/graphing?lang=en is a fun site for graphing, you can break a formula up into separate steps like f(x)=x+1,g(x)=f(x)/2, etc and graph each one along the way. Loads of math geeky fun.
You need to login in order to like this post: click here
Yeah. Well, my advice. If you can’t figure it out. Then don’t stick to it for too long. Let it rest, continue on your build and get back to it when you feel fresh and ready pick it up again. This is a hobby, and it should be fun and enjoyable.
You need to login in order to like this post: click here
You mean the hobby shouldn’t involve me screaming profanities at my computer late at night and kicking my unfinished pincab? Hmmm… interesting….
You need to login in order to like this post: click here
I think I have determined I’m not crazy. DirectSound3D does not map well to this application, what we really want is to send a sound to four of the surround sound speakers with a volume attenuated to the distance of the sound source to the speaker (rather than the listener). Surround sound is much more complex than that. I’d try to describe it but I would probably have most of the details wrong but suffice it to say sound is not split between speakers orthagonally, it more cone like. This is why adjusting the x pan was affecting whether sound was coming out of backbox or side speakers, the sound source was moving in and out of this cone. Although I am not quite sure how/if this explains the R/L channels of the backbox getting reversed.
So what is to be done? I think VPX 10.7 switched sound effects to be using the bass.dll library, I don’t know if that is DirectSound3D based or not and might behave differently. I will try it out sometime soonish. Can I do anything with VPX 10.6? Well I was able to hackish in something that avoids the cone of space that goes to the backbox speakers but that might be sacrificing a lot of the side to side pan effect. Could that be improved? I could maybe adjust for the cone depending on the current y position (right now it basically uses the widest end of the cone). Or maybe the playfield could be divided to 4 quadrants and those mapped to different subareas of the DirectSound3D sound stage that behave in the way we want.
I could still have that all wrong. But I found a DirectSound3D positioning test app which helped make sense of things to an extent, although some aspects still seemed strange.
I’m also getting pretty good at Attack from Mars through this testing, but I think v3 must be a lot easier than earlier versions, i hit the center shot much more consistently. I don’t know if that is realistic or not.
You need to login in order to like this post: click here
Bass.dll has been used for a very long time with VP. However the version of that linked library might have changed between 10.6 and 10.7. That is something I’ve seen several times during the last years. Sometime these changes gets reverted, you would have to talk to the devs about details. Mostly Toxie in this case.
Yes, I agree. v3 physics from JP makes the tables play easier, but, there is also positive sides to his physics changes. Physics seems like something we are not able to agree 100% on. Personally I believe that the best way going forward would be to either make more parameters to VP code, or change its physics for especially flipper ball control. I think that the so called nFozzy/Roth is the best solution, but, unfortunately it is scripted and seems to work differently from install to install depending on how fast your computer is. One essential thing that was done is the introduction of fast flips, but, that alone isn’t good enough. It basically means, don’t wait for pinmame for the flippers. Leads to other issues, flippers do often control more than just the bats. Not easy to emulate something that needs the speed of a solenoid in real life.
You need to login in order to like this post: click here
Well I kinda have a first version done. I think its a good improvement over the original and I think it can be reduced to about a 1 line change in VPX. The features of this version of 7.1 ssf (the 4 speaker modes are unchanged as they use the backbox speakers) are:
- no mechanical sounds are played from the backbox speakers
- sound effects on top of half of play field play only on side speakers (exciters)
- sound effects on bottom half of play field play only on rear speakers (exciters)
- side to side panning works as expected
What isn’t quite right about this? There is no fading between the side and rear speakers so as effects cross the vertical midline there is a sudden shift to a different set of speakers. Since the speakers are so close together on the cabinet it doesn’t seem to matter (to my ears anyway, i can’t really detect it), it would be somewhat jarring in a room size surround sound system. I was hoping to find a range that would fade between those speakers but I can’t find anything usable so far. I am continuing to look into this. The other issue I see is the side to side panning is maybe too subtle. I don’t want to exaggerate it but I might fidget it a little.
The biggest difference you’ll hear from the original is the removal of mechanical sounds from the backbox and the side speakers being used a TON more than they were before so you may need to adjust their volume over what you had.
When I am done I’ll talk to the authors and see if they are willing to update 10.6 and if they want to have a separate option for 7.1 ssf alternate or if they would want this a stock change or if they don’t want it at all.
EDIT: I imagine this will have some effect on SSFImpactor and SSFThump, might need some volume level adjustments.
You need to login in order to like this post: click here
Update: I was able to adjust the side to side panning. It might be too exaggerated now but I can adjust the formula easy enough now. I think its pretty good, there is a good range where it fades out of one side and into the other. So last thing is to see if I can get any kind of fade going top to bottom, but honestly I don’t notice the shift… I would need a table maybe with jet bumpers in the middle with half above and below the midline and then it might conceivably be noticeable. I’d need to learn how to build a test table and I don’t need that kind of rabbit hole right now, I’ve already gone too far afield.
EDIT: I found a range that side and rear speakers fade between without any sound going to the backboxes. So I just have to map the playfield range to that area. This should give us full vertical and horizontal fading which should give us a nice effect that the ssf was envisioned to provide. I also think I figured out how to add a SSF+enhanced option to the sound settings user interface if @toxie would want this as a configurable option.
You need to login in order to like this post: click here
Would not mind getting a copy to try out
You need to login in order to like this post: click here
When I have something done will contact you to be the first guinea pig. Wife made me catch up on chores about the house so not much progress today. I have a good range to map to get a fade effect but for some reason not many (any?) sound positions get requested in the middle third of the table. Its a rather conspicuous gap. I think we are close to the end.
EDIT: Found why there was a missing range (now I need to re-verify all the other work I did was needed as this could have been messing up a lot of things). Basically vpx had a tiny check to keep some values from hitting 0.0, the table scripts feed pan an exponential curve and a big chunk of that curve is (naturally) very close to zero and was getting clipped by that check.
You need to login in order to like this post: click here
I think I have it working! Pan (horizontal) and Fade (vertical) are mostly continuous across the playfield without any sound going to the backbox speakers. Slightly less than the top 1/3 of the playfield is forced 100% to the side speakers without any fade to the rear speakers but there would have been very little fade in that range anyway (sound leaked to the backbox speaker if it wasn’t forced in that range).
I am going to make a new sound option mode for this as it is conceivable one would want the old format. If you actually put front exciters/speakers in the playfield and had the backbox on a second sound card (if you can even do that) then you wouldn’t want to restrict mechanical sounds to the side and rear speakers. I think the effect would still be a little off but I wouldn’t want to take that option away from anyone.
Then its test test test test test test. Try some EM machines. Not sure where chimes are supposed to play from. Not sure where music plays from on music titles, etc. Just want to make sure everything works as I think it should.
And way too much Attack from Mars gameplay… I’m now suffering a little anxiety over the thought of aliens coming to visit :)
You need to login in order to like this post: click here
Not all EM’s have put the chimes where they most probably are on the real machines. So, don’t put too much into it if you find them in the backglass or by default location 0,0. Excited to hear that you at least seem to have come up with something you think is working.
You need to login in order to like this post: click here
Took a little trial and error but I added a new multi channel output option in the Audio Settings that will make vpx use my system. Tonight the testing begins. If anyone has suggestions for tables with potentially good surround sound effects, unusual sound features, or anything like that I would love to test them. My list right now is generally to check chimes and music tables to make sure they aren’t adversely effected by these changes.
You need to login in order to like this post: click here
Impossible to recommend anything. There are so many different tables out there. If I suggest some, it would be what I like and probably not the variation that you are looking for. Music tables should of course play the music in the backglass. Default, there are a lot of originals where it is put on the table. EM’s, chimes, is a sample you might want to put to the backglass ( which is wrong – I know ) because exciters can’t reproduce sound as clearly as sounds through regular speakers. Of course, my opinion on the subject.
You need to login in order to like this post: click here
Well first round of testing suggests this is working pretty well. All the issues I discovered turned out to be issues with how well the table supports SSF. A little disappointing SSF isn’t more widely supported. Makes me wonder if it would be useful to make a DOF version like SSFImpactor but does all the sound effects not just the impacts. Idle thought and a rabbit hole I don’t want to go down at least until I build my cabinet all the way.
I’ve cleaned up and commented the code. I will make a test build and send it to the couple people interested to get their opinion (I can adjust the pan/fade algorithms a little). Toxie is willing to include this in a 10.6.2 build if it checks out, DJRobX confirmed the issues I was addressing were valid and gave helpful advice, so I think we are on a good path.
One open question is whether this should replace the existing 7.1 surround option (DJRobX’s opinion) or should be a separate option. I was thinking separate option so as not to affect anyone who might have things just as they like it. It’s not a wildly different effect by any means, with all music and rom sounds on you might not really notice it too much. If you turn off rom and music sounds you can certainly hear the difference. In any event the test version has it as a separate option to make it easier to compare the versions. Perhaps the testers will have an opinion.
You need to login in order to like this post: click here
Good to see that you’ve already contacted the guys I would suggest :) I agree, not all tables are optimized for SSF. But, I expect you are aware of the vpx scripts repo ? Most new tables do support it, but, maybe not always are optimized. https://github.com/sverrewl/vpxtable_scripts
I would be lying if I say that those scripts I refer to are too. Those just improves on what was released. Most tables could do with a better variation of sounds, like TOM. And recently, many tables do.
You need to login in order to like this post: click here
I was not aware of that, awesome. Now I am interested in that sidecar option… is that still not available?
You need to login in order to like this post: click here
Yes. Standard since VP 10.4 – so, just hit that green button to download all the scripts as a zip. Named exactly like the table and it will be replaced on the fly. That also means. Be very careful of trying to rename a script from there. If you need to do that, either the table has been updated since or you don’t have the “correct” version.
I prefer to exact all tables that I use at some point in time. You can open a cmd prompt and run … vpinball.exe -extractvbs <tablename> and you will get a *.vbs.
You need to login in order to like this post: click here
1 user thanked author for this post.
Forums are currently locked.
