Dyno runs needed :P

[Warlock]

Well, it's just the sound samples that are mono: the GPB sound engine is stereo.

True

So it's in general up to the engine to reproduce these effects: you don't have to capture them in the samples.

Also true, the sound engine seems not good enough for this, unfortunately.
For me, sound is 50% of a racing experience, very very important for an immersive simulation.

But for GPB I don't think ambient-specific processing is very important: we don't ride too much in tunnels, under bridges etc. Most of the time it's open air, on a track.
It's different for WRS, where you're sitting inside a car.
MaX.

Sorry to disagree on this. I think these effects are crucial for a realistic sound (apart of good samples and sounds of course)
No need to be inside a tunnel to hear them.
I would say these are open air tracks, and imo the effect is quite noticeable:

https://www.youtube.com/v/zT0hX7fBKl8

https://www.youtube.com/v/gT4XEdydo2E

[Hawk]

+1 Warlock! 

I agree, sound including ambient sound is a VERY important part of simulation to enable immersion. I point in particular the difference between the sound for the onboard rider to the sound heard from another rider who is close by. The lack of quality(the lack of ability to simulate ambient sounds over different channels in GPB at different levels) is partly the reason we currently get engine sound washouts in GPB when other riders are near. So sound and ambient sounds correctly simulated are VERY important in a simulation.

Hawk

[HornetMaX]

But for GPB I don't think ambient-specific processing is very important: we don't ride too much in tunnels, under bridges etc. Most of the time it's open air, on a track.
It's different for WRS, where you're sitting inside a car.
MaX.

Sorry to disagree on this. I think these effects are crucial for a realistic sound (apart of good samples and sounds of course)
No need to be inside a tunnel to hear them.
I would say these are open air tracks, and imo the effect is quite noticeable:

But your examples are from spectator point of view, I was talking for the rider (so either rider pov or 3rd person "chase" pov).
For spectator pov yes, more advanced sound processing may bring something. But do we really care ?

I agree, sound including ambient sound is a VERY important part of simulation to enable immersion. I point in particular the difference between the sound for the onboard rider to the sound heard from another rider who is close by.

GPB already has positional sound. For bikes and rider/chase view, there's not a lot one can add.

Implementing some sort of "ambient occlusion for sounds" (i.e. the effect of a nearby building on the sound) is a fairly complex stuff (and may eat up some significant cpu).
For some games it's interesting (e.g. FPS), not sure for us.

One thing that may be of interest is having multiple sound sources on the bike (e.g. exhaust, engine, airbox): bt can you imagine how difficult it will be to create a proper sound then ?

I agree sound is very important (no doubt) and I agree sounds in GPB seems to be a bit below par to some other sims, but I have no idea what could be improved aside having better samples and sound-knowledgeable people to work on them.

The lack of quality(the lack of ability to simulate ambient sounds over different channels in GPB at different levels) is partly the reason we currently get engine sound washouts in GPB when other riders are near.

What do you mean exactly with "different channels" ?

MaX.

[Hawk]

GPB already has positional sound. For bikes and rider/chase view, there's not a lot one can add.

Then why do we get the same engine sound on-board as off-board and same engine sound drowning out your bike engine sound from another bike? Surely this would not happen with true positional sound, as in because the sounds are coming from different positions relative to your own they would sound different; if they did then that would make a huge difference in realism of the sound.

Implementing some sort of "ambient occlusion for sounds" (i.e. the effect of a nearby building on the sound) is a fairly complex stuff (and may eat up some significant cpu).
For some games it's interesting (e.g. FPS), not sure for us.

One thing that may be of interest is having multiple sound sources on the bike (e.g. exhaust, engine, airbox): bt can you imagine how difficult it will be to create a proper sound then ?

I agree sound is very important (no doubt) and I agree sounds in GPB seems to be a bit below par to some other sims, but I have no idea what could be improved aside having better samples and sound-knowledgeable people to work on them.

I would rather have to learn how to create complicated sound files in GPB than have to make do with sound that just doesn't sound correct. 

What do you mean exactly with "different channels" ?

MaX.

I mean being able to have separate channels of sound that when mixed and synced properly create the sound your looking for. Like real music composers do when they create music on a pc. Because creating an engine/bike sound is very much the same as creating a piece of music... Well that's the way I see it. I mean having different sounds for each engine cylinder firing, exhaust emmision and carburettor inductions, the spinning of the dry clutch assembly, even the sound of the chain spinning.... it's all in there as part of the engine/bike noise and vibration sounds varying under different loads depending on throttle position and load of the engine. Complicated yes, but to someone who knows what they are doing in creating sounds and using the correct software to create these sounds it is very possible; even to break down these component sounds from within a real sound sample is possible given the software to do it. From there you could then begin to fill-out and build the sound your looking for before implementing it into GPB.

Granted Piboso would probably/maybe have to totally redesign the way GPB calls on sounds to do this? But I think the realism that would result would be well worth the effort. Would this overload the CPU? I doubt it given the power of CPU's these days. But some of the CPU work in GPB currently could possibly be offloaded onto the GPU, because I'm sure the CPU is doing graphical work that todays GPU's are design and able to handle to save CPU time. Or is that a limitation of OPENGL in that the graphics work is CPU intensive compared to what Direct X is capable of with offloading a lot of graphical work through the GPU? I really don't know but that would be interesting, but that's another subject altogether. Lol

Hawk.

[HornetMaX]

If it gets that complicated, nobody will be able to create sounds.

By the way, GPB already has a separate sound for "transmission whine", as it has sounds dedicated to tyres squeaking and brakes whining.

And the sound in GPB is positional no doubt: just take a spectator view and listen to a bike pass by.

CPU cost would be a hit only if someone aims for ambient sound occlusion: having 2 or 4 extra "channels" to mix is not a problem.

Anyway, looking at what's happening on KRP forums after beta13 release, I guess Piboso has other stuff to take care of at the moment. And way more important than sound, unfortunately.

MaX.

[Hawk]

If it gets that complicated, nobody will be able to create sounds.

By the way, GPB already has a separate sound for "transmission whine", as it has sounds dedicated to tyres squeaking and brakes whining.

And the sound in GPB is positional no doubt: just take a spectator view and listen to a bike pass by.

CPU cost would be a hit only if someone aims for ambient sound occlusion: having 2 or 4 extra "channels" to mix is not a problem.

Anyway, looking at what's happening on KRP forums after beta13 release, I guess Piboso has other stuff to take care of at the moment. And way more important than sound, unfortunately.

MaX.

Oh I know when watching in replay from a spectator point of view you can definitely hear the difference(I think they call it the Doppler effect?), but when on-board the difference is basically nil, and that is not correct. Even the sound from spectator view after the bike has gone paste you just doesn't sound right(it's too dulled), even 2 stroke engines sound like a four stroke. I just think that the sound setup is probably what I'd call a beta setup and hasn't been fine tuned at all as yet? But then maybe this should be obvious to us all with GPB still in beta format. 

But yeah, agreed for sure there are many things in GPB that are way up the TODO list rather than this sound issue, but it's always good to discuss anything that riders would like to see improved in GPB for the future. 

Hawk.

[Warlock]

But your examples are from spectator point of view, I was talking for the rider (so either rider pov or 3rd person "chase" pov).
For spectator pov yes, more advanced sound processing may bring something. But do we really care ?

I do care  , i can imagine race videos with espectacular sounds , that could make me drowl  , ...me and potential GPB buyers.
Yes i talk about spectator pov, but it would also affect onboard
Why people do go to IOM as spectator,......those sounds are just music for the ears 

Indeed this is more a wish than anything else, because we all know it's not going to happen before 2030 

Anyway, i don't know how this sound engine things works, or how to implement a new one. i would say its a library of classes and functions similar to any other programming language and a tool to compile them or similar, don't know.
Also don't know if GPB uses a comercial one, or custom designed.

[Warlock]

Very interesting interview about sound development i found:
They even talk about REV , a tool i discovered some time ago ( that captures a sound ramp, based on granular synthesis. We talked about this in other posts in this forum) and its integration in Wwise sound engine
http://designingsound.org/2014/08/vehicle-engine-design-project-cars-forza-motorsport-5-and-rev/

[HornetMaX]

I do care  , i can imagine race videos with espectacular sounds , that could make me drowl  , ...me and potential GPB buyers.

Yeah yeah, true. But I don't want to watch GPB replays: if I have to watch I can watch real races

Also don't know if GPB uses a comercial one, or custom designed.

AFAIK, GPB uses openAL (http://en.wikipedia.org/wiki/OpenAL). It's a generic thing (i.e. it's not dedicated to racing and not even to games), but it's free and cross-platform, I guess these 2 characteristics sealed the deal with Piboso.

Very interesting interview about sound development i found:
They even talk about REV , a tool i discovered some time ago ( that captures a sound ramp, based on granular synthesis. We talked about this in other posts in this forum) and its integration in Wwise sound engine
http://designingsound.org/2014/08/vehicle-engine-design-project-cars-forza-motorsport-5-and-rev/

Very interesting indeed. But if I read correctly what Greg is saying in the 1st part, all of the new stuff is used mostly in sound creation: the implementation is still mostly loop based, like GPB.

Which more or less was my point: put some black magic in the samples and GPB can sound good (maybe not stunning, but much better than what we have today on most of the mod bikes, some of which I can't ride because my ears start pissing blood after 500 meters ...). Translated: get good raw samples (note to some: you'll never succeed ripping from a youtube video), use REV (or whatever) and then slam the samples into GPB.

Interesting also what Stephen Baysted said about 1st person view: as I thought, most of the time you just filter (comb + low-pass) the "normal" sound to account for the earplug and helmet.

As you can see, nobody mentioned real time occlusion. Just googling it, I stumbled on this: http://docs.cryengine.com/pages/viewpage.action?pageId=18384659
So it is doable but still not sure it would give us alot (for an FPS, it does of course).

[side thoughts: cryengine can be "bought" by developers for 10$/E a month and a Linux version is apparently under development].

MaX.

[Hawk]

I do care  , i can imagine race videos with espectacular sounds , that could make me drowl  , ...me and potential GPB buyers.

Yeah yeah, true. But I don't want to watch GPB replays: if I have to watch I can watch real races

Well certainly I'd like to see GPB get to a quality were when races are streamed live, people would like to watch the GPB race events almost the same as real MotoGP events, similar to what SimRacing have with their live streamed rFactor2 F1 Grand Prix's. But that won't happen till GPB V1.0 is released and further more, that we have a grid full of regular riders that are within a second of each other in race lap times(not hotlap times). Then I can see GPB really taking off and coming to life. 

But right now I feel GPB is just maybe coming out of it's embryonic stage..... It's better now than before, but not yet ready to be born into the world. But unfortunately GPB is a very slow growing project indeed. 

Hawk.

[Eagle]

I was wondering. Can we use 2 or more samples (in the scl) for the same sequence/moment ? For 2strokes that have multiple sounds with different variations, it would help for more precision.

Exemple:

- 1 sample for the mosquito sound common to all 2s.

- 1 sample for the crankshaft sound (vary depending on the cylinders number).

That's 2 samples in the scl playing on the same sequence of rpm in game, but more realistic render.

Just a thought..

[HornetMaX]

I was wondering. Can we use 2 or more samples (in the scl) for the same sequence/moment ?

Yes, of course. It's already the case when you transition from one sample to the next: one is fading out, the other is fading in, hence both are playing at the same time.

No idea if there's a limitation (max number of "simultaneous sounds").

MaX.

[Eagle]

Thanks Max !

Mmhh.. Now there's still the separated sounds to record. xd