Give us better sounds - PLEASE !!

  • Thread starter steamcat
  • 4,667 comments
  • 353,163 views
And is still a Frankenstein's monster of recycled samples - all three versions were. It does not represent PD's current philosophy on sound: simulation.

Yes, but like I said, the effort they did for it it's what I expect for final GT6 sounds on release day... anything more than that it's a bonus! I'm praying for them, to make the breakthrough in due time!

What are people's thoughts on this "completely different way" of generating sounds?

If it's what I think it is, we can look forward to much improved customisation and general variation due to tuning, as well as improved expressivity from the sounds when driving.

Honestly, I expect them going in a "mad scientist" way, like complicated calculations with RPM, amount of fuel/air in the mixture, ignition time, explosions, temperature... etc...

I know, completely crazy but if it's simulation, who knows?
 
What are people's thoughts on this "completely different way" of generating sounds?

If it's what I think it is, we can look forward to much improved customisation and general variation due to tuning, as well as improved expressivity from the sounds when driving.

I think they're gonna make it simpler, and easier to the PS3 hardware.
 
Yes, but like I said, the effort they did for it it's what I expect for final GT6 sounds on release day... anything more than that it's a bonus! I'm praying for them, to make the breakthrough in due time!



Honestly, I expect them going in a "mad scientist" way, like complicated calculations with RPM, amount of fuel/air in the mixture, ignition time, explosions, temperature... etc...

I know, completely crazy but if it's simulation, who knows?

I like the way you think ;)

Now to imagine Kaz didn't even mention sounds and readjust my expectations to suit. Where'd I leave that cricket bat..?
 
I like the way you think ;)

Now to imagine Kaz didn't even mention sounds and readjust my expectations to suit. Where'd I leave that cricket bat..?

Hehehhe I don't know, from your post history you seem to know A LOT about sounds... good question!

The goddamn INTAKE SOUNDS... if PD just add them... it would make a world of difference. Seriously, I want to buy NFSMW2012 just because of the sounds. Of course, for them it's easy... no manual transmissions, way simpler physics, etc... But listen to the sounds! Damn. The best I've seen lately. Like you told before about the E92, the intake sounds are very pronounced and it's a very important part of how this car sounds...

Of course, talking about the samples, not the sound engine itself...

 
Last edited:
Yeah that's the complete opposite of the approach PD seem to favour. Everything is hand mixed and "animated", all the expression it's capable of is only that which the designers put in.

It sounds very high quality, even though the sound doesn't match the viewpoint. But it's limited flexibility for a sim, especially one with tuning at its heart, means it just wouldn't work for GT.

A simulation approach requires putting in the appropriate physical parameters and letting the system run, so that all the possible expression is a function of your interaction with that system.

While I don't expect PD to go that far, the driveline modeling in iRacing is a perfect example of sound simulation, even though it was meant more for the driving experience. There should be a lot more of that sort of thing going forwards.
 
When I think of bad engine sounds in GT, I think of hardware limitation..
They need to maintain the gameplay at 60 fps with only a PS3, and as much as possible opponents on the track.
More cars means more sound channels to be mixed, and if they used too many samples, I don't think the PS3 could handle it. they need the RAM for other "more important things" like textures, and everything, they need the Processing power for better fps, more opponents, etc..

That's why when Kaz said about "completely different way of generating sounds", I think they just wanted to make it less complicated to the hardware but sounds better than the current.
 
I just got the surround sound in my basement to start working. I started up gt5 and changed to a Lambo and a 787B and HOLY CRAP. Surround sound makes a world of difference. You can feel the rumble (literally) and at low RPMs every car I have tried has sounded amazing. The sounds really aren't that bad now lol xD
 
What are people's thoughts on this "completely different way" of generating sounds?

Given that they couldn't master the "normal" way of creating sounds, I don't have high hopes for a good result from them trying anything fancy.

Best of luck to them, but you don't expect As from the kid that's been getting C-s just because he's now in the Advanced Mathematics class.
 
Your student analogy is ill thought out, but I appreciate the point.

PD have not demonstrated (evidence of) what you consider to be mastery (whatever that is) of the "normal way" (sample based synthesis) of making sounds. But you might find it interesting to know that this method is quite abstract; it has no physical basis.

Simulation, on the other hand, is very different in that it is immediately relatable to the real thing. Of course, the accuracy of that depends on the model they're developing, and it remains to be seen exactly how they're going to break it down.

Given their approach to the driving physics improvements (sound is also physics), I'd say it should be very interesting.

EDIT: some related reading; "next gen sound synthesis"
http://www.ness-music.eu/
 
Last edited:
Your student analogy is ill thought out, but I appreciate the point.

PD have not demonstrated (evidence of) what you consider to be mastery (whatever that is) of the "normal way" (sample based synthesis) of making sounds. But you might find it interesting to know that this method is quite abstract; it has no physical basis.

Simulation, on the other hand, is very different in that it is immediately relatable to the real thing. Of course, the accuracy of that depends on the model they're developing, and it remains to be seen exactly how they're going to break it down.

Given their approach to the driving physics improvements (sound is also physics), I'd say it should be very interesting.

EDIT: some related reading; "next gen sound synthesis"
http://www.ness-music.eu/

Could it be related to the news of PD collaborating with other companies for GT6 ? Maybe a sound studio ?
 
Sony has enough sound studios...

True, but in the past PD have been in house with sound design, maybe this time with GT6 they decided to cooperate with one of the Sony's owned sound studios - maybe for technical assistance with sound simulation or field recordings ? I really hope this is the case, as it will undoubtedly improve the sound quality.
 
Am I only one who cant stand tire squealing?. Engine sounds are fine as long as I could mute the squealing.

Don't push the car so hard then

--

edit

simulated sound - advantage over samples, it varies to load and drive train conditions


 
Last edited:
Your student analogy is ill thought out, but I appreciate the point.

PD have not demonstrated (evidence of) what you consider to be mastery (whatever that is) of the "normal way" (sample based synthesis) of making sounds. But you might find it interesting to know that this method is quite abstract; it has no physical basis.

Simulation, on the other hand, is very different in that it is immediately relatable to the real thing. Of course, the accuracy of that depends on the model they're developing, and it remains to be seen exactly how they're going to break it down.

Given their approach to the driving physics improvements (sound is also physics), I'd say it should be very interesting.

EDIT: some related reading; "next gen sound synthesis"
http://www.ness-music.eu/

The short version is that given that they have not yet managed to create good sounds by any method, we shouldn't be holding our breath just because they're pursuing some new method.

Presumably, said new method is difficult in some manner, or everyone else would be doing it. Also presumably, sample based synthesis has something going for it or every man and his dog wouldn't be using it.

Not that it's not cool that they're trying something new. But if I was going to put money on someone to bring forth a new method of sound creation, it'd be on someone who has demonstrated skill with the current art. Not because the skills would transfer, if it's a truly new approach the team is going to be starting from scratch either way. Simply because someone who has already accomplished decent sounds using samples has demonstrated the skill to make the sounds happen. Be that in a technical manner, or in a political manner (making sure his team has the time and resources necessary, that the game can support the sounds and they're not being cramped by other departments, etc).

Give PD 10 years, and they're probably as capable as the next team. Ask them to deliver by November, and the track record says it's probably not going to happen.
 
Great video Diegoborges. NFS has some really cool samples and a "simple" way to make car engine sounds sound realistic and spectacular in some ways. I like how they do it. At least from outside and inside tunnels.

I just add this one. Imagine if GT6 sounds like this...that Gallardo inside the tunnel is incredible.

 
Foxiol
Great video Diegoborges. NFS has some really cool samples and a "simple" way to make car engine sounds sound realistic and spectacular in some ways. I like how they do it. At least from outside and inside tunnels.

I just add this one. Imagine if GT6 sounds like this...that Gallardo inside the tunnel is incredible.

<a href="http://www.youtube.com/watch?v=PxSXTle0uq4">YouTube Link</a>

I know, right? Nfs mw has the best engine sounds I've ever heard in a game.

GT6 with those sounds would be an orgasmic game for me than.
 
This is from pCARS, and is just now being implemented:

Volumetric Throttle System &#8211; This is a new feature that calculates torque output using an approximated manifold pressure. Side effects include air restrictor effects, possible better metrics for sound system to use, fun engine response changes with the atmosphere among other things. One of the things it will also do is change the 'neutral' throttle position - where the engine is producing zero net torque - to be much lower in the pedal's travel. You'll have much better precision over the torque and the sensitivity curve will change naturally through the rpm range. A great side effect is that you get much more of a feel for the clutch bite point from standstill. Because you have a certain amount of air going in to the engine at idle and air is directly related to torque, as the clutch starts to bite and rpm drops the air available produces more torque.

I wonder if this is something PD is also investigating.
 
I am really pleased with his answers. Even if the process of creating sounds differently doesnt fully make it to gt6, they at least acknowledged the issue.

On top of that, I am very curious to see what may be different considering he said those were basically just placeholder sounds at e3. Cant wait and this news gives me that much more hope and excitement for gt6.

If they can give my Corvettes/Mustangs some proper meaty v8 sounds, I'll be one happy camper.
 
Research is one thing, actually implementing them is another.

Guess I'll have to have the game first before I can make any concrete judgement on it.
 
This is from pCARS, and is just now being implemented:



I wonder if this is something PD is also investigating.

That, in my opinion, is a perfect example of the non-physical nature of sample based synthesis. All you need is rpm and throttle position (perhaps boost if your samples cover that). If your model doesn't work with those inputs, it's not an accurate model - not to mention torque variations are already accounted for in the samples.

I doubt many teams are more suited to this task than PD, as most teams use an absracted sampler like FMOD that handles all the technical aspects and housekeeping, leaving more or less a purely artistic task. Except that many of the controls have no direct relation to how the real sounds are made.

PD's sound engine has been different in almost every game - they're well used to difficulty. The beauty of a physical system is that it should be easier to use intuitively as all the controls are real physical things.
 
The only thing illuminating from that talk are the budgets used for the games he demo'd. Everything else is well known to anyone in the field; he seems to be referencing music tropes, so his audience must be mixed.

I like how T10 are thinking about speed of sound modeling and HDR mixing, both of which were in GT5:P. I've always said that T10 favoured a Hollywood approach, and now it seems it's an actual target, even though "they don't sound like real cars".

I don't agree that you should mix to the worst hardware ("lowest common denominator"), either. If the same applied elsewhere, driving aids would always be on. Simulate the sound and add accessibility options for the casuals.

I enjoyed his crazy particulate sampler system for dirt noises, though, and his general enthusiasm and disdain in the right places!
Now to see what Linear Predictive Coding is and how it compares to other direct synth methods, assuming that's what it be.
EDIT: Source-filter model.
 
Last edited:
Back