Sound Update (PD is now hiring! + Email!)Answered

[smitty33]

I'm at a loss to understand what so many people want as far as sound is concerned. It must be near impossible to make every car sound exactly right

 

[Griffith500]

Quite; mostly because it's impossible to make only one car sound exactly right!

But I think we all have lower standards than that anyway.

 

[MLRSparco]

A simple answer would've been fine,no one asked for the history of how terrible the sounds were recorded.Any other company would've came out and gave us some type of insight on what they were planning for the sounds like X amount of cars will get the update in X amount of time,or we plan on only updating the premium cars rather than standards,but PD always wants to drag something out for no reason.

 

[MGR]

Recordings schmordings. An engine rev'd in neutral captured with a clip tie microphone could deliver better sounds than what's currently output in game.

The answer lies within Polyphony's 'Synthesising Process' and design concessions made to ensure the game can run on the designated hardware.

 

[Johnnypenso]

A simple answer would've been fine,no one asked for the history of how terrible the sounds were recorded.Any other company would've came out and gave us some type of insight on what they were planning for the sounds like X amount of cars will get the update in X amount of time,or we plan on only updating the premium cars rather than standards,but PD always wants to drag something out for no reason.

If you know, "PD always wants to drag things out for no reason", and then they go ahead and do so, why complain about it?

Recordings schmordings. An engine rev'd in neutral captured with a clip tie microphone could deliver better sounds than what's currently output in game.
The answer lies within Polyphony's 'Synthesising Process' and design concessions made to ensure the game can run on the designated hardware.

Which has been discussed to death and is the reason most of us believe there won't be any big changes in sound modeling until PS4.

 

[MLRSparco]

If you know, "PD always wants to drag things out for no reason", and then they go ahead and do so, why complain about it?

Which has been discussed to death and is the reason most of us believe there won't be any big changes in sound modeling until PS4.

Because complaints gets things done as you can see,if we don't complain nothing happens why people get mad because we want the game to be better or want the company to be better is beyond me.

 

[Griffith500]

A simple answer would've been fine,no one asked for the history of how terrible the sounds were recorded.Any other company would've came out and gave us some type of insight on what they were planning for the sounds like X amount of cars will get the update in X amount of time,or we plan on only updating the premium cars rather than standards,but PD always wants to drag something out for no reason.

The answer to your particular question is: they'll all be updated at the same time, when it's ready.

PD are ready to start talking about the sound, officially, so that's good news in my book.

 

[Johnnypenso]

Because complaints gets things done as you can see,if we don't complain nothing happens why people get mad because we want the game to be better or want the company to be better is beyond me.

They are in the midst of answering the question as we speak. What else do you want done? How about we wait until Kaz is finished typing, then take up arms if we aren't satisfied with his full response?

 

[Teflonicus]

I don't see how this question can be "answered" when no-one, outside of PD, knows what the actual answer is.

I look forward to reading part three though, and I hope the information it contains hints to greater things for the series. I am a little surprised that we have to wait for the question about the issue that we've already been waiting on to be answered though.

 

[stuff1138]

They are in the midst of answering the question as we speak. What else do you want done? How about we wait until Kaz is finished typing, then take up arms if we aren't satisfied with his full response?

Not to mention the fact that Kaz was racing for 24 hours in Germany this past weekend and it's only Tuesday.

 

[VBR]

Not to mention the fact that Kaz was racing for 24 hours in Germany this past weekend and it's only Tuesday.

Of course, his hobbies are much more important than answering any concerns his customers have about his game, especially concerns that are years old & have been carried from GT5 to GT6. Racing first, answer questions later, MUCH later!

 

[stuff1138]

Of course, his hobbies are much more important than answering any concerns his customers have about his game, especially concerns that are years old & have been carried from GT5 to GT6. Racing first, answer questions later, MUCH later!

Haha well Hobby or not I'm sure he was pretty busy.

 

[MLRSparco]

They are in the midst of answering the question as we speak. What else do you want done? How about we wait until Kaz is finished typing, then take up arms if we aren't satisfied with his full response?

A game plan would've been nice I mean they have been working on this since before GT6 was launched .All I'm saying is that his first post didn't exactly fill me with confidence,we still have no idea if it'll be for GT6,if they don't have an answer then they don't have an answer but don't pull or chains and drag us along feeding us bread crumbs.We asked a question and I would've hoped the question he chose to answer he would do it in one fell swoop.

 

[TayeezSA]

Haha well Hobby or not I'm sure he was pretty busy.

R&D with added fun

 

[TayeezSA]

I've just thought of something. With this new Sound update, how will it affect mods like the Exhaust for example? GT as of late in my opinion can really butcher the sound of the exhaust if you mod it.

So in this case, what does PD have planned?

I believe it may work like this:


Factory Exhaust - Pure factory exhaust and engine note. I think this sample will just be taken and modified upon.

Sports Exhaust - Minimal Volume increase,More Bass with light burbles along with pops than the Factory exhaust

Semi-Racing exhaust - Significant volume increase, balanced bass and treble, light exhaust drone, more pops burbles and bangs.

Racing Exhaust - Drastic Volume increase (Not enough to shatter eardrums though) , More gritty treble over bass but both are increased, heavy exhaust drone and turbo whistle. Also heavy pops, burbles and louder bangs. Essentially a straight pipe.

I think that would a good setup for the Exhaust model plus it saves plenty of time which means a sooner update.

 

[stuff1138]

So I guess after the 24hrs of the Nurburgring they had to stop off at the FIA thing in Munich...That explains the extended delay on answering the sound question completely.

 

[Speedster911]

PD promised us to gradually update the sound before the launch date. To be fair some does sound realistic from the GT5 although lacking some "effects". But for the update to GT6, all I see is just a very few numbers of it. Where's the rest?

I mean for the example, most of the new cars in GT6 sounds just like PD left the "placeholder" (CopyPaste) sound present from the Demo.
PD said it just a "placeholder" to be replaced in the update.
More noticeable at cars like M3 GT, Huayra, Quattro Rally Car, and more.

That's what I'm worried about. Where is it? When we finally get an update for the sound? If PD can't update a small group of cars, how can PD provide for the rest?

(Lotus 97T is the best example of upgraded sound recording. Im sure PD can apply it to other cars 👍)

UPDATE: Thank you all for voting this question until this become No.1. As for 19 June, Kaz post a blog addressing this issue, which means this is the first question to be answered.. Be patience because part three / complete answer of the post will be coming shortly.

You can see it here: pitstop.gran-turismo.com/en/article/26

Well, it's a start I guess.
I will eat my hat if PD fixes the sound completely in the PS4 version of GT. I might even drop the idea of picking up an X1 just for F5.

As far as the article goes, PD's approach to sound recording is just wrong. Neither do I think T10's approach of putting it on a dyno to record the sounds, is any better, even though it does produce some eerily realistic results.

I know this is cumbersome and very time consuming, but it's definitely worth a try: take EVERY car to a drag strip that's well over a mile long. Record the car while it's actually moving in all the gears. Engines sound different when under stress and inertia. This way, a lot of elements can be captured accurately, notably the engine whine and exhaust rumble as well as pops, crackles and sputtering, between gear shifts.

Engines sound a little different too as you ease off the throttle; you hear and feel that exhaust rumbling and the engine "whirring and whining" as it takes a breather. Finally, every car should be put in neutral and revved through the entire range to get the idle sound right, in addition to getting that sweet "exhaust snap" and "engine growl" right, as you're revving it up in neutral.

PD, please give these things serious consideration. Start with less cars initially, so that you guys get the performance, handling and SOUND characteristics down to the point of perfection. Is it impossible? Certainly not. Can it be done? YES, but it would require a lot of time and energy on your part.

I would be gladly coming back to GT if it sounds like a proper driving/racing simulator too.

Start small, release more cars as they're modeled accurately. Quality over quantity ought to be one of the company's motto by now.

Just my two guys, just my two. I want GT to give me that same excitement and thrill when it was first released and how much it improved up until GT4.

 

[TayeezSA]

Well, it's a start I guess.
I will eat my hat if PD fixes the sound completely in the PS4 version of GT. I might even drop the idea of picking up an X1 just for F5.

As far as the article goes, PD's approach to sound recording is just wrong. Neither do I think T10's approach of putting it on a dyno to record the sounds, is any better, even though it does produce some eerily realistic results.

I know this is cumbersome and very time consuming, but it's definitely worth a try: take EVERY car to a drag strip that's well over a mile long. Record the car while it's actually moving in all the gears. Engines sound different when under stress and inertia. This way, a lot of elements can be captured accurately, notably the engine whine and exhaust rumble as well as pops, crackles and sputtering, between gear shifts.

Engines sound a little different too as you ease off the throttle; you hear and feel that exhaust rumbling and the engine "whirring and whining" as it takes a breather. Finally, every car should be put in neutral and revved through the entire range to get the idle sound right, in addition to getting that sweet "exhaust snap" and "engine growl" right, as you're revving it up in neutral.

PD, please give these things serious consideration. Start with less cars initially, so that you guys get the performance, handling and SOUND characteristics down to the point of perfection. Is it impossible? Certainly not. Can it be done? YES, but it would require a lot of time and energy on your part.

I would be gladly coming back to GT if it sounds like a proper driving/racing simulator too.

Start small, release more cars as they're modeled accurately. Quality over quantity ought to be one of the company's motto by now.

Just my two guys, just my two. I want GT to give me that same excitement and thrill when it was first released and how much it improved up until GT4.

Which gives me a less tedious idea and made me remember a bit of high school science. Funny enough I thought of this while riding a bicycle

Horsepower is the rate at which work is done correct? Now the best way to record exhaust and engine audio under load would be best when the car is doing work. Unfortunately in real world conditions like on a dragstrip would be unfavourable due to wind and other factors.

So since PD has the idea so far to use a Dynapack, maybe it would be best to add mass in the mix. Mass in the sense of modifying the Dynapack in such a way that when the engine is under load, a mechanism transfers power from the car on the Dynapack to lift up weights which makes up the test car's total mass so it simulates real world conditions of an engine under load. When an engine is under load, it's essentially doing work to pull its own mass yes? That way you can get a very close to accurate recording of an engine under load in favourable conditions.

I dunno if that came out properly but you guys should get the idea.

 

[stuff1138]

Which gives me a less tedious idea and made me remember a bit of high school science.

Horsepower is the rate at which work is done correct? Now the best way to record exhaust and engine audio under load would be best when the car is doing work. Unfortunately in real world conditions like on a dragstrip would be unfavourable due to wind and other factors.

So since PD has the idea so far to use a Dynapack, maybe it would be best to add mass in the mix. Mass in the sense of modifying the Dynapack in such a way that when the engine is under load, a mechanism transfers power from the car on the Dynapack to lift up weights which makes up the test car's total mass so it simulates real world conditions of an engine under load. When an engine is under load, it's essentially doing work to pull its own mass yes? That way you can get a very close to accurate recording of an engine under load in favourable conditions.

I dunno if that came out properly but you guys should get the idea.

I think actually resistance would be better than mass. Since it's the friction of the tires gripping the road, the air/wind, gravity, etc. that causes the resistance against the car. So if there was some way to add resistance to the Dynapack that would make it closer to reality. Think of it kind of like an stationary bicycle where the pre-programmed system adds resistance when you are virtually going up a hill.

 

[TayeezSA]

I think actually resistance would be better than mass. Since it's the friction of the tires gripping the road, the air/wind, gravity, etc. that causes the resistance against the car. So if there was some way to add resistance to the Dynapack that would make it closer to reality. Think of it kind of like an stationary bicycle where the pre-programmed system adds resistance when you are virtually going up a hill.

Mass and Resistance would be great. I wonder how they'll pull it off in terms of resistance though. Mass I rate would be the most sensible and easiest.

On a side note.... Can we please have the option to change between Equal Length and Unequal Length Headers for Subaru. I mean listen to how sick this sounds with Unequal Length Headers

 

[Griffith500]

What if I told you there is a way to account for the differences in the video below, without using a single recording (other than for reference)?



So the stock sound at the start (0:18), then there's the Capristo mid pipe and muffler system (just louder and excessively raspy) with X-pipe pull at 6:12, then the one without the X-pipe at 7:45. The stock car does not have an X-pipe.

You can account for the differences in headers / manifolds in the same way, also, including the boxer throb.
For example you can do this, or this, or this, or this, or this, or this etc. without having to record them especially, once you had a working "model" for the engine in question, or even just engines in general.


You cannot really dynamically faff with things like rasp or X-pipes and manifold changes that easily using the same samples. Any "system" that can, would be built on ideas that would work for synthesis from first principles, without using samples at all. But you're still stuck with the overall "colour" of the samples you started with, one way or the other. I know what I would, and did, do.

Going the "forward" route of synthesis like that allows for infinite variation, and hence customisation. The only thing stopping us would be the control interface required to harness that "infiniteness", but it could be simplified, and greatly if necessary; we could even still just have "presets" provided by PD, just like with the current exhaust upgrades.

 

[TayeezSA]

What if I told you there is a way to account for the differences in the video below, without using a single recording (other than for reference)?



So the stock sound at the start (0:18), then there's the Capristo mid pipe and muffler system (just louder and excessively raspy) with X-pipe pull at 6:12, then the one without the X-pipe at 7:45. The stock car does not have an X-pipe.

You can account for the differences in headers / manifolds in the same way, also, including the boxer throb.
For example you can do this, or this, or this, or this, or this, or this etc. without having to record them especially, once you had a working "model" for the engine in question, or even just engines in general.


You cannot really dynamically faff with things like rasp or X-pipes and manifold changes that easily using the same samples. Any "system" that can, would be built on ideas that would work for synthesis from first principles, without using samples at all. But you're still stuck with the overall "colour" of the samples you started with, one way or the other. I know what I would, and did, do.

Going the "forward" route of synthesis like that allows for infinite variation, and hence customisation. The only thing stopping us would be the control interface required to harness that "infiniteness", but it could be simplified, and greatly if necessary; we could even still just have "presets" provided by PD, just like with the current exhaust upgrades.

Why do I have a feeling that Griffith studied Sound Engineering? That's very interesting and it makes sense at the time.

What do you have to say about using mass to replicate real world conditions though?

 

[Griffith500]

Why do I have a feeling that Griffith studied Sound Engineering? That's very interesting and it makes sense at the time.

What do you have to say about using mass to replicate real world conditions though?

I taught myself, because I "needed" to know how it all works!

I don't really see the difference between work done on an eddy current dyno, vs. work done on a water brake, vs. work done on the road. What matters for sound is throttle position and rpm, then the acoustics of the recording environment: the drier the better.

That said, you can use a big flywheel to "simulate" the mass of the car, but that's only workable with the engine actually out of the car (directly couple the engine to the flywheel). It's best used for dynamic situations if you're going to go to that trouble; for recording, you usually want static conditions anyway (unless you're using granularly resynthesised "ramps" instead of traditional samples, but a dyno pull is perfect for that).

 

[smitty33]

Do you really want PD to do all this work 1200 times to improve the sound marginally. What a waste of resources when there are much more important issues that need to be resolved.

 

[Robbks]

I think actually resistance would be better than mass. Since it's the friction of the tires gripping the road, the air/wind, gravity, etc. that causes the resistance against the car. So if there was some way to add resistance to the Dynapack that would make it closer to reality. Think of it kind of like an stationary bicycle where the pre-programmed system adds resistance when you are virtually going up a hill.

Ummm,
the dynapck (or pretty much every other) dyno is able to simulate all of that..?

They have a setting called a "Ramp Rate" which is able to restrict how fast or slow the engine can accelerate.
So they can take multiple passes at different ramp rates to simulate any load conditions they wish.

the only small issue they face is that they don't allow a vehicle to accelerate at it's natural rate.
Some parts of the rev-range it will accelerate faster (peak torque) and other areas it will accelerate slower.
But it's teh same a putting a slight incline or decline on a piece of road.

 

[1241Penguin]

Do you really want PD to do all this work 1200 times to improve the sound marginally. What a waste of resources when there are much more important issues that need to be resolved.

There aren't 1200 cars with unique sounds though. R34 GT-R's for example can make use of the same sound samples. Same goes for first generation STI's, all NSX's, RX-7's, and so on.

 

[Johnnypenso]

Do you really want PD to do all this work 1200 times to improve the sound marginally. What a waste of resources when there are much more important issues that need to be resolved.

There are a multitude of issues that is true, and I doubt anyone thinks this is the top priority. However, sound engineers don't model cars nor write game code nor do anything but work on sound in the game so whether this is worked on or not shouldn't have any effect on the development of the rest of the game.

 

[Speedster911]

Which gives me a less tedious idea and made me remember a bit of high school science. Funny enough I thought of this while riding a bicycle

Horsepower is the rate at which work is done correct? Now the best way to record exhaust and engine audio under load would be best when the car is doing work. Unfortunately in real world conditions like on a dragstrip would be unfavourable due to wind and other factors.

So since PD has the idea so far to use a Dynapack, maybe it would be best to add mass in the mix. Mass in the sense of modifying the Dynapack in such a way that when the engine is under load, a mechanism transfers power from the car on the Dynapack to lift up weights which makes up the test car's total mass so it simulates real world conditions of an engine under load. When an engine is under load, it's essentially doing work to pull its own mass yes? That way you can get a very close to accurate recording of an engine under load in favourable conditions.

I dunno if that came out properly but you guys should get the idea.

Brilliant! I hope they borrow your idea at some point. Bad sound really robs away from the experience, whether it's a proper sim or just a driving game.

 

[TayeezSA]

Ummm,
the dynapck (or pretty much every other) dyno is able to simulate all of that..?

They have a setting called a "Ramp Rate" which is able to restrict how fast or slow the engine can accelerate.
So they can take multiple passes at different ramp rates to simulate any load conditions they wish.

the only small issue they face is that they don't allow a vehicle to accelerate at it's natural rate.
Some parts of the rev-range it will accelerate faster (peak torque) and other areas it will accelerate slower.
But it's teh same a putting a slight incline or decline on a piece of road.

You learn something new everyday, problen almost solved

 

[Griffith500]

Do you really want PD to do all this work 1200 times to improve the sound marginally. What a waste of resources when there are much more important issues that need to be resolved.

They'll already have been doing it, "for the last few years". Content production is not the issue here.