Predict: The Next Generation Console Tech - AUDIO

Well, that's GT6 so I'm not too surprised the sound is still like that. With more memory hopefully they'll add more details. Gear changes sound weak, everything is too linear sounding and pitch shifty with no randomness to it.
 
I doubt GT6 will sound like the real counterpart but it is unfair to compare an offscreen video with actual real life footage

The source of the GT6 video is this one:

http://www.youtube.com/watch?v=2WZxx4XTzFI

And although the video was captured off-screen, the audio was captured more or less via direct-feed according to the video description, see:

GTPlanet said:
[...]

Sound was recorded as single-channel (mono) audio from TV's headphone jack. It sounded better in stereo from the TV's speakers.

[...]


;)
 
Well, to be fair, in GT5, with a BIG home cinema system, you could have really nice sound (put it into "big theater" mode and turn UP the volume).

Though the in car sounds still sound muffled, the outside audio is GREAT. Also in replays. Never understood that. My stock Mazda sounds louder than this.
 
When you start a car it always sound great, it's the loops that are ugly and short. (I like the Ford GT sound though)

I hope more memory and sound processing power will help them make better synthesized sound, or at least more interesting composite sound which don't feel like cheap loops. So maybe we'll have to wait for the PS4 version.
 
I hope more memory and sound processing power will help them make better synthesized sound, or at least more interesting composite sound which don't feel like cheap loops. So maybe we'll have to wait for the PS4 version.


IMHO, Need for Speed: Most Wanted (2012) for example has much better engine sounds than Gran Turismo (especially in tunnels). And it's running on PS3, not PS4.

Just listen to the following three videos for example:



[YT]0CN9x31XAAw[/YT]



[YT]a5jBI9ECiGs[/YT]



[YT]HspYzoVamBs[/YT]



That is worlds apart to what Gran Turismo sounds like IMHO.
 
^ Good examples. Though I wonder how much of it is due to not having to have a dozen or more cars on the same track at any given time like GT does.

In all fairness, GT probably had to simplify their sound capture process because of the sheer amount of cars in their game, whereas NFS can focus on perfecting a much smaller list. Still, I'm not sure it's a good excuse anymore. Their capture process and sound engine needs definitely needs improvement.

Here's a good article on the guys that were behind the sound of NFS.

Over the years we’ve changed our recording techniques in search of methods which made the cars come alive. We used to record cars by placing them on on a dynamometer, but this gave us a pretty “stale” lifeless sound (perfect for programatic, scientific cars though :) Now most of our recordings are done on a track, with dyno recordings filling in some gaps. We employ some of the best car recordist to ensure we get the result we need as there’s only 1 opportunity at times. As always it’s a team effort.

For NFS we’ve recorded 150+ cars, anything from regular to high end 1 million dollar race-cars. Each of those is a new experience as you always have to find the optimum recording spots. Besides that we’ve done car crash recordings in multiple ways, several sessions of skid recordings, road surfaces, bumps, turbos, rattles etc. You name it on a car, and we probably recorded it.

The bolded couldn't be more true :smile: FWIW I think both GT and Forza adopted the dyno recording method. But I wouldn't be surprised if GT only captured engine revs for a lot of their cars done earlier and just reused and tweaked those sounds.

I'm afraid though that PD would basically have to go back and re-record all the cars. Hopefully they're doing that before GT7. You can't polish a turd... the source samples need to be good.
 
What seems to be really missing is the aerodynamic noise, which would face an uncanny valley if pre-recorded. Some of that wind+gust technology in KZSF coupled to the audio engine might make for some very realistic wind noise simulation as well as flutter making for a more realistic experience both gameplay and audio wise.
 
The secret sauce behind it

DS: Another great feature on the Need for Speed series is the implementation. How is that process done? And what about the audio engine used on the game?

CD: On NFS we use the majority of EA’s tool-set; AEMS, Pathfinder, Renderware Audio etc. Beyond that the NFS team wrote their own mixing tool starting in 2004 which also has been used in numerous other titles (Skate, FiIFA, NBA street etc.). This tool can modulate over 2000+ parameters each frame without eating up much CPU and acts as an interactive dub-stage equivalent. It was developed by John Twigg and Cliff Kondratiuk. It is still in use today, but in the next few months will be replaced with a more refined version. Splicer is another tool written by the NFS crew. It basically can read Protools sessions and allows very quick multilayer sounds to be created.
Splicer

Aems is our MAX/MSP equivalent runtime. Originally developed during the SSX years by EA Tech. NFS audio programmers recently wrote a new interactive music tool, taking full advantage of beat-matching, overlays, multi-tracks, animation-matching etc. This will be used in a future NFS. We also have a proprietary way of playing back car audio. This was developed by Patrick Ratto. It’s been our work-horse since 2004 when the first review on NFS Underground 2 (official U.S playstation magazine) read “The sound effects are without any shadow of a doubt the most amazing I’ve ever heard in a videogame. The rasp of the engine is so unbelievably believable that you can actually feel the reverberation of the exhaust as the ower surges through it. It will change your expectations of what a game can sound like”. Now this engine is in it’s 3rd version, and we’re doing a re-invention on it. Most games are still working on refining car audio sound reproduction, we’ve moved into making the cars’ audio come alive, be real, be as expected. We’re using a whole new tool-set for this, and some magnificent audio artists.

They're playing chess while everyone else is still on checkers :oops:
 
Necro-thread, but it wasn't worth creating a new one to discuss next-gen audio (everything wanted in this thread may come to pass for XB1X/PS5!).
Just seen this:
Directly access and manipulate audio data in Party voice chat
3/3/2020

The Windows and Xbox (XDK) versions of the Party SDK now support real-time audio manipulation, allowing you to intercept and modify voice chat audio. This enables you to implement custom audio effects, such as spatializing players or applying voice filters.
https://docs.microsoft.com/en-us/ga...tworking/concepts-realtime-audio-manipulation

With 3D spatial audio, we'll be able to place player's voices within the 3D game space. That's pretty awesome. "Over here!" would provide an audio indicator of where someone was. Well, if close. Over distance, they ought to be using virtual radio to talk. That should definitely add an element of immersion, with the need to talk quietly to not be heard by people's being snuck upon.

Edit: How would this work? Would processing have to be done on the server when mixing multiple voices? Or can we realistically steam a dozen independent voice channels? I guess the quality is very low. Quick Googling, XB1X uses Skype's chat tech which is 24-128 kbps. At 50 kbps per voice, a dozen voices might be doable without impacting the other game data?
 
Last edited:
Back
Top