- 8,715
There are copyright issues, in that the samples produced belong to the studio that made them. Supposedly, in the main, car manufacturers don't actually care what their cars sound like in games, in the sense that it's the legal and marketing teams responsible for that aspect. They must not be enthusiasts... GT never seems to have fallen foul for having inaccurate sounds, or else they would have been remedied pronto.
The GT86 comparison is a clear indication of the kinds of problems that crop up when making games with so much content. You make small changes to the sound engine, and all your presets / patches / tweaks etc. go out the window. In that case it's a simple matter of the exhaust tip noise overpowering the recorded sample. It's one number that needs correcting; annoying that they missed it, but the sound components are essentially the same between the games, just in a different balance.
AES doesn't create samples, and if you're complaining about individual cars' sounds, you're unlikely to be complaining about the ones that use AES. Statistically speaking.
FMOD does not like fast audio ticks, it's fully event based and the streaming provision is not especially flexible. PD stream the physics into the audio engine for crisp response, like iRacing (who tried FMOD and abandoned it because it needed too much reworking to give them the features they needed; e.g. faithful following of the transmission flex simulation).
FMOD is "just" a plug in environment for handling sound sources, it does not exactly have a sound of its own; plenty of games use FMOD and sound awful. The assets matter slightly more in this case.
GT's sound engine is fantastic, one of the best around. They just persist in using the wrong samples, and too few of them. That latter point is subtle: GT ships with more samples than are used in-game. It seems to skip some as you rev through, so it sounds thinner. Almost like it's still using the PS1 era configuration. The samples sound OK in other games, when all of them are used.
DiRT Rally uses a granular synthesis system which is different from normal sampling.
It plays little snippets from a "ramp" recording (engine revs constantly rising / accelerating or falling / decelerating) for excellent fidelity at all engine speeds with no sample crossover or blending issues. But it suffers at constant rpm, sounding a bit like a stuck audio CD (remember those?), unless you use little tricks. Those tricks effectively involve decoupling the physics from the playback, which is not ideal for player feedback.
The GT86 comparison is a clear indication of the kinds of problems that crop up when making games with so much content. You make small changes to the sound engine, and all your presets / patches / tweaks etc. go out the window. In that case it's a simple matter of the exhaust tip noise overpowering the recorded sample. It's one number that needs correcting; annoying that they missed it, but the sound components are essentially the same between the games, just in a different balance.
AES doesn't create samples, and if you're complaining about individual cars' sounds, you're unlikely to be complaining about the ones that use AES. Statistically speaking.
FMOD does not like fast audio ticks, it's fully event based and the streaming provision is not especially flexible. PD stream the physics into the audio engine for crisp response, like iRacing (who tried FMOD and abandoned it because it needed too much reworking to give them the features they needed; e.g. faithful following of the transmission flex simulation).
FMOD is "just" a plug in environment for handling sound sources, it does not exactly have a sound of its own; plenty of games use FMOD and sound awful. The assets matter slightly more in this case.
GT's sound engine is fantastic, one of the best around. They just persist in using the wrong samples, and too few of them. That latter point is subtle: GT ships with more samples than are used in-game. It seems to skip some as you rev through, so it sounds thinner. Almost like it's still using the PS1 era configuration. The samples sound OK in other games, when all of them are used.
DiRT Rally uses a granular synthesis system which is different from normal sampling.
It plays little snippets from a "ramp" recording (engine revs constantly rising / accelerating or falling / decelerating) for excellent fidelity at all engine speeds with no sample crossover or blending issues. But it suffers at constant rpm, sounding a bit like a stuck audio CD (remember those?), unless you use little tricks. Those tricks effectively involve decoupling the physics from the playback, which is not ideal for player feedback.
Well it sounds easy anyway
