So what are the significant sound improvements that XB1's SHAPE can offer compared to previous gen?
I cant think of anything that can be a game changer other than something that mimics binaural
Dynamically adjusted sound.
The difference between pre-recorded (pre-baked) and realtime modified (realtime).
For example, lets take something as simple as footsteps.
1. They can be modified by the materials (not just wood, but pine or mahogany or cedar or balsa wood [
] and not just stone, but granite or cement or obsidian or glass or ceramic or vinyl or whatever) being traveled over. Thin material layer or thick material layer. Solid or has holes in it? Bumpy or smooth? Made of just one material or multiple? Rather than being pre-recorded you can modify them in realtime.
2. The environment. Are you indoors? Outdoors? Are there walls? One 1 side? 2 sides? 3 sides? 4 sides? A roof? If indoors how large is the volume of the room?
3. If in a room, what are the material characteristics of the walls? Wooden walls reflect sound differently than Metal walls. Perpendicular walls reflect sounds to the user differently than angled walls. Is the wall a smooth surface? If not, how is the wall shaped? Does the walls have holes?
4. Are there objects in the scene? What are the material characteristics of the objects? Vinyl reflects differently from cloth which reflects differently from leather which reflects differently from wood which...etc... What are the shapes of the objects?
5. How do different frequencies react to all of the above.
6. Is distance going to finally be modeled correctly? Low frequencies travel farther than high frequencies. Add in occlusion. Low frequencies can penetrate occluding objects much more easily than high frequencies.
7. How does the sound in the environment affect the objects in the environment? Is there resonance in some objects? Will a suitably powerful low frequency sound wave cause objects in the environment to vibrated thus generating more sounds?
8. Something as simple as the Doppler effect on sound. Something so simple, yet so incredibly lacking in games that it bugs me everytime I hear a siren from a police car, ambulance, fire truck, etc.
9. The list goes on and on.
And then add in every single other source of sound in the game (channels). Remember audio processing does NOT equal the number of voices sound hardware can support simultaneously.
It really is similar to whether you prefer pre-baked shadows or want realtime shadows. Do you prefer pre-baked lighting or would you prefer realtime lighting? Prebaked animations or actions modified by physics calculations?
Some of the above is done on a very very simplistic level in some games. And only on a VERY simplistic level, because a lot of that stuff is very computationally intensive.
Even with Shape, it'll only allow a small fraction of what I described above to be implemented. But a little bit is still infinitely better than the relatively flat and crappy sounds we have in games today.
Of course, you could attempt to do it the old fashioned way of recording footsteps on a wood surface, metal surface, stones, etc. But to do anything even remotely reflecting a fraction of what I mentioned above would require terabytes of sound files.
That is why audio processing is important, IMO. To people that didn't get to live through the early days of audio processing on A3D or EAX, the sound in games today probably sounds good. As well A3D and EAX was still very primitive back then but still sounded better than what we have today. To me? All audio in games for the past decade has sounded like crap.
Then again, will any multiplat developers make use of Shape considering the PS4 would likely require significant GPU resources to try to replicate the effects? If not, then it's a bloody shame as audio has been seriously lacking in games.
Then again perhaps the PS4 audio hardware is also powerful, but I'm not holding my breath on that one. Sony would have said something about it if it had been.
Regards,
SB