Wwise/sound design problems, part 2: the scope of the issue widens

[gerunoa 18]

TL;DR: the "Wwise sound set with close to 300 spatially placed sounds" advertised on the product page is incompletely implemented. In addition to the previously discovered issues, per my own testing the vast majority of sounds are apparently not "spatially placed" at all. 

 

First of all, I just want to say that overall I've been enjoying this plane so far, and even if there are some specific areas that aren't quite up to par yet, I think it's got great potential. It's encouraging that you responded so quickly to the criticism about the engine sample transitions and will look into it. But as you do, I'm afraid you'll need to widen the scope of your diagnosis regarding the Wwise implementation, because in just my first 20 minutes playing around with the aircraft I've already discovered another significant sound problem... 

 

To experience the issue for yourself: 

 

1. Load up a plane cold and dark - just for the sake of making my larger point I would suggest loading either the stock msfs 172, or the Twotter's direct competitor at the moment, the Kodiak...

2. Flick some switches in the cockpit, move some levers. Notice how the sound of each control has a distinct spatial position both relative to the cockpit AND in relation to the camera as you move around, e.g. when a switch is on the left side of the screen you will hear it on the left, and then when you move the camera so the switch is on the right side of the screen, you will hear it coming from the right (or if you have a good spatial sound implementation, you can hear a switch in front of you vs. behind you, and even kind of above/below!) 

3. Now load up the Twotter, again cold and dark. 

4. Flick some switches and move some levers. Notice how they ALL seem to come from exactly the same spot, which is inside your head, and they don't move with the camera, or at all for that matter. Note that it doesn't just break immersion (particularly in VR), but that the samples blend in with the other ambient sounds and are not as distinct as when they are properly placed spatially. 

5. Just for completeness' sake, start the Twotter's engines, and notice that the engine and propeller sounds ARE in fact properly spatially placed. Just not anything else.

6. Brace yourself, load up the Aerosoft CRJ and realize that in all likelihood every Aerosoft plane currently under development has the same problem. 

 

So, just a thought on this, please don't shoot the messenger, but I'm just gonna say it. There's a commonality with this spatial issue and the crossfading issue. First of all, we don't need to be real life Twin Otter pilots to easily detect the problems and be turned off by them within a few moments of loading up the plane for the first time. They stick out like a sore thumb simply because we... play games and... enjoy sound. I think you'd agree not just that ideally someone on your end really should've found and fixed these problems before consumers ever had the chance to, but also that it's genuinely odd that no one did, given that presumably the Twotter team has tested the plane in msfs, has multiple sets of ears, and cares about the product they're making; and that there should be some reflection internally about how these issues got through. My point is that this is about more than inevitably having to polish a few rough edges after any release- for whatever reason, you have some sort of problem with the Wwise process that's been keeping these beautiful samples you've recorded from being experienced in the sim as they were meant to be. Bluntly, someone either doesn't understand how to employ Wwise competently or is unable to do it for other reasons. If you were some giant company like Microsoft pushing a minimum viable product out the door, this would all make perfect sense, but it seems like there's actually a genuine misunderstanding somewhere in the sound production pipeline in this case and I really hope you're able to resolve it, for the sake of making the best planes possible. 

 

One final thing, regarding what I said about using real pilots for feedback. Don't get me wrong, It's great that you do that, and you should keep doing it. It's important to remember, though, that real pilots have been checking out flight sims and giving them the "Oh WOW! It's SO realistic!" endorsement for decades now, and it was never exactly untrue per se, even when they were talking about 8-bit sound and vector drawing. One day we'll be having the same arguments about super hi-fi VR and spatial audio implementations that make what we have now look just as primitive. In the present, we hear the problems easily because as simmers we have more nuanced expectations for a sim sound engine than a pilot does (unless that pilot is also a sim/audio nerd). It's just two distinct things and you need both. They hear other problems better, and give relevant feedback, but it's all for naught if you don't have someone effectively articulating these problems related to the game engine.

 

[ReproCase 17]

Ah, this makes sense. I was struggling to pinpoint what was different about the spatial sound, and was thrown off by the fact that the engines were working correctly.

Does this issue persist if you move the volume sliders to resolve some of the compression, as I describe here?

 

[gerunoa 18]

There is simply no spatial data for the samples in the first place, so no unfortunately that doesn't affect anything.

 

As I mentioned in my post this is just one of multiple sound related problems that will require attention; I hope this will be acknowledged by the appropriate people... 

[fahdriyami 16]

I cannot emphasize how important proper spatial sound implementation is. Especially for people like me who like to fly in VR.

 

I will hold off on purchasing the Otter until the sound issues have been fixed. Good plane otherwise, but this is too important an area to not get right.