I look at this in terms of dynamic range. What DaveWS did is make some louder then others, which is basically a good thing, but you can approach this in different ways. Making the heavy engines sound louder in comparison to the weaker engines (clipping occurs easily), or making the weaker engines sound lighter in comparison to heavy engines (less clipping). The best thing to do is to start from a mid-type engine, and level it at a realistic presure level when using a decent audioset or monitors at full level. Then match the rest of the engines around this sound. The most loudest engine may NOT clip at ANY time but certainly will make you deaf over a long period of time. This is basically the best way if you have serious sound equipment and don't want to hear any distortion when mixing 20+ of these engines together. The disadvantage of this is that every sound of the game needs tweaking to get everything right and it seems like everything sounds very weak when using 'normal' pc audio set's. When using outputlevels that max out at 100dB it will sound very good and wont destroy your ears. The result is perfectly balanced, only way less powerfull to real life measured over the whole range. When upping the output to 130db+ it all becomes very real. Don't forget you won't get blasted away by 130db's of soundpressure the whole time, this will not happen. You basically don't get past 90db's at all, but when things get really serious it is ABLE to output such force.
The same thing is done with well mixed audio in Movies, voice levels should just sound like normal voices, while an explosion should output the same pressure as an explosion would do in real life, which basically is groundshaking. Disadvantage is you need high output levels to hear all the low presure sounds. This is why they came up with dynamic range COMPRESSION, so at low amplification levels you are still able to hear voices while explosions wont wake your neighbours.
LFS as it is now, is none of these at all, but is mostly like the compressed version to suit common soundsets and is a logical choice to go by. But, the main thing that desturbs me the most, is that most people who edit the sounds are making them louder and heavier compared to what they started with, while if you want to make it louder, just turn up the volume of your soundset and lower the volume of every other sound, and if you want to give it more low frequency pressure, just lower the mid/higher frequency sounds. Since almost every engine sound is already clipping on it's own, imagine what it sounds like with 20 of them on the startgrid. This can't be right, and THAT's why a BF1 sounds like it's destroying your ears, while it actually destroys your audio equipment. In this case, when sound clips, it can't make a nice sine-curve anymore, instead it reaches the output barrier of the integrated synthisizer and becomes flat at the top, often refered to as a square wave. This distorts the originally ment sound. To some this might sound like a heavy roaring sound which might even be noticed as sounding very 'cool'. Well, it isn't and you really want to avoid this by any means possible.
But anyway, i'm currently working an a full range soundset for LFS + CSR combined. This will take loads of time and since i'm mainly a FOX driver, that one will be finished first
while the rest will be leveled and tuned from there. The pack won't be very suitable for everyone and even people with high powered soundsets could experience something totally different and wrong. This is often the case with audio, it's all made at one place with it's own characteristics and accoustics. Which concludes it will sound best where it's been made. A shame really, but the pack will be placed online eventually so people can judge themselves.
The biggest problem i encounter though, is that it seems the audio levels of LFS or CSR have just 8 bit's of precision, meaning just 256 levels. Extending audio levels to high range makes this noticable and you start to hear rough changes in amplitude while it should be a smooth curve. 16bits of precision will suffice though. It could be a CSR only related issue, which will force me to drop CSR, and that's a shame though bacause although it's digitally recorded sound, it allows for very cool layers of sound that could expand the synthisized sound of LFS in a way it totally blends together. This also keeps the most important dynamic feel of the synthisized sound which makes LFS more real then it's counterparts. Combined sound is the way to go from my standpoint, still hope an application like CSR will be integrated in the future, it also get's rid of the soundlag that way. ...Such a shame i can't program, allways need to depend on others to fulfill my dreams. Guess the only solution is to be able to do EVERYTHING myself. Maybe i will at one day
Enough talk now, back to business.
