Why not use .mp3 instead of .wav? I have this question for a long time, since I use the shift + a feature moderately, sometimes the .wav repeats itself so many times that it start not working correctly, like on almost every turbo dump valve sound.
I think it would work better if it was like in car sound remixer, with the audio playing only once, and you could choose to play more than one sound on the turbo valve for example.

https://www.youtube.com/watch?v=hObefKmXBQA like in this video, I used CSR and could make a better sound than using the Shift + A function.

Misleading title and thought process. Propably makeing a video pinpointing the blow-off valve sound looping issue may bring better results than asking for whole sound synthesizer system redesign to reasemble CSR.

For a long time (only until about 5 years ago), MP3 was problematic due to patents around the world.

i might be wrong here, but wouldn't decompressing mp3s use more CPU time?

Quote from johneysvk :

i might be wrong here, but wouldn't decompressing mp3s use more CPU time?

Very marginally, yeah. But you'd probably just decompress them once and store them in RAM. So it'd make car loading slightly (imperceptibly) slower and still store it in RAM.

While adding a bunch of complexity dealing with an mp3 decoder to put sounds back to WAV (essentially) before use.

Quote from johneysvk :

i might be wrong here, but wouldn't decompressing mp3s use more CPU time?

Potentially, but a nonexistent issue on modern computer hardware.

Compressed formats could load even faster because there's less data to be read from disc, and depending on codec used, old hardware wouldn't take a sweat eitherway. But again, the title is misleading, I knewed it will mistakenly turn into codecs debate, meanwhile the post author wants some behaviour that's totally unrelated to format used, read it closely. He propably came to the conclusion that since CSR could load up mp3s, and LFS is sticking to wav than that's the cause for his issue, but it's totally unrevelant.

That's why you first explein your experienced vs. wanted behaviour in plain words, than you try to unravel some implementations details on how to achive your goal, because the second part is often times invalid due to not sufficient knowledge, either by not grasping the matter long enought or just not knowing LFS internals (as no one should know).

But since the debate started, I personally see no reason to FLACise shit. Compressed [lossly] formats wouldn't really work, Opus breaks stereoimage, which is important part in raceing, mp3 has too many flaws to even point out them all here, mainly cuz it's too old, so don't even make me laugh bout bringing this 90's tech here, and aac has some legal shimmers. No other candidate of which I'd've been aware of, but if I'm not aware of it, it means the library support and general toolchain may be wonky, so clearly not the best path either. Meanwhile FLAC would almost provide the same functionality as the wav, in most instances it would be identical thought, while always bringing some deweighting.

FLAC is compressed too, just uses lossless compression.
While the cpu usage to decompress the files might not be huge it is still an unneeded strain on the CPU, for no real benefit.
storage and ram use are not a big issue given the size of the WAV files used.

doesn't seem to me like efficient use of CPU time

the way I wanted shift + A to work was to play the audio file once, because when the audio loops it gets a little strange
maybe the format used really didn't has anything to do with this, I just didn't really know lol 🤪

Quote from johneysvk :

FLAC is compressed too, just uses lossless compression.

Ah, yeah, I meant lossy compression ofc.

The CPU_TIME compare to .wav would be really unoticeble considering you'd only have to do it once as car loads, as in RAM you'd store raw bytes PCM eitherway.