Small introSmoothing is just "moving window" average. Based on time inverval (aka convolution).To be a little "scientific" samples are in similiar lenghts. Time can vary +- 1 or 2 s, but most sounds have around 10s. I also tried record same patters, which is terrible thing to do on labs, becoues sometimes there are loud noises somebody hiting some metal or alarm. (Yes I did offline, its part of ambient so I had to record it again)
Some results:
Some first conclusions are that: CT4 are best, based on average results in noise suppresion.But I should probably look up the peak results as well, probably best choice will be 1 or 5 ms smoothing to reject noise.Tips for analysis
You probably want headsets that have best noise suppresion like ambient or rain including gunshots, but headset should boost walking, bush noises, or other pmc related noises. In my opinion better headsets suppress gunshots more, but thats my taste.
Warning about data
Please keep in mind, that tree noises are "rng" based. Walking same path for each sample could give different results. I also lost ComeTac2 audio file. sad
Pics
2 pics for you. Same results but different scale. One in detail for each plot and second with same scale on every plot.
Any debate or questions, just post them and I will answer them. I may add some sounds later, but who knows reall. Im open for suggestions
Lol i should play tarkov instead of wasting time for graphs. I sometimes stream and get angry using M32, that some shity headphones, they barely supress
Source: https://www.reddit.com/r/EscapefromTarkov/comments/xxnm77/i_did_research_on_headsets_in_tarkov/
