Hi everyone, I appreciate any help you can offer. When you have no clue about audio like myself it’s overwhelming to know where to turn. The first route I took was calling forensic audio places and for $500 they could maybe clean up a file. I called another place and before we spoke price he wanted to hear a sample the answer was he couldn’t clean the audio up anymore than it was and it was clear enough understand what was being said. Except I can’t make out exactly what is said. The audio comes from a camera located in my home where there was potential uninvited people who entered my home while I was out of town. Without knowing what was said I have no idea who it was. Any direction would be so helpful. If anyone would like to give my 12 second audio files total of 4 a try I would be happy to forward to you and if you can help I have venmo and will pay you agreed upon amount.
Just as an idea, try and play back with treble as boosted as much possible
It's probably too late now, but if you'd like to put the audio files somewhere I can pull them down I'd love to give it a shot.
I've been playing with Nvidia Maxine GPU audio processing. It seems to be able to do things not possible with simple frequency filters.
your not too late and your help is so appreciated… I will attached here.
Here is the file… thank you so much again..
https://drive.google.com/file/d/10UYad5qe2j8muinKtmxpdXWh3t-N3dAM/view?usp=share_link
I just saw the files. I'll give it a shot.
I had fun trying everything I could think of but I couldn't get a an improvement.
Sorry, but thanks for the puzzle!
@watcher did you get any improvement? I’ll take anything lol I have more puzzle trivia’s. Do you work for a company or would you be interested in helping me with about 20 Min total approx.?
I have some files that are basically understandable for the most part I looked into the software you are working with and it sounds legit but above my pay grade. I just don’t want to post all my sound files here for public to hear lol ya never know. I really appreciate you trying either way but if you were able to improve the sound at all let me know okay thank you
I had high hopes for the Nvidia "Dereverberation", it worked great on their demo sample but I didn't get any detectable improvement on your audio; I even setup a script to loop the output back for another round of input and ran that a thousand times, sampling every 50 rounds. The lesser sounds attenuated each round until there was no audio at all by 500 rounds. At around 300 rounds the loudest audio was still surviving but it was no more understandable than the original.
I tried various Low Pass, High Pass, and Notch filters. I "Normalized" and "Frequency Shifted", all to no useful result. To my ear the original was still at least as good. Audio was never my specialty so you may still get better results from an audio guy.
In answer to your other question; no I don't work for a company. I've been retired for a number of years and apparently have trouble NOT dabbling.