Community of Sexsoundlovers

What do people use for reducing noise in long-distance recordings? I use Audacity, but as much as I play around with the options, I simply don't like the results. Maybe it's impossible to get what I want, but I thought I would see what others are doing...

izotope rx

helps you to "see" the sound and has some good tools. a bit expensive however but the de-noising are great.

For example here is a screen shot with one of your recent files.

I use Adobe's DeNoise. I did a quick review of it after Adobe introduced it in 2018. They purchased a very expensive algorithm for it.

viewtopic.php?t=7856

It works really well, is so easy and quick to use with no artefacts, unless you boost the percentage reduction to silly levels.

Thanks I will check that out

I am looking for audition to play with denoise. The demo I saw worries me a bit as it may thing some of the stuff we want to keep is noise. It seems to doa great job with the defaults in taking noise out of normal speech but the problems with what we are doing is we are looking for that one needle in the haystack, and it may well think what we are after is noise, say in the case of going at it with the TV on and having speech as what it wants to keep, or in the case of just picking the loudest noise as what it figures you want. It seems to have frequency selective control which may be helpful. If I find a demo copy to play with I will do just that. It would be cool if they are using something akin to the Nvidia de noise stuff, to have a way for you to submit your own stuff for it to learn what is good.

reggind wrote: ↑31 Oct 2021 03:03 I am looking for audition to play with denoise. The demo I saw worries me a bit as it may thing some of the stuff we want to keep is noise. It seems to doa great job with the defaults in taking noise out of normal speech but the problems with what we are doing is we are looking for that one needle in the haystack, and it may well think what we are after is noise, say in the case of going at it with the TV on and having speech as what it wants to keep, or in the case of just picking the loudest noise as what it figures you want. It seems to have frequency selective control which may be helpful. If I find a demo copy to play with I will do just that. It would be cool if they are using something akin to the Nvidia de noise stuff, to have a way for you to submit your own stuff for it to learn what is good.

Appreciate what you are saying about ambient sound and know some members hate any tampering at all. For me, if it improves the listening experience of the sex, I'm happy with it in moderation. You may have already seen it, but if not, here's a demo I did after being asked how I do noise reduction by a couple of members.

viewtopic.php?t=10274

It was the simple default way, but a lot of sophisticated tweaks are available in DeNoise. With Premier Pro I even use it to enhance the sound of some video clips. Such as certain UKnightman ones with wind noise on the mic and similar others.

If you want to play with the latest version of any Adobe apps free, you can sign up with Adobe for a month's free trial.

I looked at the trial and I think the trial is down to 7 days and it is waaay to expensive for me to rent

I just found the state of the art in AI based denoising. Sadly it is limited to 200 minutes for free, and you have to upload the file and download the result. But the results are amazing. They have some demos on the home page: https://audo.ai/ if you are at all interested in de noising scroll down on the home page and check out the demos. They also have a playground where it will de noise the first 5 minutes of a file without any type of account.

On the minus side, this is not at all tunable yet. I tried it with a couple of things and it totally nixed quiet things but it also did well in other places. I tried loading both the de noised and the original in audacity and flipped the phase of one of them and had some very interesting results being able to hear things that were under noise that I had not heard before. It is worth checking out and if you have a loud file that is corrupted by music this may be a solution. More playing is requied.

If you have a few check it out.

reggind wrote: ↑29 Jan 2022 07:10 I just found the state of the art in AI based denoising. Sadly it is limited to 200 minutes for free, and you have to upload the file and download the result. But the results are amazing. They have some demos on the home page: https://audo.ai/ if you are at all interested in de noising scroll down on the home page and check out the demos. They also have a playground where it will de noise the first 5 minutes of a file without any type of account.

Thanks a lot for the information. Sounds a good option and I will give it a go in the next few days. 200 minutes for free is pretty good. By the look of it, with the playground option, if you used up the 200 mins and didn't want to pay, You could just cut a file into 5 minute chunks and then join it back up.

Just had a good look and reggind is right, it is amazing.

There may have been some changes to the site, since reggind first looked, or maybe it was me being thick. For a start I couldn't find the area reggind described as a playground, unless that's the demo area, but that wouldn't work for me, so I created an account. Using that I was given 20 minutes a month free, although for $12 a month you get 10 hours a month, so I'm thinking about it.

Using the denoising app I found the results to be great and it was quick too, although I found that using it at 100% was a bit too much. By the time I had played around a with a video clip I had used up my 20 minutes, so don't think I had got the best results that may be possible. I intend to create another account and try with a file that's around 30 seconds long. That way I hope to find the best settings that work for me.

Created a new account in Audo ai and took a 30 second very noisy bit of audio from an outdoor video. It's a couple having sex in a children's playground climbing apparatus, with only limited sex sounds. Here is the original untouched sound, Adobe Audition's Denoise at 50% and Audo ai's at 50%. Have a listen and see what you think.

The Audio Ai version seems to be in mono. I can't say at the moment if that helps or hurts but that jumped out at me right away.

tweaktweak wrote: ↑08 Feb 2022 00:07 The Audio Ai version seems to be in mono. I can't say at the moment if that helps or hurts but that jumped out at me right away.

Interesting. I didn't notice that, but you are right. Audo ai has converted it to a mono track. Just to confirm, the original is stereo and I didn't change anything.

Looking at the original in Audition, you can see that although the left and right channels are similar, there are also differences. In essence, that does mean that Audition gives the more authentic 'live' sound, but as I said, I didn't notice.

Interesting comparison, and the two hours thing is if you sign up for the API, don't worry you don't have to do any coding. In the API window they have an upload file button, and it is pretty much the exact same interface as the time limited demo on the home page. No adjustments just it does it's thing and you download the results. For stereo break it apart and run it through twice if it has a nice stereo effect you want to preserve.

reggind wrote: ↑12 Feb 2022 06:32 For stereo break it apart and run it through twice if it has a nice stereo effect you want to preserve.

Nice idea for retaining stereo! Thanks for explaining the API two hour thing too.

Just after I've applied a process to something I find it difficult to be objective when listening to the results, as I wonder if it's my imagination kicking in. Having listened back now though, I think Audo ai and Audition produce similar results, but would be interested in the opinion of others.