Hello,
I have a mp3 files that have noise/crackling within the voice and I would like to remove it so the voice sounds clear.
Which tool could I use for this? the Noise Removal tool only works if I can isolate some noise and make a noise profile of it, but since the noise is only within the voice and not the mute parts it’s hard to make a noise profile.
I don’t think that Noise Removal will help. The sample rate is 22050 and what I’m hearing sounds like aliasing on the sibilants or some other artefact of the MP3 conversion. There’s very little hiss in the recording at all. Low-pass filtering doesn’t help either.
I think I know what the problem is even though I didn’t get to hear your sample … I don’t know why my media player closed on me and didn’t play anything.
I spent a few years using simple recording apps to record my work because I’m always on the go and I learn best that way. I had lots of background noise in my recordings and nothing helped. I searched for an application that would give me quality sound recordings because listening to annoying background noise repeatedly becomes torchorous.
My problem was solved by Audacity’s help material. There’s alot I haven’t read yet. I clicked on the Help tab on Audacity 1.3.10 … then clicked Quick Help … then clicked Record which is the second bullet point …then clicked Your First Recording … I scrolled down to Identifying Various Cables and Plugs … it shows plugs of analog microphones … I didn’t know much about microphones but I decided to try a USB headset with sound cancellation and and a unidirectional mic … MY PROBLEM is now solved … I get beautifully clear recordings with no interference which was from the pc itself … the plugs record analog sound and pick up on internal noises of the pc …
I don’t know which of the help areas I read from but there was advice about noise cancellation microphones. It said that if you can stop the menacing noise from being recorded in the first place, then there will be no problem to find solution for.
I bought my headset from Argos … its product no.675/3650 in their catalogue … it’s comfortable as well
The woman’s voice was recorded with a certain hiss level in the capture channel and the man’s voice was recorded with the microphone too close to his lips. If you listen carefully, you can hear his breath briefly after some of the responses. On some of his words, there is significant though brief overload distortion.
The show was processed extensively to get from the raw capture performances to the show that you have. The time between the dialog segments is suspiciously free of any noise at all, meaning it was artificially produced – it’s not at all natural.
If you don’t have access to the original voice material, I think that’s as good as the show is going to get. I believe it’s possible to clean up the woman’s voice hiss with the noise reduction tools in Audacity 1.3. You might be able to help the man’s overload with “Clip Fix” as the distortion is not too severe and there is very little damage.
You could do all that to the original, perfect quality, high sample rate capture clips and WAV files.
Once the performance was cut, produced, restricted in sample rate and then burned into MP3, that’s the end. The errors are burned in and permanent.
The time between the dialog segments is suspiciously free of any noise at all, meaning it was artificially produced – it’s not at all natural.
This makes sense!
I know a guy who was able to fix such a mp3 file and the result was really good, but the sad thing is, I don’t have his email anymore - so I would like to try it myself. (I think he was either using RX iZotope or Adobe Audition)
How can I remove this hissing with the noise reduction tool? Using this tool on silent parts is easy, but how can I take a noise profile if I can’t isolate the noise. (The noise is always together with the voice)
The hissing is compression artefacts, note the hiss precedes the woman’s voice, (this pre-echo is a compression artefact).
I’ve had a go at removing the hiss, (not using Audacity), but really I’ve just substituted one type of distortion for another… ‘A Paris’ (before-after cleaning) wav in zip.zip (512 KB)
You’re better off with the original hissy version, otherwise you could end up speaking French like Stephen Hawking.
One more vote for “put up with what you’ve got”.
You can only go so far with noise reduction before the cure is worse than the illness. Someone has already taken the noise reduction as far at it can go without the damage being worse than the hiss. On top of that, 22kHz sample rate and MP3 compression drop the sound quality too low for making any useful improvements.
<<<The hissing is compression artefacts, note the hiss precedes the woman’s voice, (this pre-echo is a compression artefact).>>>
I don’t know that I would agree. We know that the piece was highly produced and edited and I noticed the hiss during the woman’s voice. If the woman was as off-mic as the man was too close, then the channel gain would have been increased giving a hissy performance during her entire studio duration. The hiss is there before, during and after and selectively cut to get the unnatural silences between the voices.
This is exactly the magic place where Hollywood gives up and hires two people to reshoot it. We have how many people across the earth working on this now?
thank you guys for trying it out! I think I have no other choice than sticking with the distorted version if even the pros here can’t do it.
Thanks again!