Mixing voice with background static

E

Elyst

New member
Greetings,

Let me get straight to the point, I'm trying to edit some voice recordings I did, but has usual there's always some nasty background static that keeps getting in the way.

I'm using Reaper and I just tested it's noise reduction FX, you "teach" him what noise is and he removed it. However a part of the wave is also removed from the voice of the person speaking, which you would expect to be good meaning that the noise is no longer there but ... the voice now looks strange... slightly distorted perhaps...

My guess is that by removing the noise from the voiced parts I also removed a bit of the acoustics it did and that's why it now sounds different and "strange".

Is there a way to fix this or do I have to "bite the bullet" and decided wether I prefer a shifty voice or some background static?

PS: before this I was using a noise reduction that set a threshold and muted any sound lower than that... it was ok since you didn't hear anything in the pauses... but the static was there when the person as speaking as well as some unwanted sounds like swallowing or other similar stuff the mic caught.

PPS: The purpose for this voice is to accompany a slide presentation reading it. I believe usually in these things the voices are carefully treated so nothing but the voice is heard. I might be being a little perfectionist by trying to remove all the noise... but I wanted to do this right :) :listeningmusic::thumbs up:
 
Fix the problem.

Find out what is causing the 'static' or background noise and stop it BEFORE you hit record.
 
mjbphotos said:
Re-record the vocal track!
Click to expand...
I don't think I can do this... I probably won't be with the person who did this again... only in case of emergency... so I'm using that as a last minute resort.

anoopbal said:
I think if you get the Izotope Audio repair or something smilar you can easily do it. iZotope RX 3 | OVERVIEW
Click to expand...

Thank you, I'll look this up. So I'm fixing the audio after I "screw" it by removing the static?

Seafroggys said:
can you upload it so we can hear how it sounds? It may not be too bad.
Click to expand...

Perhaps in a later post next sunday/monday, I don't have much time now.

TimOBrien said:
Fix the problem.

Find out what is causing the 'static' or background noise and stop it BEFORE you hit record.
Click to expand...

I believe the source may be the fan of the laptop close to the person who was being recorded, which could not be avoided since she didn't have anywhere else to read the text from. I know it was rather "unprofessional" (even as an amateur I have to admit it) but I wasn't sure how to solve this issue in the heat of the moment I just assumed it could be taken care in post. (I know I know, my bad).

Either way, and this being a omnidirectional mic (also a very bad idea even for a semi treated environment) I think there's always some static noise getting caught.


Thank you overall for the replies. I know some mistakes were made overall in this recording... but i needed to work around some of these issues.
 
As someone else stated try izotrope RX3. It is pretty amazing and you can download a fully functioning demo.
Brad
 
As a follow up question. How do you guys usually treat your audio?

I ask this because I was using some senheiser headphones plugged to an alesis io2 express interface.
But yesterday I wanted to show some of those audio files and I used the sound card within my mac and it sounded completely different, I could pretty much hear the person breathing (which is something I don't want), also the noise removal seemed to damage the voice itself a lot less.

I hunch is that when i was using the interface I had a low to medium volume, and in spite of that I could hear just fine the voice and even some "clapping" lips has the person started talking.
When I used the mac's sound board I had the volume almost up to max... and the sound was coming out of the speakers instead of headphones.

So, should I take a wild guess and assume how people will listen to that audio and test it in approximate conditions?


PS: Not sure this is relevant or perfectly normal. The sound initially recorded was low (I couldn't see the wave in Reaper even if I zoomed in), so I normalized the tracks which rose the volume for the whole track I think... at least the wave is now easy to see and I can properly edit it.

Thank you for all the help so far :)
 
An idea: you might try to use a 'deessing' Approach. Find a narrow EQ band that holds the noise (or some if you have a strange fan). Drop These bands completely. Subtract the result from the original voice (by doing a Phase Inversion). You now should have two tracks: one containing only the noise and one with the (strange) voice. Now apply a gate/expander to the 'noise only track'. Set the threshold in a way that normally the noise is suppressed well enough. Mix the two tracks.

You should now get sth that suppresses the noise quite OK, but only as Long as the Speaker does not speak in These freqs. Might work...
 
Do this: start recording with nobody making a noise. Capture the noise in the first 10 seconds of the recording. After 10 seconds, start talking, playing whatever you want to record. When processing the recording, use the first 10 seconds to set your baseline noise profile for the software to filter out. It works for me all the time!
 
You must log in or register to reply here.
Back
Top