Another day full of stress.
I have recorded a vocal at home but after recording there appears to be lot of (digital?) noise .
as the singer lives not around the corner there is no chance to do it over.
The vocal at it self is good to so what plugin/program can I use best to remove this noise?
Payware and free, any suggestion will do good as I’m on a time limit.
This is a bit like cutting too much wood and wishing you could magically glue it back on.
The short answer is no, there is not tool that will 100% fix this problem. There are some fancy filters out there, but not do the job without degrading the quality of the non-noise part. Same applies if you’re trying to EQ out the noise from being noticed, you end up killing the tone quality of the vocal.
If the noise is subtle, then you can get away with a gate to kill the noise during the pauses, and perhaps a bit of light LP filtering above 11k. This may turn out ok in a busy mix, not so good in a quiet mix. But the sad thing is, you’re really stuck with that noise, and the only way to get it pure again is to redo the take, with improved engineering.
copy the link to your browser, links from kvr act funny when posted straight into this forum.
http://www.kvraudio.com/get.php?mode=results&st=adv&soft[]=e&type[]=69&f[]=0&f[]=au&f[]=dx&f[]=ladspa&f[]=rtas&f[]=vst&linux=1&osx=1&win=1&free=1&com=1&un=1&sf=0&receptor=&de=0&sort=1&rpp=100&compact=1
and post the recording, maybe someone here can help you.
can you tell where the noise is in the frequency range? more like a broad range or narrow?
personaly i’d stay away from denoiser-plugins. rather use a combination of (narrow-band) notch filters and take out the noise at those frequencies that aren’t masked by the vocals dominant frequencies anyway. add a gate for the pauses between the vocals.
all that mr_mark_dollin says is true though - no matter what you do you’ll always degrade the vocals as well. so maybe think about if you do something at all. if you don’t hear the noise when all the other signals (music…) are playing along with the vocals, just leave it like it is. headphones is the best way to check this.
I remember a recent trick I used for some quiet clean guitar recordings that had a high noise floor. Most of the noise was noticeable in the upper frequencies, as it was white noise from an unbalanced line. So I used the signal follower that reacted off the transients of the guitar sound, and linked it to a LP 24dB filter (which has a open and gentle feel). I’d set it up so the filter would glide down from 20khz to 4.7khz when there was no transients present. This meant that the crisp sounding attacks of the guitar were preserved but the lower decay freqs would be preserved while the noise was ‘masked’. Sounded fine by itself, and when the recording was put into the mix then the noise was almost entirely unnoticeable. Tweak the release of the Signal Follower to give the most musical feel, say around 300ms or longer.
I don’t think this would be as quite as good a method for a noisy vocal recording, but the concept might be generally useful. Good engineering and preparation generate the best results.
I know this problem and mostly use AmadeusPro ( (Soundeditor for MacOS, cheap, shareware, but even demo version has no limits and can save with 30 days trial) for such cases. Adobe Soundbooth or Adobe Audition do have similar features though and they are hybrid pc/mac. Maybe somebody working with one of them can tell it more precisely.
With AmadPro you can very easily and clean get rid of low humming/buzzing noise (50/60Hz) and a it also has a very smooth whitenoise (high freq) remover with realtime preview. In case your noise is in a wider spectrum you can also take a few ms “noise stamp” of the unwanted stuff and subtract it by any degree from the vocals.
As said, it is allways a trade off, but with tweaking the settings accordingly, the results can be ok without destroying the material to much
good luck
Audacity has something called “noise removal”.
It works in two phases -first you are selecting on sound what is noise. Later - you are executing filter on your full sound.
Works pretty nice.
waves x or z-noise also has this learning function which ‘samples’ the noise first, enabling you to subtract this noise-profile from the original material. They use the restoration suite here in school to clean up old tapes.
+1 for this although the results are variable. I’m sure that wavelab (and indeed most audio editors) has a similar feature but Mr Dollin’s bang on the money (again! 2nd post is genius!) = Make sure the source sound is better next time
Good luck!
if its a static digital noise use Izotope RX and see graphic spectrogram and select the noise then delete…
if u can post seom short version of the VO the i can help you about that.
Good equalizer techinc can help this problem too.!
But nothing is 100%, if the vocal fit the music right the noise will be gone as well some background noises
send me and i will chek them peace
Select area of waveform with noise but no vocal, repeat to end of song, invert phase. Classic noise cancellation.
Um… the thing about noise right… is that… it’s like… random innit 
He said it was digital static like buzz, pretty constant really. I used that technique when I recorded a piece for a college project and had an annoying buzz that I couldn’t find in my effects rack. Worked a charm. If it’s bleed from noisey things around then I haven’t got a clue.
No he didn’t, the only description of the noise he’s given is this:
“there appears to be lot of (digital?) noise .”
Digital noise… Well technically quantisation noise is digital noise but obviously can’t be talking about that. Digital (brick wall) clipping can sound a little like noise but obviously you wouldn’t get it in quiet parts. You are making assumptions from things not actually said!
Also your method, even in the very very rare case that it’s not noise but static bleedthrough from something generating a steady tone/repeatative waveform you still have the phase align when you invert and paste mix over the entire waveform otherwise you have equal chance of adding to the existing noise, rather than subtracting from.
Method is generally used for creating acapellas from the original song and it’s instrumental version. Most the time it’s going to do sweet FA when trying to battle anything even vaguely noise like!
i just saw this on news blog, and on picture it looks very good - WAVE ARTS DIALOG