View Full Version : Disguising poorly recorded vocal tracks
andyjayne
03-20-2006, 08:45 PM
I've recorded vocals using a poor microphone for one of my electronic pieces but I wish to use them without the quality standing out so much. It's not a case of hiss reduction, the vocal track hasn't got much hiss at all. More so that it sounds as though it has been sampled at 8~11KHz. Regardless of how well I get the vocal track to blend in, the only way I can make the quality of the samples not stand out is by reducing the volume and making the tracks muddier.
Asides from getting a better microphone, does anybody have any ideas how I could go about doing this? Perhaps there is a method to make samples clearer or sharper, or an alternative approach?
Cheers.
zircon
03-20-2006, 10:46 PM
Can you post some sound examples of what you have to work with?
suzumebachi
03-21-2006, 05:21 PM
you could purposely go for a kind of lo-fi sound (a la blur/gorillaz). it's easy enough. just EQ out the lows and some of the mids (or in this case, sometimes it's better to raise the highs and "heat up" the signal) then slap on some analog saturation or a tube distortion sim with very slight gain.
this probably wouldn't work very well for some styles though.
Legion303
03-22-2006, 11:33 AM
I'm surprised no one's mentioned it, but using what you have should be your last option if you have a bad recording of anything. Try your best to re-record it (even on existing equipment) first if possible. Other than that, Zirc or Compy probably know a few studio tricks to help the sound a little, but that's almost always going to be inferior to a better original recording.
-steve
Reasoner
03-22-2006, 04:04 PM
yep, you can't fix a recording from a bad mic. There isn't a magical plugin that's going to make a sound coming from a Shure SM57 sound like it's shooting out of a Neumann M147 with a Millenia Preamp through a Mytek A/D Converter. Of course unless you've got a good 6 or 7 grand to drop on all that, better to just work with what you've got.
zircon
03-22-2006, 07:56 PM
To be fair, however, it's very much possible to make recordings from WORSE mics than 57s (which, along with 58s, are still probably the most commonly used mic in the world) sound at least halfway decent. But in order to recommend a methodology for approaching this, it's important to hear what we're working with.
PriZm
03-22-2006, 08:35 PM
well an idea I'm having if they're sounding lo-fi is to drown them in reverb.
Otherwise it's probably a hardware problem. If you're like me and have a shitty laptop soundcard (I have sigmatel C#), it might be a good idea to switch to desktop or/and change soundcards.
analoq
03-22-2006, 08:51 PM
There isn't a magical plugin that's going to make a sound coming from a Shure SM57 sound like it's shooting out of a Neumann M147 with a Millenia Preamp through a Mytek A/D Converter.
there's several mic modeling plugins and hardware units that aim to do just that. good as the real thing? probably not. but it is a viable avenue.
and like zircon is saying, there's an SM57 in every recording studio in the world. usually they're for mic'ing guitar amps, and sure they have better pre-amps at their disposal than the eurorack crap you'd find in home studios... but still. don't be dissin the 57.
Perhaps there is a method to make samples clearer or sharper, or an alternative approach?
ok, you're saying it "sounds" like it's been sampled at 8-11khz. that basically equates to having no frequencies in the higher range, or at least no clarity in the overtones/fricatives/sibilence.
take your sample, pitch it up an octave (while maintaining the same length) and mix it back in with the original. you'll be mixing it in very slightly or perhaps bandpassing whatever frequencies provide the clarity. this is how most exciters work, others work by adding a short delay to lower frequencies, which you can try as well if you have the tools for it.
from there it's all bread and butter: eq/compress/verb the track the best you can.
and invest in a decent mic, preamp and audio interface in future.
cheers.
Reasoner
03-22-2006, 09:19 PM
I'm not trying to say the SM57 is a bad mic. I used to have 2 of them. I was just using it as an example. A $2500 M147 vs. a $90 SM57 is obviously going to have a few advantages to it (for vocals at least,,,,snare drums are another story). It's important to note that no amount of tweaking or post-processing is going to be able to completely improve and upgrade the sound of your source signal. But I agree that it is better to work with what you have, and if a $200 or so plugin (or free) is going to give you that extra boost in quality, go for it.
andyjayne
03-22-2006, 09:43 PM
ok, you're saying it "sounds" like it's been sampled at 8-11khz. that basically equates to having no frequencies in the higher range, or at least no clarity in the overtones/fricatives/sibilence.
take your sample, pitch it up an octave (while maintaining the same length) and mix it back in with the original. you'll be mixing it in very slightly or perhaps bandpassing whatever frequencies provide the clarity. this is how most exciters work, others work by adding a short delay to lower frequencies, which you can try as well if you have the tools for it.
That worked pretty damn well with what I have. I applied some filtering after that technique and got the sound I was after but I still couldn't get the clarity I desired. I think the issue is with both my mic and soundcard, although primarily the former. I'm going take the main advice to record it again but using a better mic (someone elses) and this time I'll be recording into a friend's hardware sampler instead.
If I get the same thing I'll post the samples here, but other than re-recording it analoq's suggestion worked quite well. Listening to it again it does sound as though there is a shelf on the higher frequencies, this is most likely due to the quality of my mic.
Edit to actually clarify what worked well:
Pitching the sample up an octave and mixing it in at a lower volume and then duplicating each, adding a slight dampening filter to one set of 2 and using vocal definition compression techniques on the other. Shaping the vocals using a "male pop vocals" EQ preset then levelling out the overall sound. Finally duplicating each of these and panning either very slightly out from the center with one of the dampened ones in the center.
vBulletin® v3.8.1, Copyright ©2000-2013, Jelsoft Enterprises Ltd.