-
Welcome to the General Discussion forum for UAD users! Please note that this forum is user-run, although we're thrilled to have so much contribution from Drew, Will, and other UA folks! Feel free to discuss both UAD and non-UAD related subjects! 1) Please do not post technical issues here. Please use our UAD Support Forums instead. 2) Please do not post complaints here. Use the Unrest Forum instead. They have no place in the the General Discussion forum. Threads posted in the wrong forum will be moved, so if you don't see your thread here anymore, please look in the correct forum. Lastly, please be respectful.
EQing out harshness on vocals?
- Thread starter keano
- Start date
Try using a multiband for this like the Precision Multiband. EQ would be most useful if it's a problem ALL the time. Another thing you could try is make a second channel and copy all the parts that sound abbrasive to that channel and EQ those sections/words differently. Kind of like a using a mulitband compressor but can sound better in the right situation.
I'd bet it's the 1.5-3.5K area you need to adress here.
I'd bet it's the 1.5-3.5K area you need to adress here.
Harshness mostly means too much info in the most sensitive treble hearing range OR distortion in that area (4 - 8khz) . If it's just too much you can take out around between 4 - 10 khz. If it's distortion that's a whole other problem; might even be attacked best with a warm overdrive... Hard to tell what to do without the example 
Btw, the 'air' on vocals lies between 12-16khz BUT to make it nice in the digital domain you will need to put a high cut after that (above 16khz).... or use Pultec which will do this for ya
cheers
Budy
Btw, the 'air' on vocals lies between 12-16khz BUT to make it nice in the digital domain you will need to put a high cut after that (above 16khz).... or use Pultec which will do this for ya cheers
Budy
Screaming and growling are most sensitive at 1,5-3,5K depending on the voice/style and will usually need some kind of multiband process to sit right. The reason for that is that normal compressors won't \"see\" that much of the extra harmonic content that is added since screaming and growls are just added harmonics to the \"main\" voice. So you can have a compressor reduce 5db on a voice, and if the singer starts to deliver in a \"screaming\" way, the ear will hear that as quite a bit louder because of the extra harmonic content, but the compressor might even compress LESS on those parts compared to normal voice since when modulating the voice in that way also decreases the energy of the fundamental compared to normal voice. The compressor will \"see\" the loudest part of the signal and adjust according to that, and since the loudest part of ANY vocal, normal or screaming is the fundamental.. that is what's going to trigger the compressor.
So what you need is something that also controls the area where the dominant harmonic content of screaming recides. So this is the perfect use for frequency dependent compression. It's a fun phenomenon when you try to stick a \"normal\" vocal up there where you have bands that generate a very thick sound image that usually is accompanied by growls and/or screaming vocals. The vocals always need to sit at a certain level as not to make the rest of the music sound \"small\" in comparison. And so when you have that vocal level where the rest of the music sounds its best, you find that you just can't have a too cleanly recorded or sung vocal since it won't cut through. No amount of compression and eq will fix this, just the right kind of \"dirt\" generated by the vocalist will make the magic happen, even on a clean voice.
Like boody says.. if it's 4-8K it's probably a problem all the time, since that range of frequencies isn't anything that will change a lot with vocal technique and delivery if we're talking normal vs. screaming.
So what you need is something that also controls the area where the dominant harmonic content of screaming recides. So this is the perfect use for frequency dependent compression. It's a fun phenomenon when you try to stick a \"normal\" vocal up there where you have bands that generate a very thick sound image that usually is accompanied by growls and/or screaming vocals. The vocals always need to sit at a certain level as not to make the rest of the music sound \"small\" in comparison. And so when you have that vocal level where the rest of the music sounds its best, you find that you just can't have a too cleanly recorded or sung vocal since it won't cut through. No amount of compression and eq will fix this, just the right kind of \"dirt\" generated by the vocalist will make the magic happen, even on a clean voice.
Like boody says.. if it's 4-8K it's probably a problem all the time, since that range of frequencies isn't anything that will change a lot with vocal technique and delivery if we're talking normal vs. screaming.
Dan Duskin
Established Member
Knowing what frequency to cut is entirely down to the singer... i.e., nobody can advise you without having a listen.
Sometimes Spectrum Analyzers are helpfull in this situation.
Open Cambridge, set a mid-treble band to the narrowest Q value, then boost it all the way! Hold the shift key, and smove the knob up and down until you find the offensive frequency. Then widen the Q a bit, set it to 0dB, and slowly start reducing level of that band below 0dB.
Multiband is a quick fix, and it sometimes works wonders. If you have more time, find the offensive frequency with the above listed methods.
Sometimes Spectrum Analyzers are helpfull in this situation.
Open Cambridge, set a mid-treble band to the narrowest Q value, then boost it all the way! Hold the shift key, and smove the knob up and down until you find the offensive frequency. Then widen the Q a bit, set it to 0dB, and slowly start reducing level of that band below 0dB.
Multiband is a quick fix, and it sometimes works wonders. If you have more time, find the offensive frequency with the above listed methods.
honestly, I'd say don't worry so much about it unless it's REALLY really bad. There's no reason to go through all this multiband compression / eq / compression crap otherwise.
If you wanted a 'nice' sounding vocal, I don't think you'd be screaming. Put an 1176 in all buttons mode and turn everything up all the way. No one will be any the wiser.
Matt
If you wanted a 'nice' sounding vocal, I don't think you'd be screaming. Put an 1176 in all buttons mode and turn everything up all the way. No one will be any the wiser.
Matt
saemskin
Established Member
you know, I was just thinking about the topic question, and it seemed to me that maybe you are tracking too hot?
I was recording some vocals the other day. It was a Rode NT-2A into a Gyraf G9 I made. Now, I was thinking to crank up the input gain all the way to get some great saturation. This might have sounded kick ass on some nasty bass line, but on vocal it was awful. If I pushed at all and the signal sent yellow on my Echo metering thingy I got some crackly harsh vocal. Which is not something I would ever try to EQ out. It just wont work. So I adjusted the input gain back down to about 60%, and backed up a few inches from the mic. So I was about 8 or 10\" away.
Much nicer sound that way. No eq, some gentle LA2A compression of about 2 db's and I'm off and running. Now if only my voice didn't sound like I swallowed a bubble I'd be all set. O
Think about trying to fix this before you even start recording. Maybe your mic is too bright for the voice, maybe you need to step back? Could be anything.
I was recording some vocals the other day. It was a Rode NT-2A into a Gyraf G9 I made. Now, I was thinking to crank up the input gain all the way to get some great saturation. This might have sounded kick ass on some nasty bass line, but on vocal it was awful. If I pushed at all and the signal sent yellow on my Echo metering thingy I got some crackly harsh vocal. Which is not something I would ever try to EQ out. It just wont work. So I adjusted the input gain back down to about 60%, and backed up a few inches from the mic. So I was about 8 or 10\" away.
Much nicer sound that way. No eq, some gentle LA2A compression of about 2 db's and I'm off and running. Now if only my voice didn't sound like I swallowed a bubble I'd be all set. O
Think about trying to fix this before you even start recording. Maybe your mic is too bright for the voice, maybe you need to step back? Could be anything.