The dreaded 3-4k of a female vocalist

[Kinh]

This has mystified me for sometime and my mixes as a whole have suffered as result.

Eqing the 3-4k range. This is where it peaks on the ee vowels, causing unwanted spikes.
The methods tried are:
Dynamic Eq
Automate Eq to cut when the ee comes in
Multiband compressor
Cut the 3k from my "color" send

I spend hours just on the ees alone and then I play the track on cheap ear buds or my car and out they come, louder than anything else.
I've listened to some jpop girl singers who tend to have squeaky voices and I notice the sharpness is consistent throughout their vowels. The oohs and ahs are just as harsh as the ees so you dont tend to notice them.

How do you guys fix these unwanted spikes? I want to get a consistent 'bright' tone without losing energy. There's gotta be a better way.

[Zukan]

Is this for submitted stems or content you are recording?

[CS70]

How do you guys fix these unwanted spikes? I want to get a consistent 'bright' tone without losing energy. There's gotta be a better way.

Do you mix in a room with speakers or on headphones?

[Kinh]

How do you guys fix these unwanted spikes? I want to get a consistent 'bright' tone without losing energy. There's gotta be a better way.

Do you mix in a room with speakers or on headphones?

Both

[Tim Gillett]

It really helps if we can hear what you are hearing. Can you link to an example of the raw voice straight off the mic, then your processing of it, then an example of what you consider good processing of a similar voice?

[resistorman]

Also, the choice of mic and mic technique really matters.

[Mike Senior]

This is something I encounter in project-studio recordings a lot, and usually deal with it with dynamic EQ or spectral processing such as Sknote's SoundBrigade. I've written about it in Mix Rescue a lot too. Here are a couple of recent examples:
https://www.soundonsound.com/techniques ... earranging
https://www.soundonsound.com/techniques ... -mcdermott

[Sam Inglis]

This is one of the few situations where I find parallel processing can be really useful. If you duplicate or mult the vocal, you can heavily filter the second vocal track to focus on the 'wanted' part of the mid-range. This varies a lot from singer to singer but might be something like 600Hz to 2kHz. Cut out everything outside this range, apply a generous amount of compression, and add it back in at a fairly low level to thicken the voice without emphasising the nasty resonances higher up the mid-range.

[James Perrett]

I probably sound like a stuck record but try ReaFIR in compression mode with the threshold reduced in the problem range. It is built into Reaper or part of the Reaplugs plug-ins if you don't have Reaper.

[Mike Senior]

I probably sound like a stuck record but try ReaFIR in compression mode with the threshold reduced in the problem range. It is built into Reaper or part of the Reaplugs plug-ins if you don't have Reaper.

Forgot about that, but I agree -- a big +1 from me there too.

[Tomás Mulcahy]

This seems to be something Oeksound Soothe could deal with.