How to EQ voicelines properly?

[Kpheeyat]

Tips? tutorials? All help welcome.
I'm not really sure where and to what extent should I drag the EQ's up or down.

[Back-From-Purgatory]

I'm by no means a master at EQing vocals... but I can tell you that it heavily depends on the singer, the style of music, the instruments used, etc. etc...

There isn't some single magical EQ setting to make all vocals sound good. This goes for pretty much everything in music.

[DJ-Chilvan]

even if you don't use FL Studio, this tutorial should help. Youtube's your friend=)

[mrbean1500]

be sure to cut away any unecessary frequencies... especially if you're recording the vocals yourself

[moose3642]

Protip: never push any frequency band up by more than 3 dB. You can easily ruin a mix with EQ. Equalization is a subtle art.

[moose3642]

Oh, and here's a good method for finding frequencies you want to cut:

1- Make the bandwidth relatively small.
2- Boost that bandwidth really high up.
3- Sweep around the freqencies until you hear something you really don't like.
4- Reduce that band by somewhere from .5-3 dB below 0.
5- Widen the bandwidth.
6- If you're satisfied, stop. Otherwise, go back to step 1 and work on another frequency band.

[Decibel]

At 7/8/11 12:30 AM, moose3642 wrote: Oh, and here's a good method for finding frequencies you want to cut:

1- Make the sandwich relatively small.

I seriously read it that way. Gah, internet.

Anyway, be sure not to boost any unnecesary frequencies, You don't want the listener's ears popping.

[Kpheeyat]

At 7/8/11 12:30 AM, moose3642 wrote: Oh, and here's a good method for finding frequencies you want to cut:

1- Make the bandwidth relatively small.
2- Boost that bandwidth really high up.
3- Sweep around the freqencies until you hear something you really don't like.
4- Reduce that band by somewhere from .5-3 dB below 0.
5- Widen the bandwidth.
6- If you're satisfied, stop. Otherwise, go back to step 1 and work on another frequency band.

I know it might sounda bit much to ask, but I can't really visualize what you mean ;D Could you... visualize this somehow ;O? I can't tell what you mean by "something I don't like" ... like what? :P

[midimachine]

more often than not with vocals you can really benefit from completely removing low frequencies, and lowering the level of low mids.

But as BFP said it's really dependent on the vocalist and the rest of the mix

[Kpheeyat]

It's not really for any "mix" or vocals.
It's a normal voice. :p
Would knowing that make you have it easier to help me :p?
I'm still rather green here and I can't really "feel" what's wrong and right in sound =P

[Usernamemyarse]

HERES SOME ARTICLES from platinumloops about the matter, though they are about rcording and mixing rap/rnb vocals, there might be some stuff to pick up from there.

[Kpheeyat]

So I made myself some preset EQ.
Basically it looks like this.

[Kpheeyat]

Basically what I tried to do is follow the tips here to filter out the non-voice sounds(below 40 Hz) and boost the 125-250 Hz and 2-4 KhZ ranges.
Is that... good? =P

[Usernamemyarse]

Could you post some sort of a sample?
Its kinda hard to judge the quality without listening to it.

If you decide to post, upload two files, one "dry" (no EQ) and one "wet" (EQ).

[moose3642]

At 7/8/11 11:40 AM, Kpheeyat wrote:
Babysit me through this.

Okay! I wrote this for you. It pictorially and colorfully walks you through the sandwich modifying process.

https://docs.google.com/document/d/10ZS3 7ZwyOrylgCLOyDSsIP1BPFhTNP5CnP7SlAPTaDw/
edit?hl=en_US&pli=1

Ps, to the mods: these things should be in the stickies, since they come up so often.
Also, there should be some stuff in the stickies on the elements of a mix: lead, fill, foundation, rhythm. And some stuff on song organization.
There might be stuff about that in the stickies already, but I wouldn't know, because I haven't read them.

[Kpheeyat]

At 7/9/11 02:27 AM, moose3642 wrote:

At 7/8/11 11:40 AM, Kpheeyat wrote:
Babysit me through this.

Okay! I wrote this for you. It pictorially and colorfully walks you through the sandwich modifying process.

https://docs.google.com/document/d/10ZS3 7ZwyOrylgCLOyDSsIP1BPFhTNP5CnP7SlAPTaDw/
edit?hl=en_US&pli=1

Ps, to the mods: these things should be in the stickies, since they come up so often.
Also, there should be some stuff in the stickies on the elements of a mix: lead, fill, foundation, rhythm. And some stuff on song organization.
There might be stuff about that in the stickies already, but I wouldn't know, because I haven't read them.

some comprehensive sound editting tutorial + the voice acting/audio advertisments/wanted threads. Those could use a sticky, yes.

I am very thankful, but the thing I didn't understand was *which* frequencies do I prefer to have boosted or which ones I "don't like". Well, except for the very low frequencies, since those seem to be unnecessary.
I'll send the mp3 later.

[midimachine]

If you're just doing voice acting stuff then yeah don't play around too much with it. I'd avoid boosting anything at all personally BUT if you have a pretty raspy voice that boost at ~180Hz could be nice, but if you have a raspy voice then I'd avoid boosting high mids too... personally! I'd like to hear both a wet and dry sample (i.e. EQ'd/not EQ'd) to compare and maybe help out more particularly. Also, have you mentioned what kind of mic you use and where you record yet? That kind of info is also really useful to know.

And for fun: the red line is roughly what I feel sounds good for my own voice with my own microphone when I'm mixing it into my own music EXCEPT where the line plateaus at ~140Hz it would be more like -6dB than -12dB cause I wasn't paying attention the the y-axis when I drew that.

[Kpheeyat]

I refined my EQ. Here's how it looks like at the moment.
The attached sound file is:
1) normal voice(no editting)
2) My EQ voice
3) My EQ voice + compressor
4) Midimachine's EQ voice
5) Midimachine's EQ voice + compressor

As you can see, I'm also experimenting with the compressor. From my understanding, does it take the loud voicelines in check?
I suppose this should make things easier to note. My aim is to get some kind of "preset" EQ I could use. Well, until I get a reflexion filter, anyway.

[midimachine]

oh man, I didn't think you'd actually use that, since I posted it to show how different factors and situations affect how you should EQ a sound. That's kinda embarrassing!

although I still like it personally because I have a fetish for highpass filtering

As long as you're thinking about compression, you should learn about how they work. Also, de-essers!

[Kpheeyat]

I don't understand a thing from these 2 wiki pages, frankly :P
So is the current EQ(the one I did) ok?
And what should I use the compressor for? For what instances?
This technical language doesn't really reach me. At all =( I prefer some practical advice. Like If X happens(something noticeable on the waves?), I do Y. This "when X feels too much" explanation doesn't really tell me anything, especially since I don't have quality headphones, only some lame speakers =P

I know, I'm an antitalent when it comes to engineering. Or with anything mechanical, actually...=/

[Kpheeyat]

So I dug around the EQ(again) and further refined the EQ based on concrete information here and
here is the result.
1) No EQ.
2) EQ pic 1.
3) EQ pic 2(smaller)
I removed the boost around 125-250 Hz as that seems to make the voice sound dark. Which is good but only for that kind of voice.
I'm getting closer, I think.

Tips?

[moose3642]

The wave file you posted clips out at certain points. I don't know what you did to the parts that clipped, but try not to let that happen. You can tell that the audio is clipping if it goes above or below the track's visual boundaries. Most DAWs have a little red light on the dB meter that lights up if the audio clips while playing. You can also listen for distortion.

Towards the end, the voice sounds a bit tinny. That's good if you have a guitar or other instrument in the mid-low frequency range that you cut out, because it allows the guitar to take its place in the mix. You can also EQ the vocal range that you like out of the guitar, so that they both sit in their own little spaces.

If you don't have another instrument in there, don't make it sound tinny, unless that's what you want.

On that note, almost all of audio engineering is arbitrary. It's about tweaking and listening for what you want to hear. It's not about searching for magic numbers. There is no "right" way to do EQ.

You should actually try the tutorial I sent you before posting again. Try using REAPER (it's free but nags you until you pay) or some other DAW with realtime EQ feedback, rather than Audacity, which doesn't. Select your audio, turn looping on, start playing, and tweak the EQ. When sweeping, you will suddenly come across a frequency band that becomes very loud. If you like that part of the audio, keep it or boost it a bit; if you don't like it, cut it. Again, it's an arbitrary process.

[Kpheeyat]

At 7/10/11 03:54 AM, moose3642 wrote: The wave file you posted clips out at certain points. I don't know what you did to the parts that clipped, but try not to let that happen. You can tell that the audio is clipping if it goes above or below the track's visual boundaries. Most DAWs have a little red light on the dB meter that lights up if the audio clips while playing. You can also listen for distortion.

Towards the end, the voice sounds a bit tinny. That's good if you have a guitar or other instrument in the mid-low frequency range that you cut out, because it allows the guitar to take its place in the mix. You can also EQ the vocal range that you like out of the guitar, so that they both sit in their own little spaces.

If you don't have another instrument in there, don't make it sound tinny, unless that's what you want.

On that note, almost all of audio engineering is arbitrary. It's about tweaking and listening for what you want to hear. It's not about searching for magic numbers. There is no "right" way to do EQ.

You should actually try the tutorial I sent you before posting again. Try using REAPER (it's free but nags you until you pay) or some other DAW with realtime EQ feedback, rather than Audacity, which doesn't. Select your audio, turn looping on, start playing, and tweak the EQ. When sweeping, you will suddenly come across a frequency band that becomes very loud. If you like that part of the audio, keep it or boost it a bit; if you don't like it, cut it. Again, it's an arbitrary process.

This isn't music we're talking about, it's a simple voice... something not *as* varied as music. Hence it's easier to get the "right" EQ for that, no? Please correct me if I'm wrong. All other tutorials seem to suggest what frequencies to reduce/increase to change the brightness/darkness, presence/clarity or sharpness/dulnesss of voices...
All voiceactors do seem to increase around the same frequencies... etc.
I think I'll switch to Adobe Audition soon.
Any other tips =P?

[seel]

At 7/8/11 12:25 AM, moose3642 wrote: Protip: never push any frequency band up by more than 3 dB. You can easily ruin a mix with EQ. Equalization is a subtle art.

Laughable.

[Kpheeyat]

At 7/10/11 04:11 PM, seel wrote:

At 7/8/11 12:25 AM, moose3642 wrote: Protip: never push any frequency band up by more than 3 dB. You can easily ruin a mix with EQ. Equalization is a subtle art.

Laughable.

Actually, voice actors on many different websites something similar. they said the "safe" point is 6 dB but most changes are 3-5 dB.

For now I'll stick with reducing the unneeded low frequencies of 50-100sh Hz (my mic doesn't receive 0-50 Hz so that's good I think) and experiment on the cuts/boosts some other time. Like when I'll be doing a paid job, since this EQ doesn't affect things *greatly*, just "accents" what is already done :p

[Kpheeyat]

Like when I'll be doing a paid job, since this EQ doesn't affect things *greatly*, just "accents" what is already done :p

Or, well, at least when it comes to voices. Quite possibly this may not apply to music.