Guitar Gear Forum

Hey guys,

My band played a show last Friday at a venue that's known for solid sound (especially for a borderline dive bar) and for giving bands a soundboard recording of their set upon request. I have a recording of the set in my possession. I need some help. Here's the situation:

1) Most of the set is pretty fucking good (enough so that I want to be able to share some songs with fans and use them for booking more gigs).
2) However, the vocals are just a touch loud, presumably because they were the thing that needed the most help from the PA. Not unbearably loud, but just a little louder than I would like.
3) It's a two track stereo recording, so obviously there's no option for simply pulling the vocal slider back a few dB.
4) The vocals are panned down the middle, but so are the bass guitar and kick drum, so that limits the utility of the karaoke trick.
5) I'm working in Adobe Audition, which has a pretty good suite of filters/EQs/compressors and other effects.

Thoughts? Are there some frequency ranges I should be looking at where a judicious boost or cut would help the vox take a step back? Would a little multiband compression be more useful for that? If it's helpful, I'm happy to post a sample song here.

Thanks for any advice you can give!

well.. vocals tend to live over a pretty broad range- especially male vocals.. but between like 200-2k for fundamental stuff. i'd start by recessing there-- but guitar ALSO lives there.. which is really problematic! so no one shot eq or compression solutions gonna work.

you COULD try running a multiband compressor in there too- compressing the band, and pulling it down on the whole.. but you also lose a lotta dynamics in the drums.. particularly the snare and toms... so that kinda blows as well. undynamic drums sound super fake.

BUT i'd say you're onto something using mid/side compression in all honesty. take the channel panned down the middle and work with that FIRST- and do it in parallel so you can mix in the clean AND the compressed version to try to keep your fundamental tones and dynamics realistic, but be able to also clean up the too-loud vox. then find that fine line where you mix the comp'd signal back in enough to smooth it out without ruining your original dynamics.

do you have any mastering tools in adobe?

Isn't there a plugin that removes vocals from stereo recordings?

This sh!t is fascinating, but I am zero help, just here to learn a thing or two.

Looks like there are some YouTube tutorials... search for "Removing Vocals"
I suppose you could open a session in your DAW and create 4 tracks.
Use track 1 & 2 for L&R with vocals removed.
Create two more tracks, 3 & 4 for the original, L&R tracks as they are now.
During playback, you can adjust vocal volume by blending in more of the original recording.


Never tried it but, this is from a ProTools tips site:

"If you need to remove vocals from a song, first you need to open your DAW and load the song. Then, you need to split this track so go to Track and click split to mono. Now, mute this first track. Next, you want to remove the stereo by panning each the bottom to 0. Now, flip the phase by going to the bottom clicking EQ plugin. Now the vocal is pretty much gone.
Now you need to nudge one of the tracks forward so it will sound a little bit better over all. So now go to the nudge value and set it to one sample by selecting samples and clicking 1. Move this over two samples and see how it effects the sound. Now try 1 sample. You will have to toggle this to see what works best for each song. You will never completely remove the vocal because you cannot do this but you can knock it down so someone can sing over the vocal. You can also add in some bass if you lose it. Go back to the original track and put in a EQ plugin and roll in just the bass. This will help adding in back some of the bass that was lost. Follow these same steps will all of your songs and you can easily remove vocals from a song and make karaoke tracks!"

Removing vocals is possible, but it also affects aspects of the recording overall, ime.

I would mess with compression to try and push the rest of the mix forward a bit. Multi-band probably would be most successful, but even a regular compressor should give you some results

OK, after an agonizing few hours of no response, I got some responses! Awesome. I did some stuff, and I'll share my process and the results. So here's what I ended up doing, in order:

1) Used a parametric EQ to roll off pretty much everything below about 36Hz. It's just inaudible energy down there anyway, so I always do that when mastering live recordings. (Doom guys would hate me. I don't care. Fuck your beard.)

2) Applied a multiband compressor, with three bands I'll call "below the vocals" (up to about 110Hz), "with the vocals" (110hz - about 3.5kHz), and "above the vocals" (everything past 3500Hz). The thought was, if I could get it so that the vocals weren't making the other instruments duck, things would balance themselves out on their own a bit.

3) Used a parametric EQ again, this time introducing a couple of subtle dips: -1.4dB at 136Hz, -1.1dB at 580Hz, and -1.7dB at about 2.5kHz. I found these just by trial and error with headphones on, and they're based on where my vocals fall in the frequency range, and the peculiarities of that venue and set.

4) Added a touch of reverb. Otherwise soundboard recordings without a room mic are suuuuuuuuper dry.


How'd it work? Not too bad. It isn't magic, but in the headphones test as well as the "car stereo with the EQ flat" test, the vocals are definitely less pokey and annoying. Here:

Song #1, "All The Way to Houston"
Raw:
https://soundcloud.com/str8_2hell/all-the-way-to-houston-raw/s-Jvs7Y?in=str8_2hell/sets/live-mastering-comparison/s-lU18V
Mastered:
https://soundcloud.com/str8_2hell/all-the-way-to-houston-mastered/s-2pv9H?in=str8_2hell/sets/live-mastering-comparison/s-lU18V


Song #2, "You've Got This, Kid"
Raw:
https://soundcloud.com/str8_2hell/youve-got-this-kid-raw/s-XHoOt?in=str8_2hell/sets/live-mastering-comparison/s-lU18V
Mastered:
https://soundcloud.com/str8_2hell/track-2/s-vi3yv?in=str8_2hell/sets/live-mastering-comparison/s-lU18V


And finally, because I know you're all dying to stare at my fat ass, a friend captured video of one of the songs in the set. Turns out that the best sound of all consisted of me taking the mastered audio, syncing it with the video, and using about a 75/25 mix of soundboard audio and the live audio from her iPotato.

[video]https://youtu.be/UEwx3V08NWo[/video]