This was a fun one from TELEFUNKEN’s Live From The Lab. It’s called “Buildin’ It Up” by the McLuvins. I love this song - especially the bass. The mix definitely presented some challenges. The drummer is singing, and his mic has tons of bleed. In fact, the main vocal track is labeled “Vocal/Drums”.
I wanted to try out my new Kush Audio REDDI plugin and test drive Kush’s new AR-1. In case you’re interested - both totally kick @ss! I may have to reconsider my long standing anti-subscription and anti-iLok dongle stance…
From TELEFUNKEN’s Live From The Lab. Video and Multitracks here:
Hey Mike, cool track! Just listening on cheap speakers at present. Musically it sounds really nice - can’t be too helpful or more specific on these speakers on that count…
The thing that stood out to me as needing most attention is the lead vocal. It sounds really “muted” and “closed down” in the high end. It also has some pretty obvious compression artifacts, especially when the vocalist pops the mic with “p” and “b” sounds. Maybe get that proximity effect under control before compression with some automation of low eq, and then perhaps use multiple compressors more gently on the lead vocal, and allow the top end to breathe a little more.
sounding good. Does the ‘p’ on pressure jump out? I think the vocal could use some gentle massaging to get it to fit more nicely without getting so distorted. Oh and as i am writing this Andrew enters his response. read it. it is more helpful. Thanks for sharing.
Thanks for the listen guys. Yes, that initial “P” really sticks out. I clip gained it - a lot - but I didn’t want to overdo it. Also, the trouble with the vocal track is that it also has drums - in fact, the main vocal track is labeled “Vocal/Drums”. No doubt it sounds muted because I tried to EQ some of the cymbals out.
If anyone else mixes this, I’d love to talk about how to deal with some of these things. The audio on the YouTube video certainly dealt with it well.
I’ll just come in and say that there’s some funky stuff going on with the vocal eq/compression.
I’d say forget about getting rid of bleed and focus only on making the vocals sound as good as possible. After that, then you can start thinking about killing bleed if the bleed is screwing up the mix. If it’s not screwing up the mix, then don’t put any thought into it.
What Boz said… you literally have to allow the ‘bleed’ to be a sound that you keep, that means work from the vocals down, or everything should be solo’ed with the vocal ALSO on it.
This may be tricky to start with as the drummer is singing, but the reference track is really well balanced, so work to that.
eg. Leave the cymbals in the vocal tracks, then reduce them in the drum tracks.
It still sounds a tad dull in the top end of the vocal, can’t you just chop up every line, and manually pull the loudest ones down - its quite quick after you get the hang of it.
That is always one of the first things I do. I normalize the track, then hunt down the sections or peaks that are the biggest. Then normalize again until I get a relatively even track for processing.
That sounds like overkill, try leaving 6- 10dB of headroom, don’t normalize anything unless you have to - then pull the level back down a bit afterwards, if you do normalize a track.
Those peaks are important, try to keep them away from 0dBfs while you process them. Compression will dull them too, so use your attack generously to keep them intact.
I use compression more as an effect than a signal leveling tool.
Maybe also try using a tilt EQ like this free one with your compressor to select a neutral Eq curve that flattens out the lows and maintains the highs.
With 32 bit float this depends on the particular plugin, right? We don’t need to really worry about digital clipping anymore until we’re actually rendering the track. At last least that’s my current understanding.
Maybe THIS is the problem? I was trying to be “gentle” and slowly step the levels down by 2-3 dB each time to even it out…
…and THIS! I know you’ve said before that you use tilt EQ a lot. I obviously need to learn to “hear” what you are talking about and read up on using this type of EQ.
Here is my automation for the lead vocal - Im just mixing drums and vocals with no compression. Its quite the Manhattan skyline (the green line - sorry its dark
Cool! Thanks for posting that. Just to make sure we’re talking about the same thing - is that “automation” or are you slicing and clip gaining the slices?
My initial clip gaining of the peaks/sections to deal with big peaks would look like that, only on a MUCH less granular scale, that is, I’d have maybe 10 spots or chunks. Then I merge the parts and re-normalize the track until I get a fairly level track. I guess I just do them a few spots at a time.
PS. I DO have that tilt EQ but never really use it. HOFA makes a nice one too. But I picked up Boz’s T-Bones a while ago and reach for that when I want to try one out, which isn’t very often. Obviously, I need to learn how to use a tilt EQ better!.
That’s only a third of the song so I probably have over 50 slices in there. That M80 is a little fiesty with plosives and proximity effect, the bleed does sound good though. The Heil PR35 would have behaved much better.
I’m just reducing levels, except for a coulple of really low spots.
I may have to reconsider my long standing anti-subscription and anti-iLok dongle stance…
Eh-hem! I was fiddling and checking the guitars right before I rendered it last night… will fix it tonight.
