General approach to mastering, am I going about this in the right way?

DrewPeterson7

Sage of the Order
So, I've been doing home recording in some form or other for damned near... 20 years now, I guess. It's a hobby, one of a couple, and one I don't have as much time to dedicate to as I'd like, but it's one I do take pretty seriously and over the years I've gotten to the point where I'm fairly happy with my recording and mixing.

Mastering has always been sort of a dark art to me, however. I've done the half-arsed "normalize, then lit it with a limiter and makeup gain to get it up to CD volume" thing on my own recordings, when it came time to master the album I released a couple years ago had a buddy of mine do it rather than trust myself (and normally I'd rather have a second set of ears on a project anyway), etc... But, I'm just wrapping up a folk-rock/roots-rock/blues project with my dad and uncle we're planning on sharing with the rest of our family but not really any further, and I didn't want to hire someone to do it for me or call in a favor from a friend or anything for a project that was going to get listened to a handful of times by fewer than 50 people. So, I've been going at it myself.

All that out of the way...

The general approach I've been taking is render all of the tracks down into stereo mixes, then load them up in individual tracks in a new Reaper program, stagger them out in running order so the whole thing plays back as if it was a single one-track CD, then normalize each of the individual stereo mixes within Reaper. Any fades I wanted were applied at this point.

Next, I went through and balanced individual tracks by ear to ear so even if they weren't all peaking at the same point, they sounded about as loud as each other. I haven't really done any EQ work on the indivudual songs because, well, I'm still tweaking the mixes a little bit so I'm trying to get everything right up front in the mix, and because so far everything does sound fairly good. Once all the mixes are 100% final though I may end up doing a little bit of light EQ tweaking here and there to make sure everything sounds as consistent as possible from track to track.

Then, I've thrown a series of plugins on the master bus. Off memory, I've got the Sonimus Britson bus in "loudness" mode (I like the Sonimus stuff, incidently, and used the Britson channel strip and their Burnley73 EQ heavily in this project) both for the slight EQ and to add a touch of crosstalk, ReaComp set for extremely light compression (maybe 2:1, I forget) with a threshold just attenuating the last 1-2db of the audio and with about a db and a half of makeup gain, and that into the Waves CLA 2A opto comp to add a bit of fairly smooth compression and juice the output another couple db (given the nature of the plug it's tough to state in quantitative terms how much it's doing, but the attenuation needle suggests it's taking 1-3db or so off peaks, and I'd guess it's in the 3-4db louder range coming out the other side). From there I'm hitting it with a limiter, Waves I think, to add maybe about another 3.5db to the perceived volume. Basically, stacking a couple compressors on top of each other, none of them independently all THAT hard. I've tried some other stuff - I snagged both the Waves J-whatever and Kramer tape sims in a recent sale and figured that would be a good way to add some of that "magical analog tape sound" to the mix, but honestly I thought in both cases the mixes sounded audibly better without them, more so for the Kramer than the other, and that the music sounded clearer and more dynamic on its own.

Anyway... As far as a general approach goes - line up all tracks in a single project, get them balanced against each other, then process the master bus - is this more or less defensible, or is there something stupid I'm doing here? On one hand, I get the "if it sounds good it is good" argument... But, if it could sound better if I did something else, and there's something that I'm doing doesn't make sense (for example, I could see fading the two tracks I want to fade out before they hit the series of compressors on the master bus meaning there's going to be a difference in how dynamics are perceived, somewhat anyway, between the main sections and the fades), well, I'd like to learn a bit from this project, too. :)

For now, I've been A/Bing against a couple modern Dylan albums - Time out of Mind and his Tell Tale Signs bootleg series - and while I'm nowhere near the engineer (or musician!) of all parties involved in that one, I'm at least not embarrassed by how they compare. So, I'm probably not doing anything truly stupid... But, I figure you guys will tell me, mercilessly, if I am. :) Thanks!
 
Sounds more or less like what I do. I try to use the least amount of processing possible, just adding a plugin when I feel there's a problem to solve. I'd consider putting compressors on individual tracks rather than the master bus, so you can fine tune them to the specifics of each track.
 
Sounds cool. I'd love to hear one of your tunes. Post up! I've also been experimenting with mastering lately. I have been using Waves Abbey Roads TG mastering channel strip and other Waves compressors as needed. Struggling with transients. I try to get 2-3 dB reduction through the limiter. If there are still peaks, I add on the ReaComp set to a fast attack and release and set to infinity. I set the threshold just low enough to catch those peaks.
 
I pretty much "pseudo" master everything I do, but what that means varies at lot.

So, if you are planning on any kind of mastering, just leave stuff off the master bus, as much as possible. Bounce your mix down and check loudness (LUFS) and dynamic range against your targets. Make adjustments now in your mix if it's too loud or tight to meet the criteria you'll need to match up with.

I'm a slave to the LUFS metering in various plugins and that's what I use *religiously* to "normalize" my entire tracks, regardless of whether I use Ozone 5, Ozone latest, or just some EQ, compression and limiting. I measure everything with the Orban Loudness Meter. Get your tracks to the same levels (more or less) and burn a CD w/o any kind of additional leveling. Listen and tweak.

I work to -14dB LUFS because a lot of what I do ends up syncing to a video recording and that works reasonably well overall. It's a little low for some more modern CDs but if you're peaks are ~-1dB it will generally be fine. Balancing dynamic range with overall loudness is a task that can take up a lot of hours, and does depend on the ultimate output format.
 
Since you guys asked - this isn't quite final, but it's fairly close - this is one of my uncle's song from this project. He sings and played the acoustic guitar, my dad played the keyboard part, I did the drum programming, played bass, and electric guitar. I wouldn't claim the performances, mix, or master are flawless here, but this is just a fun project we're doing for my family and my uncle in particular has never done any real recording before, so they're having a lot of fun with this. For me, considering I usually do Satriani-esq insturmental rock, it's fun to get out of my comfort zone and do some work in a different genre (and, you know, to record and mix vocals at all, haha).

[MP3]https://drewpeterson7.files.wordpress.com/2018/12/02-the-broom.mp3[/MP3]

Been referencing this against a couple Dylan albums, Time out of Mind, and Tell-Tale Signs, for perspective. They're clearly going to be better. :) I'm hoping to get everything wrapped up over the weekend.
 
Im fairly new to trying to master, is it always necessary to normalize. Hopefully thats not too dumb of a question :)
 
Normalize is just analyzing the level (loudest peak or integrated loudness) and then adding gain or attenuation to meet a target. Like, the loudest peak is -3, but I want it to be -1, so I add 2db or the average level is -16 but I want -12, so I add 4. You don't have to do that via the automagical process in your DAW. You could do it just by watching the meter and turning knobs. Normalize process can be a good way to find a starting point, but it usually needs to be tweaked by ear.

Most of the time, though, you'll find that "normalizing" for a target averege puts the loudest peaks all over the place, and that's when you end up having to compress, limit, and/or go back to the mixes.
 
Sony Vegas had a clip gain function a decade or so before Pro Tools got it, but it's a downward-only adjustment. On the other hand Vegas does have an instant normalization tool, so you can normalize then trim down with clip gain easily enough. That's really the only time I use a peak normalizer.
 
Back
Top