The Song Fight Bulletin Board

This article, The Dark Art of Mastering Music I found quite interesting, but I also wonder how much of 'mastering' is really unnecessary bullshit.

Very good read.

I know precious little about mastering. I know that the few times I've tried mastering my own songs, the output has been unpleasant. I know that LANDR does an okay job if you give it enough headroom, but that there's a noticeable difference between something auto-mastered by LANDR and something mastered by an actual mastering engineer. And I know that I paid a low-cost producer to master twelve of my songs last year, and I was very pleased with the results.

In my album, there were only a few specific moments for which I could point out what the mastering engineer had done (for instance, he widened a drum fill in "Audience," which I loved), but the overall impression was that the songs sounded more even, more polished, closer to how I'd intended them in the first place. One of the songs that had come together with some difficulty, and which I even considered cutting from the album before sending it off for mastering, came back dramatically improved and is now one of my favorites, even though I'm hard-pressed to quantify what the difference is.

So, I wouldn't call it "unnecessary." It's difficult to explain. But there was, at least in my case, a very tangible difference.

My two bits:

1. Improving your mix makes your mastering a lot easier, and there's some ways to make the mix a lot easier. Recently I've learned to actually love and make use of sidechain compression. I'll put my vocals on a mixing bus, and each instrument that interferes with vocals gets a sidechain compressor that's triggered by the mixing bus. (I'll also put an EQ on them that partially notches out the vocals' peak spectral range but the sidechain compressor brings it the rest of the way.)

2. Logic's "broadcast ready" mastering preset gets me like 90% of the rest of the way to levels and overall EQ. It basically sets up a decent set of presets for EQ, multiband compression, and an adaptive limiter. Then I just tune the limiter so that most of the song is at around -3dB and that the peaks go up to 0dB. It's nowhere near what a professional would do but for Song Fight it's more than sufficient.

LANDR seems to be a simple limiter with very little else being done. The "good" change is just loudness it seems to me. (I admit I've only dealt with a couple of tracks this way.)

Well, I can do that with a free limter. Add an EQ and a reference track, maybe some saturation and I can do better than LANDR.

I'll agree that a good mastering engineer can make a huge difference. I cannot do that. But for songfight, I rarely do more than a mild compressor, saturation/stereo spreader and a limiter to up the volume until it starts to limit then back it off a hair. Actually what I've learned to do that helps the most is mix with more headroom. I've "known" this for years, but it's a hard habit to stick with.

re: Mastering Audio. I've had the book for years and never read it. But just today I opened Mixing Audio by Izhaki again and I can tell that I need to re-read it. With a few years experience, I've already gotten more out of it. Although, a lot of it is reinforcement at this point.

Read the article in full and it was a good read. It also gives a better idea of what I should be aiming for when I "master" songs. It sure doesn't say how, though, does it? Guessing that Bob Katz book is gonna have to get read after all.