Mastering Experiments, Part 2: EQ & Imaging
This is the second half of the thrilling chronicles of my attempts at mastering the piano recordings. (Here’s part one.)
Mastering Experiments, Part 2: EQ & ImagingPaul Cantrell, pianoDownload (12:42 / 15.4 M)
I’m constantly changing things — I’ve tweaked the process since my last post, and even while making the explanation, I suddenly noticed a new EQ adjustement. It never ends. These experiments are now coming up against the limits of my ears, the point where I spiral endlessly varying some parameter or other, eventually unable to tell whether the result sounds better or worse, or even any different at all. This week, I’m going to enlist the aid of some more knowledgeable friends, and of the listening public (that would be you!), then call it good and move on.
And yes, I really am interested in how it sounds to you, on your speakers and to your ears.
Once again, Logic Express is heavily involved in what you hear, as is some custom code of my own which handles the stereo image manipulation. But the real software star here is Firium. I’m generally unimpressed with the quality of audio software: it’s typically convoluted, opaque, crashy, ridiculously finicky about its environment, and an embarrassing distant last place in getting compatible with a new OS revision or new hardware. Even much-praised Logic, while it has an excellent set of capabilities, suffers from most of these complaints. It just doesn’t feel polished; it’s certainly no Adobe Illustrator.
And then there’s Elemental Audio’s products. They’re elegant. They offer powerful capabilities through a simple, carefully considered feature set, expressed in interface that explains itself clearly and makes what’s most important most obvious, yet rewards exploration and handles exceptional needs gracefully. On top of all that, to my ears, their stuff sounds fantastic.
Comments
Your tweaks make a big improvement in the level of detail in the sound, even on my crappy speaker setup; it was already good, but this takes it (forgive the cliche) up a notch. Everything is quite a bit brighter now, but without becoming too sharp-edged or metallic. On the other hand, I can’t hear very much soundstaging difference.
One interesting revelation for me was how much sharper and brighter the M channel is vs. the S. In hearing your sample played over and over, one imaging problem I came to notice is that the repeated note in the mid-bass seems fuzzy and overly persistent; it tends to obscure the rest of the bass sound. When you split the M and S channels, this effect appeared (to my ears) almost exclusively in the S and was absent in the M, and your boosting the M made a bigger difference to fixing that problem than any of the other changes. Is this effect particular to the sample (or to your recording setup) or does it reflect a general tendency?
I totally agree, Nick, that there’s something fishy in the bass of the S channel. One of the many things little things I gloss over in this explanation (I simplify things a bit) is that I reduce an area of low bass in S. I sometimes think I could reduce it even more, but it starts harming the lovely richness of the bass. It’s interesting that the M channel made such a big difference for you; that may be entirely specific to your setup. On both my fairly well-balanced headphones and my stereo, M is actually a hair thin too thin in the bass. What kind of “crappy speaker” setup do you have? I might send you some “pick the best of these three” sorts of experiments.
You’re very unlikely too hear any imaging / soundstaging difference unless you’re sitting right in the sweet spot of two well-placed speakers. (You can also catch some of the difference listening on good headphones, though it comes off a bit differently.) The stereo image comes out of the mikes pretty wild to begin with — there’s some really serious phase mismatching (which I love, though classical engineers hate it) and timing differences between the speakers which vary from note to note. So in all of that, the changes I’m making aren’t likely to be noticeable if you’re listening on computer speakers. If you can hook your laptop up to those nice electrostatics you have, that might do the trick.
I’m listening through a set of Altec Lansing computer speakers– two small main speakers and a subwoofer. The crappiness comes partly from that and partly from the loud fan in my docking station. Unfortunately, we packed away the big Magnepans when we showed the house to buyers, so as to help declutter the living room. Once I get my PowerBook (probably shortly after Tiger releases on 4/29) it will at least alleviate the docking station noise, and I may buy some headphones then too.
I also thought the M channel alone was too thin in the bass; but what bass there was was better-defined.
Fascinating stuff. It gives a whole new definition to “ear training.” I’d make a lousy recording engineer. I find that I grow impatient, even just listening to you recap what obviously took hours and hours and hours to do. I’m so glad there are good engineers and tools out there, and that I’m not in charge of any of them!
When you do music for stereo you’ll need the right equipments for it and have many years of training