Looks good. The slope Michael is proposing looks to be around 1.5dB/oct. I've never heard of using such a slope for mixing but I can't assume he's wrong.
If you want to work on your nulls, you can divide the wavelength by 4 to find physical boundaries which might be contributing. I know 70Hz is about 8' so a null there would be partially due to a boundary being either 2' (1/4) or 6' (3/4) away. Any boundary at 4' (1/2) or 8' would help create a peak.
Know that, at least in theory, you can move or eq your speakers in complimentary ways so a null is balanced by a peak in the other speaker. I say 'in theory' because you also have to worry about symmetry, etc.
it could operate a very tight Q's over multiple frequencies making it transparent?
The answer has its roots in live sound. Feedback is generally a narrow-band phenomenon - meaning it's a single sinewave-like single, or a couple of sine components. So if we're aiming to remove feedback from a system, we want to use as narrow of a filter as practical to do it so we do less collateral damage to the surrounding signal material. A lot of folks will go to a standard 31-band GEQ which has filters about a third of an octave wide. As you can see, that is removing an enormous amount of spectrum if you're chasing a narrowband feedback signal. So that's why units made specifically for dealing with feedback offer very narrow (hi-Q) filters.
I would suggest for you as a starting point to aim for the rough target that I've imposed here with the red line. The exact amount of "tilt" necessary should be determined by ear, adjusting it until music tracks you know well sound "correct" in the space. But that should get you pretty darn close to where you need to be. I'd do this with only a few filters - a very wide cut maybe 4 dB centered on about 150 Hz, and a gentle hi shelf cornered around 3k and cut down perhaps 4 dB. That should get you very close. You may choose to flatten out the LF a bit more, as you will get more of a tilt up when you bring both sides on.
Post by Michael Lawrence on Oct 6, 2019 21:59:02 GMT
The crux of the matter is that music has a pink-ish spectrum. There is more energy lower in the spectrum and that slope is - surprisingly - relatively consistent regardless of genre. The biggest difference is the "sub bump" - EDM has a bit more. Classical has a bit less. Since music doesn't "sound correct" playing through a flat PA system, most sound systems and playback devices intended for music have some manner of spectral tilt as well. Some have more than others, and rooms tend to contribute a lot of LF tilt as well, generally with larger rooms contributing more. So the idea is to get your system in a tilt that sounds reasonably correct in your room so the mixes translate elsewhere. If your system is too tilted, your mixes will sound fine in there (because you're mixing them there) but thin and unsupported out in the wild, and if your system is too flat, you'll have the opposite problem.
This image shows the response of a large format concert PA system - one side, without the subs - "out of the box" (before tuning). If you walk up to a PA at an outdoor festival that's more or less what you can expect the system to sound like. Inside in a reverberant room, you'll see more tilt-up so you want to pull it back a little bit. Here's how we tuned the rig for the show I worked this weekend, and the FOH engineer didn't ask for an adjustment to that when he fired up his console, which means it's pretty representative of what he's been running into. There's kind of a big debate in the industry about how much to tilt, or rather where the tilt should occur. (Flat PA = tilted mix at console, tilted PA = flat mix at console). In recording, you don't draw that distinction because you don't take your console into different venues each night. Carl Tatz, who tuned many of the major recording studios in Nashville, uses a tilt that's between 6 and 8 dB over the entire spectrum, which is about what I recommend to get started. I think Carl has some info on his website. Then if you notice that your mixes are generally bass-heavy, you tilt your room up a little more, or if you notice they tend to sound thin, you flatten things back out. Over time you will find the center if you are listening on a lot of different systems. Industry is kind of chasing its own tail on this one. But at the end of the day, go with what makes your references recordings - stuff you didn't mix - sound correct in the space and you're good to go. In terms of studies of music spectrum, I would direct you to the EIA 426B standard, and some of the AES long-term average spectrum studies such as this one and this one. There's also some information in Dave Gunness's lecture here.
Post by Michael Lawrence on Oct 11, 2019 13:21:38 GMT
The one time I tuned a recording studio flat at the insistence of the owner, they called me back after a week and had me come change it In my experience, most professional studios are tuned with a slight tilt. But the real answer is to tune it however you find things best translate to the rest of the world.