Feature: Global parametric EQ (and VST support)

[mpogue2]

See also #71.

I find that I more often want to compensate for room or speaker acoustics, and less often do I want to change every song separately.

I've been thinking a lot about this recently, and I think this is what I want:

• global multi-band parametric EQ
• I'd like this in a new Preferences tab, called "Global EQ".
• similar UX to what is in #71. I do prefer the ability to show control points as draggable dots, rather than having a jillion vertical sliders (I think the latter is more complex), but I'm open to this. Lots of sliders is the easier but perhaps less elegant way to implement this.
• I also want two pulldown menus to set the initial position of the EQ sliders:
  • the speaker we're connected to: e.g. "Kustom PA-50/Powerwerks PW-50"
  • the speaker we want to sound like: e.g. "MA-220 + Yak stack"
• After that "Set Initial Curve" button is pressed, I then want to be able to tweak it from there, to taste
• The 3-band EQ name on the Music Tab then changes to "Per-song EQ", but it continues to work the way it does today, AND it adds/subtracts from the global EQ, rather than being the only EQ applied to a song. Basically, overall EQ = global EQ + per-song EQ.
• Ideally, there is a way to use the laptop running SquareDesk as a calibrated mic, to send white noise to the Line Out, and measure the actual speaker response curve, so that new speaker curves can be added easily, OR maybe there's an iOS/Android app for a cell phone that works remotely (I walk into the center of the room, press a button to play white noise or some other calibration tone, and it figures out the actual response curve of the speaker/mixer, and sticks the compensating curve into the global multi-band parametric EQ as a starting point). This could also help to set the default music levels, too!
• There's probably a corresponding mic audio path test (and another pulldown menu) for commonly used microphones (e.g. ND767/A) and output equipment (Hilton, Kustom, et.al.).

I've also measured the response of the Kustom PA-50, with bass and treble controls set to 3 positions. We can see that it's probably not as good as it could be from a square dancing point of view (look how little treble it has -- I've noticed that I tend to turn up the treble a lot), but I'm curious to compare it to the MA-220 + Yaks. I suspect that global EQ would allow us to dial this into an overall sound that is more exciting, easier on the ears, and has more intelligibility for the dancers, even all the way to the back of the room.

[mpogue2]

By the way, that purple dashed line in the plot above is actually measured data from the white noise generator I used (which is why it's not 100% perfectly flat). So, I'm fairly confident in the methodology measuring the actual current response of the Kustom PA-50 speaker. I have moved the line up and down to match the "pivot point" of the two tone controls, though. I think this shows more clearly what the bass and treble controls on the PA-50 are actually doing.

I did not measure the interaction between bass and treble tone controls, just 3 positions of the bass control with the treble control held constant, and vice versa.

There is a bump up of about 2.5dB right around 2.5KHz, which is exactly the opposite of what we want for intelligibility (see more discussion on this in #71). We actually want it to be lower at that frequency.. A checkbox for "Mike Sikorsky-style intelligibility improvement" might make it easy to add this to every song (without having to edit each one individually using Audacity or Adobe Audition, etc.). Alternately, a per-song parametric EQ (as in #71) would make it possible to tweak it on a song-by-song basis.

[mpogue2]

For comparison, here are the K8 curves. On the top is what they are when measured. On the bottom, I've moved the "vocal boost" curves down by 3dB, so we can see the effect of the vocal boost setting, without the added side effect of boosting overall volume that the vocal boost setting gives us.

This shows why I tend to run the K8 in Ext Sub/Vocal Boost mode -- we can see that this speaker defaults to putting out a LOT of bass. But, in ExtSub/Vocal Boost mode, it's much closer to flat, and then we can tweak the EQ to taste at the SquareDesk side.

Also note how much more high frequency response the K8 has. It's a much "crisper" sound. The Kustom PA-50 ($80) is rated at 50W (before it starts distorting), and it's fine for small venues, with the treble turned up all the way. The K8 ($650) is rated for 1000W of output (before it starts distorting), and a single one of these works great for large venues, or venues with a large amount of noise (e.g. a one night stand). I love the crisp K8 sound in ExtSub/VocalBoost mode, it's quite small and has excellent horizontal dispersion (to cover more squares with a single speaker). But, it's also a lot more expensive.

[mpogue2]

I'm also thinking that it might be simpler to allow a single EQ VST/VST3 component to be chosen as "SquareDesk Master EQ". Then, the user could save/load eq profiles at will. I believe libbass has the ability to load such plugins dynamically. This might be easier than trying to build the UI for a mutli-band parameter EQ -- there are LOTS of them out there already that work on Windows and Mac OS X.

For example: https://www.kvraudio.com/plugins/free/eq/most-popular

I think the key will be figuring out how to view the full GUI from within Qt. This is apparently not trivial.