Originally Posted by mogorf
Hi, since I'm not familiar with MiniDSP, allow me to ask how many mic positions do you use during measurement of the subwoofer? Thx.
Originally Posted by garygarrison
That's my question, too.
As I understand it, one mic position won't do the job, or at least won't do the job very well, unless you have a room with modes of very low (inconsequential) amplitude, or multiple subwoofers optimally placed. Move your head a little, or change seats, and the frequency response will be different. Earlier on this forum we joked about an audiophile head clamp that would help a sole listener stay within the EQ zone correctable by data from one mic. An engineer I respect answered my question concerning "how many mic positions?" by saying "Good question -- probably at least 3." Audyssey obviously thinks 8 positions will be O.K., but then they add in their proprietary "fuzzy logic" substitute for an average, arguably an improvement. Some proud victims of OCD (like me) will more or less (anxiously) trust Audyssey, and its 8 mic positions, for the overall curve smoothing, then turn up the subwoofer to taste. Of course, to EQ for an audience of more than one listener, one must bounce from seat to seat to confirm that the bass is about right from each seat, and across a decent sample of bass frequencies. This can be amusing to onlookers, so it's best done alone.
REW (what I use to get the measurements to use with the MiniDSP) it depends on what I'm trying to do, and Audyssey I always take a grid. I'm mostly concerned about just my seating position, but Audyssey does better if it gets a bit of a scope of the room so after the first main listening position measurement (this is critical because that's when Audyssey grabs its distance and delay information) I move it around in a grid of about 1.5ft increments. I take all eight measurements. Some people go wider if they are concerned about more seats, but, you have to play with it a bit and it's a judgement call. If it's normally you and your girlfriend sitting in the sweet spot love seat, then I'd probably just concentrate in that area to optimize in a reasonable area but not compromise for seats that aren't often used. If you have a big group over, yeah some people will get a non-optimum seat but they won't care as long as it isn't really really bad (like being crammed up against the back wall or something where the bass is so loud they can't hear the dialog).
For REW, I do a mix of single measurements to get started and then taking measurements in a grid like with Audyssey (MLP, left of MLP, right of MLP, those three spots about 1.5ft in front, then two spots on either side behind the MLP...so total of 8 positions), but I pay attention to how much things are moving around. When tweaking placements often one or two measurements is enough to see if you are making progress, you don't need to take a ton of measurements every time. Then when you think you have a great spot you can take a bit of a grid again and average it. I find that this is working well when keeping within a reasonably tight listening area for two people. If you are trying to EQ for a much wider area than yeah you probably need to use a combination of averaging and specific measurements to see if you are compromising one seat you care about too much to benefit the average. The wider the area, tougher the compromises, most of the time.
I recommend not trusting Audyssey if you don't have to though, at least get REW and a UMIK and see what it is doing to make sure it passes the sniff test. The reason I'm wanting to avoid Audyssey EQing my bass is because it just happens to be getting confused and making some bad decisions that cost a lot of headroom. It's doing fine elsewhere though, so, keep what we like and remove what we don't.