In Reply to: Averaging of multiple locations posted by Inmate51 on December 9, 2014 at 07:58:35:
Averaging locations makes no sense. I don't run around the room 30 - 150 times per second to listen to my stereo. I generally stand, sit or lay down in one place, and THAT is the measurement location I'm interested in.
Stereophile's measurements are taken from the listening area, not all around the room. I don't remember how big of a listening area they try to cover, maybe a sofa's worth.
I have found that averaging multiple measurements taken from around the vicinity of the sweet spot produces a response that correlates better with what I hear than a single measurement. I usually confine the measurements to an area about 2 or 3 feet wide, 1 or 2 feet deep, and 1 foot high centered on my head position when seated in the sweet spot.
Single-point measurements are full of peaks and nulls and most of them are highly location specific and shift in frequency as you move your head around. Spatial averaging will tend to eliminate peaks and nulls that are highly location specific and tend to be inaudible, leaving the ones that are consistent across the area around the sweet spot which are typically audible.
Smoothing a single point measurement using a fixed-octave filter is effective at eliminating the high frequency comb filter artifacts, but isn't as effective at sorting out which peaks and nulls are most audible in the bass and lower midrange. Also, if you first average the measurements from multiple locations and then filter the averaged result, you can get by with a narrower filter bandwidth so you see more detail.Further, measuring a response in one room does not provide a person with the ability to apply that data to any other room. In short, in-room measurements tell us more about the room than about the loudspeaker, unless it's a time-gated measurement.
I generally agree, but I still find it helpful when Stereophile compares the measurements from different speakers in the same room. Sometimes the comparison reveals something that wasn't obvious when looking at the pseudo-anechoic measurements. I also find that their in-room measurements help to shed light on the power response. For example, when there is a suckout in one axis accompanied by a flare at the same frequency range in another it's hard to tell what to expect at the listening position without an in-room measurement.
This post is made possible by the generous support of people like you and our sponsors:
Follow Ups
- RE: Averaging of multiple locations - Dave_K 15:56:26 12/09/14 (10)
- RE: Averaging of multiple locations - Inmate51 20:42:43 12/10/14 (7)
- RE: Averaging of multiple locations - Dave_K 08:57:35 12/11/14 (6)
- RE: Averaging of multiple locations - Inmate51 17:59:31 12/12/14 (1)
- I did read it, and disagreed with it (nt) - Dave_K 08:33:24 12/15/14 (0)
- RE: Averaging of multiple locations - J. Phelan 16:29:36 12/11/14 (3)
- RE: Averaging of multiple locations - Dave_K 10:59:22 12/12/14 (2)
- RE: Averaging of multiple locations - J. Phelan 12:20:11 12/12/14 (1)
- Soundstage does do THD+N measurements - Dave_K 15:05:39 12/12/14 (0)
- RE: Averaging of multiple locations - jasonpatrickbrowne 07:15:11 12/10/14 (1)
- Agreed - Dave_K 10:44:42 12/11/14 (0)