How Flat is Flat?


I have been trying to optimize speaker placement in my small, dedicated room for the past couple of weeks. I started out by calculating all the theoretical "best" positions following the advice on the Cardas Webstie, the Rule of Thirds (and fifths, and ninths) AND multiple iterations with the RPG Acoustics software. For each possible speaker/listener location, I measured Frequency Response using the Radio Shack meter and several test CD's, and conducted extended listening tests with a variety of music. Leaving questions of soundstagingf aside, I can't get very flat response regardless of where I place the speakers/seat. Response is flat (+/- 3 dB) above 100 HZ and from ~ 40-45 Hz, but plagued with problems in the 50-80 Hz range. The "dip" is in the range of 6 to 15 dB, with the worst frequency being ~ 60 Hz. I have used the SGHT Excel spreadhseet to calculate the room peaks/nodes, and there shouldn't be any nodes at this frequency at any of the seating positions I have tried. I am pretty certain it's not the speakers (SF Guarneri Homage), since they measure flat in an anechoic environment. The room is 14 ft long, by 12.75 ft wide by 7ft9inches long (dimensions chosen using the SGHT spreadsheet based on the fixed height of my basement ceiling). I have minimal sound treatment at this stage (RPG foam at the reflections points, RPG Diffusers behind the listening position, Corner Tunes at the top and bottom corners behind the speakers). My questions is...is it desirable to do better, or am I just being REALLY anal? If desirable, is it possible to do better? How (short of getting a Sigtech or similar)? Would also appreciate any comments on the best scheme for speaker positioning you've tried. Anyone have experienve with the RPG software? The Cardas-based set-ups sound good..but the bect sounding one is VERY awkward even in a dedicated room! AlexC
128x128alexc

Showing 1 response by greysquirrel

I think Carl is right. Just moving the speaker/listening position and absorbing/diffusing first reflection points will not smooth out the low frequency response. For a given room dimension there are optimal locations to reduce the peaks and valleys, but you'll still have peaks and valleys. You smooth out those peaks and valleys by absorbing low frequency standing waves (most effective in the corners of the rooms since this is a common node for all frequencies and foam won't absorb deep enough nor flat enough). This really doesn't get rid of standing waves, it just reduces the amplitude/decay of those waves, hence smoothing out the peaks and valleys you measured. Tube traps are excellent, though a little salty. I'm building my own, based on F. Alton Everest book, "Master Handbook of Acoustics". I'll let you all know if they are effective when I'm finished. Good luck, Greysquirrel