Home Theater Forum and Systems banner

1 - 4 of 4 Posts

2 Posts
Discussion Starter #1
Dear Sirs,

Although I listen to music for a long ago, just by now I decided to start studying how to correct rooms for a better listening. Just downloaded REW and started reading some messages.

I don't have a calibration mic but a Radio Shack SPL Meter that I think could help me, once I´m interested on adjusting the bass frequencies up to 500Hz. As the title says, I don't want to use the sub to listen to stereo, since my speakers (PSB Image T6) are full range and play bass very well.

I'd like to know how do I make the measurements (1 channel or the two channels at a time) and how to interpret the graphics to make the better correction possible (by EQ or even by changing listening or speakers position).

Anyone to help this brazilian newbie?


Super Moderator
9,235 Posts
Ok, just wanted to make sure there wasn’t going to be an “oops” moment when it was discovered that you didn’t have the necessary hard- or software after I gave a lengthy explanation. :D

There is a transition point where speakers can’t be equalized separately, roughly above the 3-400 Hz range. Above that point independent filters whack out the soundstage. You’re on the threshold of that, being only concerned with response below 500 Hz. Below 3-400 Hz you can measure both speakers independently and EQ them independently. If you need something done above that point, you can give it a try. But look out for any loss of coherent soundstage in the frequency range where you’re equalizing. If you hear any, then the two speakers will need matching filters at that frequency, not independent.

As far as interpreting graphs, and especially applying filters based on the graphs, that’s tricky if you don’t know what you’re doing. Basically you’re looking to take care of the worst issues, because the small ones aren’t especially audible and it makes no sense applying filters that don’t make an audible improvement.

To determine what filter parameters to apply, it’s easy enough to look at the graph and determine at what frequency and how many dB a peak or trough is. Determining the right bandwidth is a bit trickier, but pay attention to the frequency locations on the graph where the problem begins and ends.

For example, if a peak begins at 125 Hz and is back to normal at about 250 Hz, it is one-octave wide. The starting point bandwidth for the EQ filter applied will be half that amount. So a half-octave filter. If the peak begins at 100 Hz and ends at 160 Hz, that’s 2/3-octave, so a 1/3-octave filter is the starter. And so forth.

I said “starting point” because it is the electronic nature of filters to spread more or less with the amount of boost or cut applied, as you can see in the picture below. So if you’re dealing with something like an 8-dB peak or trough that’s an octave wide, the filter that perfectly addresses it will probably be less than 1/2-octave. You know the filter is a perfect fit when you’ve applied the boost or cut to the peak or trough and the outer parameters (e.g. 125 and 250 Hz in the first example) do not change. If the outer-parameter frequencies drop or raise, the filter is too wide and should be adjusted tighter.

Make sense? If you feel you need help, please post a graph and we can make some recommendations.

Here’s a case-study thread on full-range EQ you might want to review. Spridle’s Experiment

1 - 4 of 4 Posts