This may not belong here, but I will ask anyway...
I am working to set the Equalizer for a Church.
What is the best way to do this with REW and the RadioShack SPL Meter.
I am not working with a Sub-Woofer but standard amps, speakers, Equalizer etc.
I understand that this question might be beyond the scope of of this forum,
so I apologize in advance.
Thanks again for this very noob question.
Sincerely yours
Geist
Hi Geist,
I fully understand your desire to use measurements to aid you with your system EQ, but allow me to save you some frustration up front...
I make a living out of tearing equalizers out of pro sound installations. Back in the 80's, it became very popular practice to measure systems with a 1/3 octave RTA and then use a 1/3 octave graphic EQ to flatten out the response. The problem with such an approach is that it sounds like garbage because it completely ignores the entire time domain of acoustics - which is integrally tied to human perception of sound.
Btw, I'm not trying to say all system EQ is bad...just that it needs to be done with a deeper understanding of what's going on.
For starters, take multiple measurements of your room and note just how wildly the frequency response varies from position to position. One of the goals of a proper installation is to minimize the perceived variation. To sum up a lot of theory, humans hear sounds that arrive early differently than sounds that arrive late. So the first thing to arrive at the listener's ears is the direct sound from the speaker. Any reflections within the "Haas Window" (any sound arriving no later than about 20-50ms after the direct sound) are perceived as part of the original sound. So if you've got a speaker cluster up in the air and there are boundaries nearby, the boundaries will impart a change in timbre that the listener will hear as part of the direct sound. This notion is in contrast the many reflections arriving much later (like 100ms) which would be perceived as part of the reverberation of the room.
A normal frequency response measurement (like that achieved with an RTA) doesn't decipher between the direct and reflected sounds. The point I'm trying to make is that you can't EQ the reverb of the room....the reverb is a function of the acoustical space. You can, however, EQ anything that is minimum phase...which is essentially going to be your direct sound. Somewhere inbetween is a continuum where the EQ doesn't "fix" a flaw, but can possibly be perceived as an improvement to the sound.
So getting back to your goal....REW actually lets you gate the impulse response that the frequency response is derived from....which is essentially another way of saying that it will let you measure the direct sound independently from the reflections/reverb. The only caveat is that the more isolated your direct sound frequency response is, the less frequency accuracy you get....which is another way of saying that the low frequency corner of your measurement goes up in frequency as the gate gets shorter...and REW will show you the frequency accuracy.
So do you just gate your measurement and then tweak your EQ until it measures flat? Nope
Getting back to the above scenario where moving your mic around changes the frequency response....well, moving your mic around is going to change the direct sound too. You do not want to make any compensations at frequencies where there is dramatic deviation at different seats....unless that compensation results in more GBF (gain before feedback)...which is essentially another way of saying that you're tweaking the EQ to be flat on stage at that particular problem frequency. Just make sure you don't get too "tweaky" (steep Q's) when doing this because environmental changes can shift those problems around.
Also, you want to make sure you're not capturing the early reflections caused by the environment directly surrounding the microphone. You only want to capture all the direct/very early sounds emanating from the speaker cluster/array or whatever it is you've got happening.
If you have a stereo speaker configuration, then you'll need to run one stack at a time and find the common areas - double checking against the combined response. Stereo arrays are about the worst configuration you could employ, but often tends to be the common approach...
I'm starting to ramble a bit, but hopefully I've painted a picture where it's obvious that just obtaining a single flat frequency response isn't going to be a best solution...
If you don't want to dive into all the acoustical theory (and believe me, we're just scratching the surface here), then allow me a simple approach that I've found to always yield very good results...
Just use EQ to compensate for the anechoic frequency response of your speakers. If possible, pull your rig outdoors and try to load it similarly to how its flown in the church. So if your speakers are hung along the front wall near the ceiling, then take a measurement outdoors where the ground acts like the ceiling and then an outside wall of your church can act like the front wall....so basically, keeping it 1/4 space in both situations.
If you can't do this, then try to get a measurement mic as close to the speakers as possible (like 1m) - keeping in mind the coverage patterns of the various drivers (you might need to take multiple measurements to capture what the whole system is doing in the far-field).
And if you can't do any of that....just take an unsmoothed manufacturers data sheet and figure EQ out from there (taking into account polar patterns and how your power response will transmit to acoustic intensity). The manufacturer probably has a recommended filter set for their speakers too.
===================
Btw, the way modern speakers are these days, I would wager good money that you could get by without any EQ whatsoever (except for speakers that are designed with a specific EQ in mind). In fact, when I tour around with bands to venues that I didn't set up, I will usually bypass any system EQ.
Maybe I should delineate between system EQ and speaker EQ. Speakers (especially the high-end stuff) require EQ for the way the horns work and all that shnazz....I never bypass Speaker EQ as it is really part of the speaker design. System EQ, however, is an EQ that goes between the mixer and the speaker processing racks...usually it's in the form of a graphic EQ. My general rule of thumb is to never use a graphic EQ. That said, you will probably find me using it from time to time on the fly during a gig if I've got a system wide tonal change that I want to make....What I'm talking about above mostly falls under speaker and array EQ that compensates for the acoustical space directly surrounding the flown speakers.
Man, I'm seriously rambling so I'll just shut up now and let you ask questions (if I haven't scared ya off yet).
The basic things to keep in mind are the time-domain and how the sound changes from seat to seat (making it better in one spot often makes it worse in another) - and then the sound changes from day to day (so don't be aggressive with the EQ).\
Another good resource or two if you wanna do some reading:
http://www.prosoundweb.com/
http://www.synaudcon.com/website08/index.php (requires a membership fee and offers conferences where you can get hands on training)
