equalization curves

[johnpl]

Has someone experimented with room equalization using RTA analyzers? i am using a cheap Berihnger DEQ2496 for a couple of years for correcting my room deficiencies with good results. The thing that is confusing me is that totally targeted flat equalization doesn;t work for me. The sound becomes thin and i have to manually adjust some frequencies. The 50-200 hz ffrequencies have to be manually adjusted and the same applies to 15-18khz (attenuation of 6db). once these adjustments are performed things balance and the overall feeling is great. So did someone had the same experience as i did ? Does a non flat equalization curve works better for you nad if so can you give me the db curve ? Just consider that since my room is rather big (7 * 8meters) i have taken measurements from 5 positions with my mic calculating averages.This worked better than taking measurements solely from the listening position.

thank you

[Light Dependant Resistor]

Hi John
No I have not used RTA analyzers, congratulations on a great approach to assisting to get the best out of your room. As much as I dislike proprietary software there is apparently an application available for ipads and iphones
to assist, I am unsure if these are better or worse than what you presently use. A question when you find a problematic area, how do you solve it ? is it a matter of applying padding and how do you assess how much ?
or is the Behringer device then set to equalize the overall sound.

Cheers / Chris

[johnpl]

hi Cris
Behringer and all similar devices give yoy the ability to target a flat curve meaning that that automatically boost or attenuate certain range of frequencies. The second step isto manually adjust big ups and downs and fine tune certain areas. My problem is that this flat response curve the device is producing is not good to my ears -by the way i don;t have huge problematic areas meaning that the overal reverbation of my room is ok- and i have to proceed to manual adjustments for example in the 50-200hz area (6 db boost over the "suggested" by the device level ) in order to get a good overall feeling. if i don;t do that i feel the reproduction rather sterilized. This is why, to make it more clear, i was wondering if flat is not the way to go but perhaps a bell shaped (only in some frequencies) curve does the trick. I have experimeted with bell shaped curves but with limited results.
It is funny though that for many years i have been a "purist" ( the less interfering with signal is better....) and now room correction for me is the best way to go. You spend so little gaining immense benefits the only thing you need to do is get some basic knowledge on acoustics. No need to spend thousands on gear!!

[Rothchild]

You don't really need the RTA to tell you were the problems in the room are, you can derive them from first principals:
http://www.mcsquared.com/modecalc.htm or
http://www.marktaw.com/recording/Acoustics/RoomModeStandingWaveCalcu.html

a 7x8x2.5m room is going to have some funky bumps around 70-80hz and 250hz (and multiples thereof)

Using and RTA is however a useful tool for analysing any changes you make via room treatment or EQ (if you must...) Here's my cheap /free method using a computer and omnidirectional microphone:

Get Reaper www.reaper.fm
Get Span http://www.voxengo.com/product/span/
Get some test files: http://abluesky.com/wp-content/uploads/_HR-Images/BlueSkyTestFiles.zip

In reaper set up 2 tracks, one is the source (with the -20dB Pink noise file in it) and an instance of span so you can see what the curve of the pink noise is. The other is the return (with the microphone feeding it) and another instance of span which you'll be using to analyse your room - make sure you turn off the input monitoring on this track so you don't get horrible feedback.

The idea is to blast the room with pink noise (a known source with a predictable curve) and look at the frequency 'imprint' of the room on the curve. you can also walk around the room with the microphone and see how the response moves around. I'd guess that if you just moved back and forth in the room along the centre line you'd see a nice bump appear and disappear over 70Hz in the analyser.

[Macca]

Not surprising that a flat response sounds thin in such a big room. As above in theory the room will still have some gain at certain bas frequencies but I have found in practice that once the room gets beyond a certain volume the subjective effect of this gain pretty much disappears unless you have the SPL very high. At normal volume you won't get room gain at all. I say this from having use a lot of different kit in two different, but large rooms over the years.

Only guessing here but I would try a response curve that is starts off high in the bass and runs downward to the treble, a sort of ski-slope profile. That should bolster the bottom end whilst keeping the top in check.

[johnpl]

This Behringer device having an autoq function blasts the room with pink noise giving you the imprint of the room.If you want you can set a target curve (in this case a flat curve) and let the device do the equalization. In my case produsing a "sterilized" sound. The curve i am curently using is a "ski slope" from around 500 hz to 15000 hz after tweaking the original curve produced by the device.This is done by ear confirming perhaps Martin's guess. I do agree about your findings in large rooms.The thing is a bit more complicated for me since i am running a center channel between the big horns.
Thank you for your help

[User211]

I've got my response gradually sloping down as already mentioned. Try DIRAC's free demo s/w and they recommend the same thing. But the idea is you tailor the response curve to your liking.

My HF roll off is engineered into the x-over by ear by trying different inductor values.

The response I have works extremely well to my ears. If it doesn't sound right use ear over test kit or you will just end up unhappy.

[Jazid]

I read this to try to understand reasons why straight EQ doesn't work as one might hope. I found it very helpful, it explains why many solutions fall short, though of course it does not quite deliver the silver bullet either!

http://www.minidsp.com/products/dirac-series

At the page bottom there is a download link for the 'On room correction and equalization of sound systems' pdf.

[johnpl]

An excerpt from DIRAC manual says it all and gives me a clear answer for my findings. This is what other guys suggested in this thread. Which means that to fix this "sterile sound" you can manually roll off HF or set a ski slope target curve and let the machine do the work for you.

"The target curve is the desired measured response of loudspeakers in a room In contrast to measurements made of a loudspeaker during its design under anechoic (measured in free space) conditions. While high-quality loudspeakers are usually designed for a flat on-axis anechoic response, these same speakers when placed into a listening room will tend to have a downward-sloping or “tilting ”response at high frequencies ,due to the effects of limited dispersion at high frequencies and greater acoustic absorption. A completely flat in-room response is therefore usually not desirable and will tend to sound thin or bright. Start with a target curve that follows the natural behaviour of your speakers in your room, and then experiment with greater or lesser degrees of tilt in the treble region to obtain the most natural timbral balance|