Tune your system the way live pro sound people do for cheap

[john65b]

OK, so after reading this thread, I picked up a Behringer DEQ2496 and matching 8000 mic. I liked that I could feed it a digital optical signal and do all the EQ'ing in the digital domain, then output the EQ'ed signal to my DAC. All the horrible things I remember hearing about EQ'ing was due to it being done in the noisy Analog domain...

Anyway, I initially saw that "how to set up a DEQ2496" Youtube vid (the one with the guy with an Aussie accent). I have an Aikido Tube preamp (6DJ8 / 12BH7), a Krell KSA50 Clone, and a Cirrus Logic CS4398 based DAC (all of which I built myself with the help of friends from DIYAudio). My speakers are Apogee Duetta Signatures (restored by me too - no "Apogee buzz" anymore!!)

Anyway, I initially did the Auto EQ low bands only (20hz to 400hz) as the vid shows. Big improvement on the bass (these Apogees are impressive at Bass). For the hell of it, I let it do Auto EQ the whole band from 20hz to 20khz, just to see the response. Once finished, I then modified it a bit to slope the high frequency down toward the 20khz. I was impressed. I liked it enough to keep it permanent "Reference" setup.

Now I obtained a pair of Acoustat Spectra 22 locally that needed a little help. They needed a bit of work on the bias circuit (bias voltage was way low) and panels needed to be cleaned up and a few bias wires needed to be reglued to the grid, causing some annoying buzzes and noises...anyway, got them all dialed in and perfect. For teh hell of it, I then Auto EQ'ed the entire band. I then modified the HF to trail down toward the 20khz just like I did on the Apogees. They sounded great afterwards.

But the biggest surprise was that these cheap (but great ESL BTW) Acoustat Spectras now sound very very close to the Apogees. I was floored. Without the EQ, they sound nothing alike. I am convinced I can take any speaker (that can do the frequency range) and make it sound close to the Apogees with this equalizer. Kinda had a "Bob Carver" moment...

 

[canuckaudiog]

I too have been playing around with EQing.

I am using my iPhone, a $10 calibrated RTA app, J River's built in equalizer (using the graphic EQ for now, going to move to the parametric when I understand it better), and pink noise.

So far the results are rather promising. I had to take out all the room treatments recently to get ready to move to the new house, and this has brought back that sweet sound I was missing. Not entirely, I think room treatments are ultimately a better solution (possibly doing that and a combination of EQ), but definite improvement. I've been using the A-weighted SPL option which as I understand takes into account the ears natural hearing, which to me makes sense because the frequencies I am boosting and cutting coincide with what I expect a loudness function to do.

In short, this little experiment has turned out for the best! I am quite happy with the results.

 

[totem]

Once you start down this road the results can be astoundingly impressive!

 

[kevzep]

I've been using the A-weighted SPL option which as I understand takes into account the ears natural hearing, which to me makes sense because the frequencies I am boosting and cutting coincide with what I expect a loudness function to do.

When you are measuring your system, you need to use no weighting. This way you get the true representation of how your system is performing.

A weighting explained below.....

The curves were originally defined for use at different average sound levels, but A-weighting, though originally intended only for the measurement of low-level sounds (around 40 phon), is now commonly used for the measurement of environmental noise and industrial noise, as well as when assessing potential hearing damage and other noise health effects at all sound levels; indeed, the use of A-frequency-weighting is now mandated for all these measurements, although it is badly suited for these purposes, being only applicable to low levels so that it tends to devalue the effects of low frequency noise in particular.

 

[canuckaudiog]

When you are measuring your system, you need to use no weighting. This way you get the true representation of how your system is performing.

A weighting explained below.....

Well, I think the A-weighted option makes more sense because my typical listening volume resides somewhere between 65-70dB.

I think if I was listening at a higher volume, such as 80dB, using no weight would make more sense. Or am I off my kilter here? I just found I made better adjustments when using the A-weighted SPL calibration versus flat.

 

[kevzep]

Well, I think the A-weighted option makes more sense because my typical listening volume resides somewhere between 65-70dB.

I think if I was listening at a higher volume, such as 80dB, using no weight would make more sense. Or am I off my kilter here? I just found I made better adjustments when using the A-weighted SPL calibration versus flat.

A weighting is used when sound levels are measured in relation to human hearing.
For example, max SPL at the boundary of a concert site to comply with local council noise limits.
Or industrial noise levels for machinery like chainsaws, lawnmowers, etc etc.

To measure what a sound system is doing, you need to use no weighting otherwise you are not getting the actual representation. When we measure a system we want to know the actual performance.

The human hearing curve changes with SPL and this needs to be also taken into consideration.

But by all means measure the way you want, I am just telling you we do it in the world of sound system measurement.
We never use weighting when analyzing a system. The target curve when using unweighted is not a straight line.
The term flat is a figurative term, it's not literal....

 

[canuckaudiog]

Makes sense.

Well I am going to play around with it more, learning to adjust the EQ is really challenging, I am going to have to do quite a bit of listening to make this a worthwhile venture.

So far so good though.

 

[kevzep]

Makes sense.

Well I am going to play around with it more, learning to adjust the EQ is really challenging, I am going to have to do quite a bit of listening to make this a worthwhile venture.

So far so good though.

Exactly, experimentation and listening is the key. You need to trust your ears, they are the most sensitive and accurate tools you have....

What helps a lot is when you make you adjustments and you feel you are happy, go back and by-pass the EQ and compare it to what you have done, you will very quickly realize if you have gone over the top....
Its easy when EQ'ing a system to go too far and dig in too deep, you always need that nominal point to return to.....

 

[canuckaudiog]

Exactly, experimentation and listening is the key. You need to trust your ears, they are the most sensitive and accurate tools you have....

What helps a lot is when you make you adjustments and you feel you are happy, go back and by-pass the EQ and compare it to what you have done, you will very quickly realize if you have gone over the top....
Its easy when EQ'ing a system to go too far and dig in too deep, you always need that nominal point to return to.....

Thankfully with J River it's very easy to do that.

The troublesome part is focusing your hearing and trying to realize what sounds tonally correct while factoring in the recordings. It's at this point where having high quality recordings is crucial! Even familiar recordings may not work well here because I haven't heard them live. Or of course the best reference would be pink noise and knowing what it should sound like properly tuned, which I don't.

I think ultimately though I am going to gain an indepth understanding of the different frequencies in the sound bandwidth and have a greater appreciation for it. It's also teaching me a lot about room acoustics and how it affects the sound. For example right now the 1Khz and 2Khz bands are highly emphasized in the room, causing the midrange to be far too pronounced and making it sound colored. Back off those two bands, boost the treble a bit and it sounds much clearer and natural. I am doing this in a room with zero room treatments so, the EQ is making quite an improvement.


Sent from my iPhone using Tapatalk

 

[Soldermizer]

Here's a VERY late addition to this thread. First, I'm just an amateur, listen at home to stereo type of guy. Here's a plug for REW (free) and for the Behringer DEQ2496. Although I sold mine because I no longer needed it, consider that this box has been on the market for nearly 20 (!!!) years. To me, that says quite a lot about a "consumer" or "prosumer" product in an industry where models come and go a bit more frequently. If you have the measurement equipment, by all means learn to use it and tackle the problems you need measurements for, such as identifying peaks and reducing them. But there are some issues not even the best equipment in the world will address, but my method will -- because YOU are the only test equipment, and thus it is you who will determine what sounds best for your set-up.

I don't guarantee my approach will address ALL issues, but it is bound to be the cheapest. No test equipment but your own ears. No media except band limited pink noise, available free various places on the internet. The only equipment you'll need is a graphic EQ or some other means to boost (less desirable) or cut (more desirable) specific frequency bands. The general plan is that you listen to different bands of pink noise at close to your accustomed volume level. If one band is noticeably louder than the rest, lower it gradually, perhaps a few db at a go. Your goal is to have most* bands at the same apparent loudness. There are very few things that can go wrong, but here are a few:

*Don't obssess about the very low end (20-100 Hz) and the very high end (over 10KHz). Unless you have a monster subwoofer system, you probably won't be able to match the apparent levels you hear in the midrange. Similarly, unless you're a bat or a dog, your hearing falls off somewhere above 10KHz. Strive for equal apparent loudness between 100 Hz to 10 KHz. Even a clock radio can meet these extremes.

Don't try to do it all at once. You'll need to do dozens of comparisons over several listening sessions. Take breaks between pink noise sesssions by listening to your typical music but do not fiddle with EQ at these times. This is to refresh your ears, not to hunt down that last 1 dB anomaly.

This method (which I didn't invent) is easy except for the many listening trials you'll do. I did this to my system(s) years ago and I rarely have felt a need to change the frequency response of my system since. Yes, it's that good. And, if you don't like the result, it has cost you nothing but some time and listening.

 

[JoeESP9]

I bit the bullet and bought a calibrated mic in 2011?. It was <$100. It replaced a Rat Shack SLM I bought in the late 1970's. I have both although the Rat Shack is no longer used.

 

[FOH Engineer]

As a veteran of more sound gigs than I can count I can impart some useful advice. Number one for me is forget the numbers , forget the rules. Get your rig or seek out a rig to produce the human voice accurately and cleanly with maximum headroom and no exaggerated frequencies anywhere in the bandwidth used.
Next and this is a tough one, get the rig to also reproduce a piano with no stridency in the octaves above middle C.
Lastly find a great recording of a solo double bass and and tweek the low end until it sounds natural with no rumble or over overexcursion of the low end drivers
If you can accomplish those three things your system will sound correct under just about any circumstance. Forget measuring mics and spectrum analysis, blah, blah. If your rig does the above you have it right. There are no exact numbers or ideal settings. As long as there is no clipping or harsh room reflections you'll have it right. In live sound the most difficult instrument to reproduce is a human voice, it's not even close. I'm not talking about saying check over and over. I'm talking about getting a natural sounding vocal over a dense band mix, in a live room, at high volume levels. It's the difference between a below par live mix and excellent sound throughout the room .

 

[PanheadXR290]

Cool tip - thanks! :thmbsp:

Another possibility is to pick up one of the old Soundcraftsmen EQ/Analyzer units, which have the noise generator and everything else you need built-in, except the microphone, which came with them as an accessory. Get one with the mic, or find a compatible mic, follow the instructions (which you can find online, if needed) and you're set.

Some audiophile purists don't want to add the extra component in the signal path, but in many cases of less-than-perfect real world listening environments, using one of these is probably a better choice than not using one. Pros don't do these things just for the heck of it, nor just for the fun of it.

Youd get off cheaper & have much better results by simply buying a brand new DBX eq that is also a 2 to 3 way active crossover as well as loaded with the latest room correction software with pink & white noise generators ,i paid $450 for my DBX brand new,it replaced the Mcintosh MEN 220 Room correction unit and the big $ crossover i was using to triamp with.

The DBX PA-2 is the best upgrade ive done to both my rigs and cost next to nothing .