ECM8000 accurate enough for correction? ...and more

D

deafen

New member
I'm trying to correct the frequency response of my monitors/amp/control room. The method that I'm using is to play back white noise (stereo, uncorrelated, to avoid phase/comb filter problems), recording it through an ECM8000, and then doing spectral analysis in SoundForge. I'm applying the correction by strapping a parametric EQ on the master outputs of Cubase during mixing, then removing it when rendering a final mix.

Two questions:

1. Is the ECM8000 accurate enough to be used this way? I know it's what it was designed for, so I sure hope so.

2. Does the rest of this procedure sound like a smart thing to do? I've been able to flatten out my NS10M Studios to +/- 2dB from 65 Hz - 20 kHz.

On a related note, but a bit OT for the microphone forum: the response curve for the NS10M Studios in my control room had a *huge* peak (+11 dB) around 200 Hz. Does that sound right?
 
i always though one adjusted the room around the monitors to "flatten it out".......i dont think its very kosher to have an EQ in the chain to mix with......
 
Well, as you might have surmised by my choice of monitors, I'm out of money, so further room treatment just isn't an option. The room actually doesn't have much to do with it, since I'm on the long wall of a 14x20' space with a 10' chopped a-frame ceiling that comes down to about 4' off the floor.

Okay, then, I'll bite. Why wouldn't it be okay to have an EQ in the chain? I'm using a relatively quality EQ plugin (Steinberg's FreeFilter from the Mastering Edition pack), and I can get the measured response down to +/- 2dB. If I can get a more accurate frequency response, what could be wrong with it? (Yes, I know that a lot of EQ's cause phase shifting problems. I'm not hearing any with this setup, and like I said, the final mix rendering bypasses the EQ completely anyways.)

And perhaps a more important question: between doing this and doing nothing, which would be the better thing to do? Those are the two options I've got...
 
I think the best thing to do is to treat the room. That large spike at 200Hz that you have is most likely a room mode from on of your room's parallel walls. I think a decent way to fix the problem is to make some good bass traps. You can search the web for something like "build bass trap" or whatever and get a lot of good info. Like someguys made what looks to be a decent trap out of a trash can and insulation. I'm kinda goin' with a similar thing as far as bad room modes, but I would avoid EQ at all costs. Like Gidge said it just isn't "kosher".

Ian

Oh, and your final mix bypasses the EQ anyway? I'm kinda tired but that doesn't seem right. Any changes you make during the mix will be relative to the EQ you have applied and your mix will sound different without that EQ. But I could be wrong.
 
Well, good news on that 200 Hz peak. I moved the monitors (and the desk they're on) about 2 ft. further out from the wall and it disappeared. That left the measured response, uncorrected, measured by the ECM8000, at +/- 5 dB from around 50 Hz to 22 kHz. Not stellar, but certainly better than I expected from the monitors that seem to be everyone's favorite punching bag.

I was able to do another correction profile and get it to within +/- 2 dB again, but I don't know how necessary it is anymore. Especially since no one here seems to know if the ECM8000 is accurate enough to use for this purpose. Harvey? Please?

The reason to bypass the EQ at final mixdown is because the EQ is to correct for the speakers, not to be creatively applied to the mix. The final mix doesn't go to the speakers, it goes to disk, so it doesn't need the correction profile. Make sense?

And not that I don't believe you that it's not kosher, but why can't anyone seem to tell me why it's not okay to do? Do you have anything I could read or reference to find out why it's a bad thing?
 
Hmmmm?

You are sure you are not correcting for the room?

To me If you are in the room I think you are it would make the bass stand out.

If I understand right the roof slopes down tward the wall. and you are set up facing that wall? That seems like it would be a problem to me. Can you move to the end of the room to a flat wall and still keep your monitors out of the corners? Can you move your setup further from the wall? You might try hanging dense cloth (sleeping bags) from the ceiling to flaten the surface and get rid of some of the echo chamber low end amplification?

If you have enough speaker and mic cable try moving them around in different proximity to the walls (as a unit) and see what happens. You could move them so that they are facing the length of the room? This would give you an idea how much of the bass boost is the room VS speakers.

If you are saying you drop the EQ when you export the file I can understand your school of thought. If you are correcting for a room or speakers and then you remove the sound from that enviroment do you still need to correct for them? AHHH but every speaker colors sound in the real world. You are still better to learn your DAW as compared to your car stereo, boom box, home stereo ect, and adapt. If I where i your posistion I would use an external stereo EQ in route to the speakers, and it would always be there and always the same (tape over the sliders), or I would do nothing at all and learn to adapt.


Oh Ya. According to Behringer the ECM8000 are very flat. Opinions could varey though. I do not own any but would like to for the money.

F.S.
 
In the 60s and 70s, we used eq to tune the rooms, but we eventually realized that the problems were in the time domain, not the frequency domain. Think of blowing across a Coke bottle and producing a note. That's time domain. The air resonates and produces a note. Like a pipe organ, it takes time for that note to build up, and time for it to die away.

The resonant nodes in a room act the same way, slowly building up some frequencies and cancelling others. Yes, you can increase or decrease the level of those notes with eq, but the time buildup is untouched and you get "smear", you lose detail, and the music begin to sound "muddy".

These days, most studios use bass traps and a combination of absorbtion, diffusion, and deflection to get the room sounding relatively smooth to begin with. An eq can help tame the sound a little bit in the problem areas (around 50Hz to 400Hz) but the problem isn't going to disappear, and the eq will only be effective at ONE single point in the room.

There's an article in one of the new audio magazines on tuning a room using a parametric eq and the article recommends only trying to compensate for 3 or 4 node problems at the most. Any more and you create more problems than you cure. Try to cut, not boost, and you only want to cut about 6 to 8dB at the most. A 31 band graphic would be much better than a 15 band graphic, and a 4 band parametric would be better than a graphic equalizer.

I suggest you search the net and do some reading about how to properly tune a room. It doesn't have to cost a lot, but you really have to understand how tuning works before you try it. EQ is good for a lot of things, but tuning a room for flat response isn't one of them. In lieu of tuning the room, you can also try mixing (softly) on near-field monitors which don't excite room resonances as much as cranked, large, soffit-mounted speakers.

The big problem with most small nearfield monitors is a lack of accurate bass information, and learning how to compensate for what you're hearing from them.
 
GOOOD Info.

I always watch for your posts Harvey.

I guess It is sort of how a RTA only sets up one part of the room for live applications.?.? But then again much different.

As far as live is concerned I have never been able to, or seen any sound man be able to walk into a room with the board pre set up and tune to the room with the main EQ and call it good (and have a good sound). So I know it is not the all mighty answer.


Later

F.S.
 
I think quite a few couldn't understand about RTA, that's why they kept being insistent of room treatment rather than using a GE. I guess the battle is 31GE vs Psychoacoustics, since many of us are really on a tight budget.
I am facing this problem too as my room is really small, but loaded with too much gear. I am recording drums as well and my control room and my drum room is the same room, and it is also for recording vocals. I don't have much choice here as it's the only room i have for studio work. Most people flatten the room for their control listening purpose, but for recording drums and vocals, it's another story, we are talking coloration here, not flattening.
I believe many home studio owners have difficulties in regards to lag of space, and treating their rooms with huge bass traps, diffusers and acoustic foams ain't really too much an option. That's why there lies the need for a 31 band Graphic EQ to compensate for deficiencies.
On typical hifis, they normally come with a graphic eq to shape the music according to the venue. And now with regards to RTA, it's also about shaping the music to fit with the venue.
That's the right way it suppose to be, pumping pink noise or sine wave from keyboard and recording them with the Behringer ECM8000, then checking the signal dips and peaks with the software.
Now if Harvey say the problem is time dependent, then that surely is as we are dealing with flutter echos and such (reverberation issues!). The idea of RTA is for that one location we are looking for, that is our monitoring sitting position. We are not tuning the room with the 31GE but rather tuning our monitoring system to work with the room according to how our ear will hear. Like if our bass frequencies are running out of our monitoring positions due to our small rooms, and so we are hearing less bass then we should, we boost up our monitoring system with the 31GE so that we hear more of it, this will bring less possibilities of over boosting the bass while we are mixing, due to not being able to hear enuff. There is only a need to tune the 31 GE once and leave it running. Normally an analog hardware type is preferred and placed in btw the monitors and the audio out.
I know frequency reflections are working on the time domain, but i believe on the average basis it also works on the frequency domain, meaning some frequencies will tend to lag in this room while some tend to peak more, even when we move to different positions. It will be great to place some bass traps, diffusers and acoustic foams i believe, analysed after RTA on the location. Now i am currently looking at acoustic screens that are removable...
 
Harvey and the rest are dead on about correcting the room.

But to answer the question about the ECM8000... Yes it is most certainly flat and unbiased. I compared my ECM 8000 to an Audio Control RTA mic, and they were nearly spot-on with eachother.

Bass traps can be quite cheap if you build them yourself.

Look for Owens Corning 703, and cover it in black burlap with a 2x4 frame and mount it a 6"-12" off the wall, corners and/or cieling. Then magically watch that 200Hz peak tame itself.

Or for a few bucks more get Realtraps and be done with it. Plus sorucing 703 can be a PITA depending on where you live.
 
Perhaps I am missing something, but wouldn't pink noise be a more appropraite measurement tool? White noise is flat, but pink will let you actually hear what is going on. White noise keeps you from hearing the guy in the cubicle next to you in a call center. ;)
 
You use mic. calibration files with your mic / measuring program combination
so you end up with flat curve.
Matti
 
"Is it flat enough ?"
You never really know abou these things unless you have your equipment calibrated in an anechoic chamber.

I have done similar RTA analysis/compensation with calibrated mikes and monitors. It does make a difference and EQ'ing the room does help, but as pointed out before, there are also other ways to treat a room. I think they are all valid.

IMO the ECM8000 is probably flat enough that other errors and factors will far outweigh its inaccuracy. For one thing, many room resonances are very high Q, so even it=f you identify them, you need something much sharper than a standard EQ to "correct" them. Some "feedback eliminators" actually have very sharp parametric equalizers that are better for room correction than an "EQ".

It is a huge topic with lots of pitfalls and myths and lore. :)
 
I prefer to do a frequency sweep because it gives you much more detail as to what is really going on in the room and with your speakers.
 
bubbagump said:
Perhaps I am missing something, but wouldn't pink noise be a more appropraite measurement tool? White noise is flat, but pink will let you actually hear what is going on. White noise keeps you from hearing the guy in the cubicle next to you in a call center. ;)
Click to expand...

OK, I'll look stupid. Wouldn't be the first time. How does THAT work?

I mean the white noise thing, not the pink noise.
 
White noise is equal power across the sound spectrum. Pick noise takes white noise and then decreases by 3db/oct from low to high across the spectrum. Gnerally when usin gan RTA one uses pink noise as it is much easier to hear what is going on versus white noise which pretty much sounds like the same sort of PSHHHHHHHH to our ears.
 
Harvey Gerst said:
In the 60s and 70s, we used eq to tune the rooms, but we eventually realized that the problems were in the time domain, not the frequency domain. Think of blowing across a Coke bottle and producing a note. That's time domain. The air resonates and produces a note. Like a pipe organ, it takes time for that note to build up, and time for it to die away.

The resonant nodes in a room act the same way, slowly building up some frequencies and cancelling others. Yes, you can increase or decrease the level of those notes with eq, but the time buildup is untouched and you get "smear", you lose detail, and the music begin to sound "muddy".

These days, most studios use bass traps and a combination of absorbtion, diffusion, and deflection to get the room sounding relatively smooth to begin with. An eq can help tame the sound a little bit in the problem areas (around 50Hz to 400Hz) but the problem isn't going to disappear, and the eq will only be effective at ONE single point in the room.

There's an article in one of the new audio magazines on tuning a room using a parametric eq and the article recommends only trying to compensate for 3 or 4 node problems at the most. Any more and you create more problems than you cure. Try to cut, not boost, and you only want to cut about 6 to 8dB at the most. A 31 band graphic would be much better than a 15 band graphic, and a 4 band parametric would be better than a graphic equalizer.

I suggest you search the net and do some reading about how to properly tune a room. It doesn't have to cost a lot, but you really have to understand how tuning works before you try it. EQ is good for a lot of things, but tuning a room for flat response isn't one of them. In lieu of tuning the room, you can also try mixing (softly) on near-field monitors which don't excite room resonances as much as cranked, large, soffit-mounted speakers.

The big problem with most small nearfield monitors is a lack of accurate bass information, and learning how to compensate for what you're hearing from them.
Click to expand...

YES! "Acoustic" problems, "acoustic" solutions, no?

Ciro
 
You must log in or register to reply here.
Back
Top