I might be trying the Nuforce processor as well. I am just curious. I am liking the outlaw better everyday. Sometimes I think it might get a little bright but it never hurts my ears. Usually when it is bright it hurts your ears. I think the outlaw has a up front soundstage with alot of detail, but not necessarily bright. To put things into perspective I owned so called warm equipment(Mcintosh) and it had the same characteristics as my 990. That is what speakers will do because they were Klipsch and now I have M&K.

I like my 990 for home theater (I have a separate system for stereo) but do find it a little on the bright side - and my speakers are anything but bright. What I did was turn "Theater Compensation" on which helped a lot.

Will

I am now using the 7.1 direct with my HD dvd player and it sounds better than anything I have ever used. Great stuff.

Sanjay,

Long time, no read.

Help me out here...exactly what does the Meridian Room Correction do that any digital PEQ doesn't do?

Never mind how one arrives at what they intend to do to the input signal, exactly what does MRC do with notch filters to the input signal that differs from employing any other PEQ?

Bosso

Bosso,

The digital PEQ itself is simply the device/tool used to apply filters to the signal. Automated room correction systems are more than a parametric equalizer. They have to measure what the room is adding to the overall sound and then decide which problems to address, how to deal with those with and then calculate the filters needed to ameliorate the problems.

Dismissing "how one arrives at what they intended to do to the input signal" and asking only about the notch filters is akin to asking what Rodin did with a chisel that other sculptors didn't. He hit it with a hammer, just like others did, and he chiped away at marble, just like other sculptors did. But that doesn't explain why he ended up sculpting something as enduring as The Thinker while his contemporaries didn't. The same tool will give different results depending on what you do with it.

To that end, MRC is different from most automated room correction systems (many of which also use a digital PEQ) in how it measures and what it chooses to correct. As mentioned in my post you quoted, most EQ systems measure amplitude and try to flatten out frequency response. By comparison, MRC measures in the time domain, looking specifically for frequencies that have long decay times, dialing them down to the average RT60 for the frequency range it is trying to correct.

So the difference between auto-EQ systems is not how they use their filters but what problems they use their filters for. For example: if your speaker had an amplitude peak at 55Hz, most auto-EQ systems (like the SMS-1) will use their filters try to bring it down so that it is flat. MRC won't touch it. Meridian explicitly states that flat frequency response is not the goal of MRC.

For a more detailed explanation (from the horse's mouth), check out Meridian's white paper on room correction here .

The paper references an AES presentation (The Loudspeaker-Room Interface - Controlling Excitation of Room Modes, Rhonda J. Wilson, Michael D. Capp, and J. Robert Stuart, Meridian Audio, AES 23rd Int'l Conference, Copenhagen, 2003) described here . This research was the basis for Meridian Room Correction.

If you find the description intriguing, you can purchase the paper from AES.

Sanjay,

Yes, I'm familiar with the paper you linked to, and have read it when it was originally published. I must tell you that it reads much more like a marketing blurb than a white paper. Let's examine some key points...

1. The paper attempts to mislead by first mentioning that 'room equalisation' is accomplished by using a pink noise source. This is so that the impulse response used to generate the same FR graph can be differentiated as unique to MRC later on in the paper, IMO.

2. The mention of some pro sound people using a graphic equalizer to create an inverse curve in the input signal is simply the same ploy, as no one uses a graphic EQ to equalize a HT and most savvy folks know that PEQ can't 'fix' a room.

3. They say that inverse curve equalization is wrong because it:
a. 'aims to arrive at a flat response which will suck the life out of a room.'
b. 'a graphic EQ [an analog EQ] produces all kinds of undesirable effects...ringing and phase errors.' Still, they correctly mention here that 'DSP-modelled EQ [which everyone uses] can be designed to avoid these problems.'
c. And that using an analog graphic EQ makes it 'problematical to try to perfect a room's impulse response'.

4. Next, they mention that to cancel out a 10dB notch [in FR] requires 10 times the power, causing one to run out of amplifier power. This sort of info is certainly for the uninitiated.

5. They then admit that 'Indeed, EQ like this [inverse curve PEQ] only really works for one listening position.'

So, first they spend 2 pages telling us why room equalization doesn't work, then they casually admit that it works for the primary LP.

6. They go on to say that looking at room modes is a better approach. Well, if it ignores a 'peak at 50Hz', whether it's just a poorly designed subwoofer or a boundary gain issue, I wonder how it can be thought of as better?

7. We now discuss that MRC only concerns itself with low frequencies, again PEQ 101 for those who employ one.

8. Next, under 'Filtering and Resonance', we finally get to the fact that MRC is a parametric equalizer that identifies peaks and flattens them to a less peaky state:
'So the aim of our room correction system is to identify important low frequency room modes-and their Q- and construct notch filters to reduce excessive decay times due to resonance at those frequencies to something more like the overall decay time of the room.'

9. Now we read that the range is 15Hz-250Hz. What is the range of the SMS-1?

To keep this shorter than it will otherwise be, I'll just say that MRC is nothing more than a PEQ that uses measurements from multiple spots in the HT, then averages them into a single FR graph, which it smoothes and then fashions PEQ notch filters to attack the peaks from 15-250 Hz.

It doesn't address placement (room and boundary gain induced peaks), delay settings and variable phase control of the sub to realign phase (which also can correct peaks and dips), and it describes that it's designed to treat the average sized HT and to treat it such as to reduce it to below the average reverberation time.

As they admit in the paper: "Here it is important to understand that there is a 1 to 1 relationship between frequency domain and time domain, i.e. time domain does not include any additional information not already present in frequency domain."

They offer impulse response, frequency response and waterfall graphs, all of which are derived from the impulse response (clicks or chirps).

It's redeeming quality, IMHO is that it treats all channels, but 'white papers' that contain words like 'astonishing level of DSP capability' leave me wondering if I'm being educated or sold.

It seems to be the latest gadget for the upscale owner who typically throws the manual away and wants to push a button for setup, IMO.

Certainly the same tool gives different results depending on what you do with it, but, as your own illustration points out, a capable sculptor like Rodin will create a better result than the same tool driven by a computer program.

Bosso

A wee point of fact irrelevant to this discussion and to audio but I'm too picky to let it go by: Sdurani and Bossobass have both mentioned Rodin's work with a chisel. Rodin actually didn't use a chisel. Hr worked in clay, believe it or not - see Wikipedia. After he finished a figure it was then cast in bronze or chiseled in marble, not by him but by his large corps of assistants. Sdurani, you should have said Michelangelo.

While we're coming up with nitpicky art remarks (and while the debate moves beyong my experience with EQ), I'll add my own.
I read this and was reminded of a brass box that dad picked up while at a metalsmithing event in Carbondale a month or so ago - milled out by a CNC machine, but using instructions created by Richard Maudsley. Maudsley's always been know for deeply intricate, detailed designs (often using thousands of tiny pieces of tubing that he would draw through custom-made draw plates to create unique shapes, cut into lengths, and solder together). I ought to get a picture of that puppy for you guys. If Rodin was giving the computer program its instructions, the program would become the tool. (Granted, I'm twisting bosso's point all into knots for no good reason - the statement is quite true if there isn't that talent somewhere upstream - but it was funny to see the idea right after seeing Maudsley's CNC brass box.)

Hey, this isn't audio but it's fun! Merry Christmas, everyone.

As for the subject at hand, I think that sdurani and bossobass (both pretty darn knowledgeable fellows) have hit on an underlying aspect of room equalization: there is no one absolute, agreed-upon answer. We know that our rooms do things to the sound that alter what our ears hear, but how to compensate for it is a tough nut to crack. All the more reason to kick the tires on the more common approaches to see what turns up...