10 Band EQ on the 976

The filter resolution depends on the processor in the AVR. It is where the filters have to run ultimately and so they will be XT or XT32 depending on which AVR you have.

Why do you use an automated room correction system like Audyssey instead of manual parametric EQ?

A: Actually, we use both.

The SC-1 incorporates active crossovers, parametric EQ that is transparent to the installer, and Audyssey MultEQ XT which must be calibrated using MultEQ Pro software on a PC. (We train our dealers to do the calibration; it is not something the owner has to worry about.)

A few points:

MultEQ Pro has vastly more resolution than any parametric EQ. You would need something like 500 bands of parametric EQ per channel to achieve similar results, assuming you had the time, skill, and instrumentation.
MultEQ Pro optimizes in both the amplitude and the time domains.
MultEQ Pro does have a finite number of target curves, whereas manual EQ can shoot for any given target. It is important to note that our target curves are different than the standard Audyssey curves (since they represent the final "voicing" of our speakers in the room).
It is highly likely that one of our curves will be pretty close to what an experienced installer would try to do. Curve editing (part of the MultEQ Pro software) will likely get the installer the rest of the way (±3 dB from whichever target curve is selected).

There are multiple levels of implementation of MultEQ XT. If you only have experience only with the lowest level, receiver-based implementation, you don't know what you're missing. It is like comparing tone controls to true parametric EQ, saying that they do basically the same thing. They are drastically different in capability.

We find that parametric EQ is excellent for "roughing in" a system; MultEQ's additional resolution works best for the final tuning.

That's not to say that our way is the only way. Far from it. But if you want to get outstanding results quickly, easily, and consistently, it will be hard do better than the SC-1.

Q: Why did you select Audyssey MultEq XT as your room correction technology?

A: Our previous approach to room correction (bi-amping and lots of parametric EQ) delivered great results, but it was extremely time-consuming, and it required a tremendous amount of experience on the part of the person doing the calibration. The hard part is not operating the equipment; the hard part is analyzing the room's many problems and prioritizing them, and coming up with a strategy for tackling them. In fact, one of the most important lessons is learning which problems you ought not to try to fix. (In some cases, the "solution" is worse than the original problem.)

Audyssey's MultEQ XT encapsulates this most-difficult part of the process in software, in effect capturing decades of experience in a program that is installed on a laptop. In the case of the SC-1, we also use the same powerful DSP engine used in the $5000 Audyssey Professional Sound Equalizer to maximize the resolution of the resulting correction filters. The combination is amazingly powerful and effective.

The subjective result is comparable to the calibrations our own people achieved under our old system (in some ways, even better). But these results can be achieved by our dealers (after they go through our mandatory training program), in about an hour. Contrast that with us having to send our own people all over the world, and a calibration process that usually required several days.

Hmm.... Some good info to get me thinking here. Thanks!

Maybe we'll have to revisit this topic once we know more about how the 976 will roll.

IMHO, one of the great disadvantages of a PEQ is that once it is tuned to the masters ears it may sound odd to others. For instance, my 60 year old ears are less sensitive to the 4K-6K band so I may actually boost those frequencies to a point where it sounds weird to my wife who, as all wives do, has catlike hearing. The room corrections which use a microphone to set the processors DSP are at least trying to make the system neutral so that it sound natural to most people. I think that for the most part, higher end equipment tends to shy away from manually adjusted PEQ to discourage users from making their equipment sound different than intended.

When the 976 runs its levels/distances/phase adjustments will it adjust for phase on each sub channel?...
If we had a sub on each side wall I can see where it could be necessary for the phase setting to be different for each cabinet. Could something like this be an issue or is my thinking on this wrong?

I flashed back to all of the people I knew in the 70's with 10 band (full octave) eq's where the sliders always looked like smilies.

You might need to run the set up one time for each separate sub if that's the case.

We are on the same page. I want to see before and after curves to see the problems that might exist then how the correction handles everything.

users should try separate & together. My concern for users is the EQ for one sub may complement or clash w/the other.

For that matter, EQing one channel or sub at a time doesn't allow the software to adjust the channel's output particularly in the lower octaves. You may end up w/a "bassy" sounding system and not understand why.

I think you will need to EQ each channel at a time first, then use the set up mic to adjust for levels.

I would think the levels and time alignment would need to be first.

If you eq to mismatched levels your PEQ settings will be off when you change the levels.

I was thinking more along the lines of adjusting the response of each channel first, before matching all the channel levels/distances.

The 10 band EQ will allow you to adjust the relative level of 10 individual bands of frequencies [much] like a very sophisticated tone control.

Thus you could boost the midrange and decrease the high end if you so choose. This will allow you to customize the sound in your room to your specific tastes.

It will likely depend on where it is in the analog domain. I would think that all analog output will go through REW before the amps. As long as the multichannel input to the 976 is part of that domain all should be well.

If you think about it, distance settings can't be affected by frequency response because you're setting delays for each channel independently....

...So you can do distance settings before or after the EQ step.

It's different when setting levels. As mentioned before, frequency response can affect levels setting: e.g., a large peak (or dip) can throw off a SPL meter, making the channel seem louder (or softer) than it really is, resulting in the level being set incorrectly. Smoothening out the response for each channel first (to whatever extent you can) will subsequently make for more closely matched levels.

You set distance BEFORE messing w/levels or eq. You made clear why in the latter part of your post.

I start w/one channel at a time ...

...then run [test again] all together to check interactions.

Each channel gets independent signals...

...so if you're checking their interaction using the same [test] signal, that doesn't tell you anything about what it [or they] will sound like in actual use.

No. Peaks & dips in the frequency response don't affect delays like they do levels, so distances can be set before or after equalization.

So I could use a bit of an education here. I see a lot of software acronyms and discussion of using them with the 10 band EQ in the 976 in lieu of built-in room correction. Is anyone willing to shed some light on that process and how it works?

For both the 1050 and the 975, I just used my trusty Radio Shack SPL meter and the test signals to set everything up and then adjusted the .1 channel "by hand" as it never really seemed to register properly with a test signal. I just fiddled with the sub level until it sounded good to me and called it a day. That's given me great results in our home so far.

How does a 10 band EQ change this process? What does the software do? Honestly, I'm enough in the dark with this change I'm not entirely even sure what questions to ask but I'm hoping this is a good starting point for a conversation on it.

Do you eq stereo one channel at a time only?

Does REW do this??

Great post EEman.

I do have a question. Does overall phase change with frequency based phase shifts in crossover components?

You also mentioned setting for a flat response. With stereo the common practice is to set a "house curve". Starting at or just below 200hz dial in a gentle slope to +3db at 20hz. Do we not do this any longer?