"Official" Audyssey thread (FAQ in post #1) - Page 1712

At what point does it stop EQing down low, regards to MultiEQ. Would it be the knee -3db point....

Sorry, just read your post, no eq below the knee...thx

... but Audyssey normalizes the level which indirectly boosts below the -3dB point, see the pic above which shows approx a +9dB level increase at 2Hz.
So, a manually applied HPF on the subwoofer after Audyssey might be a good idea...

MultEQ searches for the -3 dB point as measured in the room and then stops correcting below that point but the normalization goes down all the way.

If I ran xt32 with dual subs (similar characteristics other than one having more output) and wanted the more powerful sub to contribute more than the other, could I simply adjust the gain on the subs after Audyssey does it's thing?

I did that without audible problems.

Agreed.

Got it. Thanks for the clarification.

Acoustical Treatments Master Thread

I would suggest that you buy a calibrated mic, mic stand with a boom arm and outfit a laptop with Room EQ Wizard (REW). As the mics require phantom power and laptops' onboard audio is rarely duplex (capable of simultaneously outputting a test signal and inputting the mic's signal), you will need a USB pre-amp with phantom power.

The mic will cost ~$85, the pre-amp ~$110, the stand/boom arm ~$50 (that's a guess) and REW is free.

I use the Dayton EMM-6 microphone (from Cross-Spectrum) and Tascam US-122mkII USB Audio Interface.

If you have any/all of this stuff, then good for you. If not, one of the first things they will ask you on the thread above is what are you measuring in your room now. Be advised.

Jeff

+1. Once you adjust the frequency response of the sub it might sound louder or quieter than uncorrected. Normalization simply turns the sub level up or down to theoretically make it sound about the same with Audyssey on or off, in terms of perceived volume.

IIRC normalization for Audyssey focuses on a specific frequency or narrow band of frequencies comparing before and after EQ in that band of frequencies. I think Chris K identified the frequency or range here but I can't recall it.

Somebody with a better memory will correct me if I misremember.

He never disclosed the frequency range for the sub but said:
we prefer not to give out that detail...It comes from an experiment using listeners to estimate the relative levels of subwoofers by ear.
But I am guessing it is 40-80Hz.
And he said:
The speaker levels are matched to C-weighted levels. That covers most of the response with a certain standardized weighting.
So I assume that narrows it down to 500Hz-2000Hz for the non-sub speakers.
And more quotes:
The right way to match levels to the satellites is with a spectrum analyzer.

MultEQ uses a spectrum method that looks at the energy in the subwoofer band and compares it to the energy in the midrange of the satellites.

Since the entire frequency spectrum per channel is not considered when normalizing the levels, the before and after Audyssey average SPL levels will not measure precisely equal, but nevertheless they will sound about the same level.

FYI: there were abnormalities found with bass processing in your model as well, fixable with a FW upfdate. This post in the model thread refers to it. jdsmoothie and batpig are the experts but there are several informative posts on that thread, including the one linked, for diagnosing the problem. If you don't have that problem in the the AVR, erratic results can sometimes be fixed with a "reset the microprocessor" procedure. BUT, if everything appears to be working properly now, sounds good and measures reasonably, I'd be very tempted to just leave well enough alone.

Following the setup Guide and using the tripod properly (insulate the feet from vibrations, as previously mentioned)-or better yet a mic stand and boom arm-has helped many folks get better and more consistent calibrations.

Interesting link. Note that the Audyssey mic used by the OP for REW without a cal file on that thread was for the 5308, so was likely not a DM-A409 like you have. But your point is well taken, it wouldn't matter if your only purpose is to compare Audyssey On/Off. And measuring the preout signal directly makes the whole issue moot.

Not true. The PBK is a subset of Anthem RC which is, in almost every way, comparable to Audyssey.

Yes, I knew about this, the firmware of my receiver was updated in March 2010. It was a very obvious bug, don't know how it slipped through QA. I work as a software developer, so I know what trouble a lost or reversed sign of a numeric value can cause. The subwoofer correction was applied to the output of the Dolby and DTS decoding stages with reversed values, resulting more gain applied to frequencies instead of the supposed cutting and vice versa. With PCM input the bug was not present.

Allthough Audyssey uses a spectrum approach, which only looks at a certain frequency range within the sub region,
for setting the subwoofer level, it could be susceptible to standing waves since the level is only measured from the first position.
And that first position *could* be located in a dip or a peak, right?
A better approach would be to do a 2nd pass sweep and calculate the levels *after* the filters have
been created and hopefully gotten rid of the worst peaks and dips.

Quote:
Originally Posted by hclarkx View Post
This is very true. And, in addition, Audyssey does "phase" or "time" correction that virtually every other EQ system does not do. Hence you get a parametric EQ from the PBK which can't hurt and on top of that you get some additional amplitude (parametric) EQ AND phase/time correction from Audyssey.


Trinnov EQ in the Sherwood R-972 corrects phase as well.

Hi guys,

I moved some furniture around my HT room and in the process of doing the 8-position calibration (Onkyo TX-NR3009 w/ XT32), I think I may have damaged the Audyssey mic.

Here is what happened: When moving the mic around the central seating area I raised the tripod (that the mic was mounted to) above my head and accidentally hit the tip of the mic into the ceiling above my head. This was such a stupid thing to do but its no use crying about it any more.

After trying real hard not to jump out the window for a few minutes, I found myself asking 2 questions:

(1) How can I tell if I damaged the Audyssey mic?
(2) Can these mics be replaced and by whom? (Audyssey, Onkyo, or 3rd party)

I do understand the calibration of the mic depends on the specific receiver that it came with but I would hate to ship my receiver to a far away place for re-calibration.

Thank you for any tips you can give me.

Actually, the mic calibration is not specific to your receiver, which makes ordering a replacement fairly straight-forward. Make your request for a replacement here: http://www.us.onkyo.com/feedback_form.cfm.

As to how to tell whether it's damaged, you could try running a new calibration.

^The mic is not specific to your processor. There is a standard Audyssey mic cal file in your processor that will work equally well with any Onkyo Audyssey setup mic of the same mic model #. Worst case scenario, just order a new mic from Onkyo.

You probably did not damage it if you only lightly bumped it into the ceiling, but the only way to tell for sure would be to measure it using an adequate measuring system against a known good one (see this thread brought to our attention by zzoli). If you purchased it fom a B+M store like BB they might switch it out for you.

My apologies, but need some clarification.

When running audessey, should I have re-eq off, and also turn THX settings off? Or leave them on?

I know this has been discussed both ways. I truly don't know how THX settings work with the calibration..
I would like to use it if THX source material benefits from it.

Thanks all...

I've read through posts, Audysseys website/documentation/faq. I'm having a hard time understanding the real world practical usage for Reference Level Offset as opposed to Dynamic Volume. Does anyone have thoughts on the two settings and how they can be used to achieve ones goal?

Hi lou, first off, RLO and DynVol are two different animals!

RLO: you need to understand that human hearing is frequency dependent and that frequency dependency is level dependent.

DynVol: you need to understand how loud TV commercials were ruining our lives till the advent of Audyssey's DynVol.

Yes, a senior moment for sure. But, I fail to see how PBK can be comparable to Audyssey when it's done in isolation of the mains. I.e., how can PBK align the phase of the subs and the mains in the crossover region? In fact, correcting group delay even partially would seem to aggravate any phase difference at the crossover frequency. Does PBK make an assumption and shoot for a 360 degree lag to bring the sub back into phase with the mains?

It seems this may be apples and oranges .. PBK works on the sub alone while Audyssey MultEQ, like the ARC system, addresses the whole audio spectrum.

I understand Dyn Vol..been using it for years.

I was told that it should be used when you want to reduce the amount of "Dynamic Volume sound effects boost" that occurs at lower volumes.

I was also told that it tricks Audyssey into thinking your listening to the movie at a louder level. So if you Watch movies at -10 Vol, then RLO should be set to -10 to balance things out.

These comments are much different than Audysseys documentation which leaves me baffled.

Whoa! What I said was that ARC is comparable to Audyssey and PBK, as a subset, does for the sub what Audyssey does for the sub. Think of PBK as comparable to Audyssey Sub EQ. Neither of the subwoofer tools can do what the full range, multichannel tools can do in integrating the sub with the other channels.


Right. So, what is wrong with using PBK in a system in which Audyssey attends to the inter-channel integration?

Gomez, all settings in the processor are ignored during autosetup. You will find the Audyssey setup guide very helpful as it addresses quite a bit of Audyssey basic stuff like this.

Reference level offset changes the way that Dynamic EQ works. It causes DEQ to make less change than it otherwise would. Some use RLO for movies because they find that the DEQ corrections are more than they need. But AFAIK RLO was invented to deal with the fact that while we actually know how loud every sound in a movie soundtrack was mixed to be, we do not know that for TV or music mixes. That is, movies are mixed on mixing stages where the systems are calibrated to reference level, but music and TV are not. In practice, music and TV are all over the map but uniformly are recorded at digital levels that would occur if the reference level were set lower. The offsets allow you to adjust the way DEQ affects your content "on the fly" so that bass doesn't become overbearing

Okay. I'm with you now. ARC was not mentioned in the original post or your post. I assumed the discussion was around PBK.

Definitely nothing wrong with stacking them as I think was clear in my original post. As long as all the A/D and D/A converters remain transparent and the delays don't become excessive, I'm all for as much correction as we can get. My aging ears/brain need every bit of help they can get.

This does raise an interesting question though. Audyssey is or at one time was limited to, as I recall, 10 db of correction up and down. There is/was surely a good reason for that that it seems is solved when the correction is applied in two stages (one by Audyssey and one by filters in the sub).

Thank you! By Dyn Eq changes, you mean Dyn Vol corrections?

In my case I have TV set to "Dyn Vol Evening" because of the fluctuation in sound. With Videogames, Dyn Vol = OFF since it throws off FPS games by boosting the surr speakers too much.

In this case, would I benefit from using REO for Video games or Cable?

You need to understand the difference between DV and RLO.

The basic underlying denominator for all Audyssey features is a flat frequency response at THX Reference standards, thus allowing your system (if it's up to the task) to reproduce the audio as the sound mixer intended.

DV basically limits the dynamic swings of the audio, basically compressing the dynamic range in real time. It is tuned towards vocal reproduction so using DV at a main volume below THX reference results in a boost to the vocals to maintain the clarity of speech, but limits how loud explosions etc. will get.

RLO, stands for Reference Level Offset and is there solely to account for the fact that only movies are recorded to THX Reference standards.

RLO in effect is directly tied to DEQ.

DEQ was developed to automatically compensate for the way our hearing differs at different volumes. At lower than THX Reference volumes, human hearing tends to be less sensitive to the lowest and highest octaves. To reproduce a perceptually similar frequency response when the volume is turned down, you need to boost the lows and highs relative to the mids. The further you turn down the main volume, the more you need to boost those octaves.

DEQ is linked to the THX Reference standard so it does nothing when the main volume is set to THX Reference and has increasingly greater effect as the volume is turned down lower.

The RLO adjusts where the Reference level is. In general, music tends to be recorded at different levels than movies are and using the THX Reference standards for DEQ could result in over exaggerated bass and highs, so using the RLO setting compensates for the different recording levels.

For example, if you listen at -10, using an RLO of -10 means DEQ is not performing any compensation. Using an RLO of -5 means it is using a compensation curve for a 5db reduction from the new Reference level.



Max