"Official" Audyssey thread (FAQ in post #1) - Page 1190

Hi

Thanks for the heads up. I already have the Pro Eq kit so no worries on that front.

What you mention about using the 886 is interesting............

I have been wondering how engaging Audyssey (including Dynamic Eq) on the 886 and then also engaging the Audyssey Sound Equalizer would work. I was worried that engaging the Audyssey in the 886 and also the Sound Eq further down the chain would confuse matters and mess things up.

However i can always go back to plan A anyway as you said.

Plan C would be to pay the licence fee and unlock Audyssey Pro capabilities in the 886.

Either way, i want to utilse the power of the Sound Equalizer in my system

I could also try and engage THX Loudness Plus in the 886 as well when Audyssey processing in the 886 is off But then that is a totally different discussion.

What you say about trying it out would not hurt. I can always go to the original plan.

Out of interest Chris, what would the negatives be from an Audyssey point of view with the approach above?

Many Thanks!

QUESTION:

ANSWERS:



Thanks for the responses.

Well, these statements stand true for the 2010 Audyssey models with MultEQ XT32 (+ SubEQ); however the XT32 and the AS-EQ1 are not the only units that can apply individual calibration to dual subs.

In 2009 the 1007, 3007 and 5007 had MultEQ XT. These XXX7 receivers DID apply individual calibration/eq to EACH sub. These 2009 receivers fall short in that they did not ping the dual subs together, like the AS-EQ1 and the new XT32.

I was hoping that the 1008 could at least have the true dual sub calibration like the 1007 had.

Hi Markus,

As I understand it we have three situations:

1. devices that only have a single subwoofer output
2. devices that have two separate subwoofer outputs in which the timing, levels and equalization is applied separately
3. devices that have two separate subwoofer outputs in which the timing and levels are applied separately, but equalization is applied to both subwoofers simultaneously

I don't think deepstang was disagreeing that the third situation is the best, but merely stating that the second situation was superior to the first.

For folks who have two subwoofers that can't be located equidistant from the primary listening location, do you agree that being able to more accurately set the timing and levels individually is an advantage over the first situation where both subwoofers have the same settings for timing, level and equalization?

Thanks.

Larry

P.S. It is not clear to me that the 1008 differs from the 1007 in its handling of two subwoofers. The only definitive statements I recall reading on this general point of separate versus simultaneous equalization of two subwoofers has come from Chris. I believe that even the device manuals do not specifically discuss this point. Beyond Chris I suppose that it will require an owner of the device in question to determine whether or not both subwoofers are being "pinged" simultaneously.

I have sent this question to Audyssey, but figured it might be beneficial to all if it was discussed on the forum. This question is in regard to Audyssey Pro and the target curves, addressing my main speakers in particular.

Setup:
-Acoustically treated room, 5000 cu. ft. with very reasonable response without Audyssey.
-Two way main speakers where the HF drivers roll off on their own from 13kHz on up.
-Actively crossed system whereby I can EQ at will. I have EQ'd the HF driver prior to running Audyssey to bring up the driver's natural rolloff.
-Pre/Pro is an Integra DHC 9.9 and it only has target curves that roll off the HF, no flat curve selectable.

Problem:
-After running Audyssey Pro and selecting Target curve with roll off #1, music lacks good balance in high frequencies. Although Audyssey Pro does an excellent job with most the frequency spectrum, I end up with the HF spectrum muted a bit much.

Potential Solution:
- I suppose a possible solution is to EQ the HF driver after I run Audyssey Pro since I am not limited to +3dB as in the target curve editor. However, in order to do this, I need to understand how Audyssey treats what it "hears" during tone bursts in order to achieve the selectable target curves.

Question:
- Does Audyssey Pro calculate the FR results to achieve the selected target curve, or does it apply the 'target curve rolloff' to what the system response is at the time of testing?

What I'm trying to ascertain is if Audyssey Pro tries to modify my HF driver's natural rolloff to meet a predetermined roll off curve, or does it add roll off to an already poor HF response? I hope this makes sense.

Yes, I know getting a better driver is the way to go....and I'm working on that. In the meantime, this will help me understand Audyssey a bit better. However, I wish Integra would allow me to select a flat response.

MultEQ (Pro or otherwise) measures the response of your speaker in your room. Then it creates a filter that tries to shape that response so it matches the target curve.


Yes

Hi Rudy,

All versions of Audyssey, including the Pro version, apply filters to the frequency response to boost or cut frequencies as need to match the selected target curve at the frequency in question. If your high frequency driver's actual in room response is less than the target curve, Audyssey will apply up to a +9 dB boost.

You can force a flat target curve on your preamplifer by selecting a THX listening mode, and turning off THX Re-EQ. The THX Music listening modes do not have Re-EQ so it is not necessary to turn it off to obtain a flat target curve.

Larry

Chris, thank you for the quick response. It is likely that Audyssey Pro is 'undoing' the EQ I set in my active crossover since my HF driver unfortunately has that natural rolloff already.

I'm now thinking the best thing is to avoid EQ'ing the HF portion before I run Audyssey Pro and let it modify what it needs to achieve target curve number one. I should then be able to use the Target curve editor as much as possible to reduce the HF rolloff. That may be a better approach that what I have been doing.

Thank you for your help.

Perhaps then, the better way to listen to 2-channel music is in THX music? Frankly, I have never messed around with that mode and will give it a try today.

I think I will run Audyssey Pro again, but remove the EQ I have set up in my Ashly active crossover. My approach of EQ'ing prior to Audyssey is just a waste of processing in the crossover.

Hi Rudy,

In most home listening environments there is a good reason to have a high frequency roll-off. That's why Audyssey designed the default target curve with a roll-off in the first place.

The reason for this is that generally the mixing environment is less reverberant than the home listening environment and therefore the high frequency roll-off is needed to reduce undue brightness on playback in most homes.

If your listening room acoustics are similar to the recording environment, then a flat curve may be appropriate. Everything I have discussed is true of any content, not just two channel.

Larry

Hey Larry,

I agree, and that is my understanding of the intent of the rolloff curves. In my room, direct HF reflection points have absorbers and so I don't get that HF reflection an untreated room would produce. I am trying to achieve the flat curve, at least so I can see how it sounds, but at least directly, Audyssey Pro and my Integra do not present the flat target curve option. I believe that is an Integra issue, not Audyssey.

I am going to remove the EQ from my Ashly crossover and run Audyssey Pro and see how that changes things, if at all. I can then use the EQ to approximate a flat HF response. In the end game, that may be too little rolloff, but right now, I have too much.

I ran a couple of tests with my new ACM1H mic (thanks Onkyo/Chris) and had the Polk FXiA4 surrounds in bipole mode and while checking the levels later on, found that the surrounds were set exactly at the same level as all the other sat's, as measured with a RS analog meter and using the test tones from the Onkyo SR707.

I then ran another test yesterday with the surrounds in the dipole mode and the surround levels then went up 2-3db. I have the 1st mic position set in the null zone (in dipole) and I believe that is the cause of the difference.

I'm not overly concerned with this as it's easily remedied by lowering the levels, but it may explain why some people complain of boosted surrounds.

Yet, I wonder, if this could cause a difference in the way that Audyssey applies the filters for these speakers, in the way they are set up-bipole or dipole-or am I just over thinking this while eagerly awaiting a Packer victory in the Super Bowl.

Packers!

Hi Rudy,

Yes, it remains to be seen whether your high frequency drivers are capable of a flat response regardless of the means you use to boost the high frequency response.

If you select say a THX Music listening mode Audyssey will set filters in an attempt to achieve a flat response. However, I believe when the actual measured high frequency roll-off exceeds 3 dB then Audyssey will start tappering off the filters and will cease attempting to boost the highs.

If you use the EQ to attempt to flatten the high frequency response does it provide a means of displaying whether the measured response is flat?

Larry

Hi Patrick,
Yes, the difference in level is because of the dipole null. And, yes, the filters created will be different (mostly in the high frequency range) because the radiation pattern of a dipole is different from a bipole and therefore the interaction with the room is different.

Larry: I can always measure the results in my room using REW. However, I am not hung up on being perfectly 'flat'. My sole goal is to boost the hf enough to make it sound more natural.

This issue is primarily due to my HF drivers and their inability to reach good response above 13kHz. I am working on getting some drivers which will have a much better FR overall. In the mean time, I am using this as a learning lesson on using Audyssey with an active EQ system.

Hi Chris,

What I find interesting is that Patrick measured an increase in levels in the null zone when in dipole mode. Since the same drivers are being used and the same level signal is being applied, why would out of phase drivers produce a greater response than in phase drivers?

I suppose the explanation is that the new filters for the dipole operation are somehow boosting levels more than in the bipole mode in the 500-2000 Hz range where SPL is measured, but this seems counter intuitive for out of phase drivers. Wouldn't the level of all frequencies be cut in the null zone when in dipole operation?

Thanks.

Larry

Hi Larry,
Unless I misread what Patrick said, I interpreted as follows: the dipoles produce a null in the direction of the mic. That causes the mic to read a lower level and therefore a higher required trim value is reported.

The level measurement and setting is separate from the filters. The null we are talking about is a "level null" not a frequency response null.

Hi Chris,

Yes, that's what I meant. Where I place the 1st position mic, is at 90 degrees or directly in line with the center of the surrounds.

So, taking all this into consideration, would you recommend leaving the surround level at where MultEQ sets it or wander over into preference land and match them to the other sat's?

OK, you had me for a minute and then you lost me... According to MultEQ the speakers are already matched to the other sats.

Hi Chris,

Thanks.

Duh, I apparently did the misreading.

I thought Patrick was referring to a measured SPL level rather than the trim levels. Yes, it makes perfect sense that the trims would increase in going from bipole to dipole operation.

Larry

Yes, but...when I use the built in test tones from the Onkyo AVR they come up +2-3db higher than the other sat's (in dipole config).

I'll just try them both ways and let my ears be the judge.

Thanks for your help.

Packers!

Hi Markus,

Can you please elaborate on your reasoning?

Are you saying that even in a situation where the two subwoofers are at significant differences in distance from the primary listening locations, that it is more important to ping both subwoofer simultaneously rather than have accurate separate timings and levels for each subwoofer?

If so, there should be an easy fix. Just use a "Y" cable to connect from one subwoofer output to both subwoofers.

Thanks.

Larry

Actually, I have corresponded with Welti on this question, and it turns out that the subwoofers can be optimized individually. That is how the BassQ system does it.

[edit] this is assuming that impulse responses are measured, definitely more complicated than level, distance, and FR response measurements. Then, a transfer function matrix can be used to solve the system using the individual impulse responses. More complicated than parametric equalization would allow, in which case the subs should be measured together. Sorry about lack of clarity.

Hi Markus,

If is not too much trouble perhaps you could point me to the relevant chapter in Dr. Toole's book.

Thanks.

Larry

Trying to catch up here and my cursory reading of replies might have missed this, but I did not see anyone make the observation that the Audyssey Reference Curve is for CINEMA. It is not uncommon for people listening to music to make comments like yours.

Some gear allows selection of Audyssey Flat which, as it suggests, moves from the cinema curve to a flat one. Note, I don't think it's *really* flat as that would sound way too bright.

Other gear allows a crude workaround to engage Flat and that is to select a THX listening mode and defeat Re-EQ. When a THX mode is engaged, Audyssey Flat is engaged; when Re-EQ is turned off, Flat remains engaged.

Mine falls into the latter category, but I have never been able to use that workaround in my OP885 .. "mode not available" or some such tripe, however I have created a target curve in MultEQ Pro that works .. for me anyway .. for all content. I, too, have a well-treated room .. viewable at the site linked in my sig.

Jeff

Ah, I misunderstood your point(s).


No, it is not, nor did he say that it is. However, the assumption that the system is linear still holds. From the Welti paper:

"SFM relies on the principle of acoustic superposition to simulate a large number of subwoofer configurations. The first step is to measure the complex transfer functions from each potential subwoofer location to each listening location. Measurements must be complex."

This ties into the impulse responses that I mentioned before.

Then, "After measurements have been gathered it is possible to calculate the expected response at any seat for any subwoofer combination simply by adding the individual contributions of each source at each receiver (we assume the system to be linear)."

The sources do not need to be measured in tandem or simultaneously.


I see what you're saying. I thought you were arguing that one had to "ping together." Neither SFM nor BassQ pings together.


Yes. Okay, I think we are in agreement.

As I recall, "Sound Reproduction" does not address this particular issue in any significant depth, other than to describe SFM in general terms.

Jeff: You are correct in that Audyssey is mainly aimed at cinema. Let me throw it out there that as the quality of some pre/pro systems have improved, more of us are using that type of controller for both cinema and pure 2 channel listening. I think that might be an area that Audyssey could exploit to its benefit!

In my case, the main issue was not so much the roll off itself, but when coupled with a weak HF driver, it really dulled the highs. After the instructive answers here, I went back and started again.

I set my Ashly crossover to flat, and removed the EQ settings I had been using to compensate for the driver weakness. I then let Audyssey just do its thing and then tested the results subjectively by listening to familiar music rich in high frequencies.

Let me say that Audyssey by itself did a much better job of EQing the HF drivers than I had done manually! I set the target curve to rolloff #1, removed the mid-range compensation, and then sat back to listen. From the start, my highs had returned to a more normal level. I fear my manual EQ had not added enough dB boost the problem area.

After listening for a while, I made minor tweaks to Target Curve Editor to boost above 13kHz a little.

Overall much improved frequency response in my listening room for music is now available. This exercise just made me realize how badly I need to get a better driver that can cover from 500Hz to 20kHz. I am currently looking at some options.

BTW, my system is now great for demonstrating the advantages of Audyssey vs. not using it.

I wish Audyssey Pro would let you see the initial results of distance and level after the first run instead of having to wait for 3 runs before you can look. It would make it easier to ensure the initial microphone position is correct down to the tenth.