Hi Larry,

I honestly can not speak intelligently with the use of on-board parametric EQ's. I have never used one. I messed with a B&K receiver that had one just to see what the resolution was which was 1/3 of an octave. Some may be good, but I would be inclined to say that on-board EQ's on most receiver's or pre/pro's just don't have the resolution you would like since you really want to find that center frequency if possible.

If your receiver is Audyssey Pro capable, I personally would prefer to see that used than the on-board, but it just all depends on which direction money wise you want to go...and if you're lucky. The positive about Audyssey is that it is quite good at what it does provided you have someone who understands exactly what they are doing with it, but the negative is that you are limited on how much "control" you have.

Regarding target curves, most of the choices you have in Audyssey can be achieved through a shelf filter, which most have on-board and those do work, or a natural roll-off. I find the default target curve usually the best choice and so do most of the people I service.

Hi Shawn,

Thanks very much for the response.

However, I don't understand your comment highlighted in blue above. I was under the impression that Audyssey Pro had all the control that one could possibly want since the target curves are completely customizable.

If you don't mind, might I press you, and the other professionals following this thread, on a point. How can any parametric equalizer, on-board or external, with only a few equalization bands, compete with the literally thousands of control points provided by Audyssey Pro room correction? In both cases, manual or Audyssey Pro, we have "a professional behind the wheel". I could see a competing room correction software product offered as an alternative, but are there really any parametric equalizer alternatives that as a professional you would recommend to a client that already had an Audyssey Pro-ready AVR/Prepro? Audyssey Pro is more accurate, works in the time domain as well as frequency domain, and is more customizable than a parametric equalizer. It seems like a no-brainer to me.

By the way, with regard to the professional calibrator I mentioned earlier who advertises on his website better results with manual calibration over Audyssey calibration, I should point out he is not a certified Audyssey Installer. So he doesn't offer Audyssey Pro calibration. What has occurred is that consumers with Audyssey enabled AVRs who are not entirely happy with how they have setup their consumer version of Audyssey hire him to improve upon what they have done. The calibrator never rolls up his sleeves to figure out what's wrong with the client's implementation of Audyssey. He merely ignores Audyssey and uses the on-board equalization. In some cases the clients actually had Audyssey Pro-ready AVRs, but since the calibrator wasn't Audyssey certified that more capable option was never made available to the clients. Based on the number of hours required to obtain reasonable manual results the Audyssey Pro option may have actually been both more accurate and cheaper.

Thanks.

Larry

If your receiver is Audyssey Pro capable, I personally would prefer to see that used than the on-board, but it just all depends on which direction money wise you want to go...and if you're lucky. AudysseyPro can be used with the on-board DSP or with the external SEQ.

AudysseyPro can be used with the on-board DSP or with the external SEQ.

Kal, I was referring to an on-board parametric EQ, such as what B&K offers, not the Audyssey system itself.

Assuming your client has gone as far as he/she can to deal with room design, seating, speaker positioning, etc., as practical considerations (budget, aesthetics, etc.) will permit, and he/she has an Audyssey Pro-ready AVR/Prepro, would you ever recommend a different room correction approach than an Audyssey Pro calibration? In other words, for example do you believe it is possible for you to perform a more accurate room correction by using the simple parametric equalizer that is also available on many AVR/Prepros?


If your receiver is Audyssey Pro capable, I personally would prefer to see that used than the on-board, but it just all depends on which direction money wise you want to go...and if you're lucky..

AudysseyPro can be used with the on-board DSP or with the external SEQ.

Hi Kal,

Your response of course is correct, but it is not entirely relevant to the particular question I posed to Shawn (and the other professional calibrators following this thread).

I was asking him whether he thought it was possible for a professional calibrator to use the on-board parametric equalizer on many AVR/Prepros to perform a more accurate room correction than Audyssey Pro.

You may recall we had some interesting exchanges with a professional calibrator on the Audyssey thread, in which he claimed to provide better results than the consumer version of Audyssey setup by his clients. As you pointed out in that thread it was an "apple and oranges" comparison since the consumer did the Audyssey calibration and the professional did the manual calibration. In this thread I'm trying to elicit some responses from professionals when dealing with an "apples to apples" comparison where the calibrator would in theory being doing both the Audyssey and manual calibration.

Larry

Hi Kal,

Your response of course is correct, but it is not entirely relevant to the particular question I posted to Shawn (and the other professional calibrators following this thread).

I was asking him whether he thought it was possible for a professional calibrator to use the on-board parametric equalizer on many AVR/Prepros to perform a more accurate room correction than Audyssey Pro.

You may recall we had some interesting exchanges with a professional calibrator on the Audyssey thread, in which he claimed to provide better results than the consumer version of Audyssey setup by his clients. As you pointed out in that thread it was an "apple and oranges" comparison since the consumer did the Audyssey calibration and the professional did the manual calibration. In this thread I'm trying to elicit some responses from professionals when dealing with an "apples to apples" comparison where the calibrator would in theory being doing both the Audyssey and manual calibration.

Larry

Larry, are you asking if I would use both, or do a test to compare one vs. the other? If the latter, I've already answered that question. However, if a system is employed, I go with only one system. I do not employ multiple forms of EQ in a system with the exception of an Audyssey pro calibration on the receiver/pre-pro and marry it with the SubEQ since they are designed to be employed together.

Larry, are you asking if I would use both, or do a test to compare one vs. the other? If the latter, I've already answered that question. However, if a system is employed, I go with only one system. I do not employ multiple forms of EQ in a system with the exception of an Audyssey pro calibration on the receiver/pre-pro and marry it with the Bass EQ since they are designed to be employed together.

Hi Shawn,

I was asking from your experience whether you prefered any other equalization approaches to Audyssey Pro, when a client already has a device that is Audyssey Pro-ready. Yes, you already answered the question, but your response (quoted below) was brief and evoked luck so I was hoping you might elaborate a bit on the pros and cons of different approaches. :D

If your receiver is Audyssey Pro capable, I personally would prefer to see that used than the on-board, but it just all depends on which direction money wise you want to go...and if you're lucky..

It should come as no surprise that certain calibrators are not fans of Audyssey calibration, and invariably will always recommend a manual approach. Perhaps they have solid technical reasons to select what appears to be a less technically sophisticated approach. Some skeptics might suggest that this aversion to Audyssey, or any automated approach for that matter, makes perfectly good $ense since manual calibrations, requiring more time, will always be more lucrative than automated calibrations.

So I hope you can appreciate that some potential clients might have some concerns if a calibrator claims to yield "better" results using a rather coarse parametric equalizer in place of rather sophisticated, high resolution room correction software. Yes, under such circumstances a potential client might value objective comparitive test results.

I understand that from a practical standpoint requesting comparitive tests using a client's equipment and room is simply not feasible or a good use of a calibrator's time. However, comparing the results of some novice client using a consumer version of Audyssey versus a professional doing a manual calibration may not exactly provide a meaningful comparison either.

I guess ideally it would be nice if a professional trained in Audyssey could show an example of what Audyssey Pro can do in the hands of a professional versus what that same professional was able to do with the on-board parametric equalizer. No, it wouldn't be the client's room and gear but just a sample of a level playing field comparison.

Larry

if Audessey has 10,000 points of adjustment and 4500 curves, but not the point i want with the shelving adjustment i want... then my outobard with 10 points and variable shelf is better.

if Audessey has 10,000 points of adjustment and 4500 curves, but not the point i want with the shelving adjustment i want... then my outobard with 10 points and variable shelf is better.

Hi,

Thanks for the response.

I fully appreciate your point in the theoretical.

From your professional experience as a practical matter in most cases have you found that Audyssey Pro's curve adjustability has been too limited to suit your client's needs such that you abandoned it and used outboard equalization?

Larry

i have no experience with Audessey.

But i have known lots of pros who are very particular about the tools they use. so in some cases they may just prefer a different product.

i have no experience with Audessey.

But i have known lots of pros who are very particular about the tools they use. so in some cases they may just prefer a different product.

Hi,

As a group do most of them tend to avoid automated calibration approaches?

Thanks.

Larry

Hi,

As a group do most of them tend to avoid automated calibration approaches?

Thanks.

Larry

I read the thread and the specifically posts that you pointed out. Interesting read. For some reason, there was some hostility in that thread against varying opinions.

I belong to the crowd who had a chance to compare both Audyssey and that calibrator's work. In my subjective opinion, his results were clearly superior.

I belong to the crowd who had a chance to compare both Audyssey and that calibrator's work. In my subjective opinion, his results were clearly superior.

Hi,

Can you please elaborate?

Was the Audyssey calibration the consumer version performed by the client and the calibrator's version was using the on-board parametric equalizer?

Was this strictly a subjective listening comparison, or did the calibrator also provide comparative measurements?

Thanks.

Larry

If you don't mind, might I press you, and the other professionals following this thread, on a point. How can any parametric equalizer, on-board or external, with only a few equalization bands, compete with the literally thousands of control points provided by Audyssey Pro room correction? In both cases, manual or Audyssey Pro, we have "a professional behind the wheel". I could see a competing room correction software product offered as an alternative, but are there really any parametric equalizer alternatives that as a professional you would recommend to a client that already had an Audyssey Pro-ready AVR/Prepro? Audyssey Pro is more accurate, works in the time domain as well as frequency domain, and is more customizable than a parametric equalizer. It seems like a no-brainer to me.


Thanks.

Larry

I'm not so sure that the accuracy of Audyssey's filters is really needed (except for bass frequencies) and it seems to me that for the mids and highs that broader filters may be the better solution.

From Audyssey's website, here is a visual of what it is doing. The red dots are the samples it takes of the FR.
http://www.audyssey.com/images/technology/graphs/graph2.gif

It then creates an almost exact inverse of the response to cancel out the dips and peaks (the yellow line).
http://www.audyssey.com/images/technology/graphs/graph4.gif

Why wouldn't you want an exact inverse of the FR? Because, it's an exact fix for an averaged reading. If Audyssey applies an exact inverse of a dip that isn't there in the spot you are listening in, then it has altered the FR in a way that is no longer accurate to your current seat.

For instance, lets say Audyssey measures a series of peaks at 600, 650, 700, 750, 800, and 900Hz after calculating the average response from multiple mic positions. Move the mic (or your head) 6" in any direction and the center point for any peak measured will move a few Hz. Since you have a moving target, why not just set a broad filter that pulls down the range of 600 - 900Hz. The overall trend will be better and you haven't caused peaks where there were dips and vice versa.

This is the averaged response of my current center speaker.
http://i80.photobucket.com/albums/j193/ptwalker/measurements/wbcentreavgnoeq.jpg

Here's what I was able to do with just five PEQ filters in the SSP-800.
http://i80.photobucket.com/albums/j193/ptwalker/measurements/wbcentrewitheqnovisfilters.jpg

Here are the filter settings.
http://i80.photobucket.com/albums/j193/ptwalker/measurements/wbcentreavgwitheq.jpg

Are the results as ruler flat as Audyssey's "after" graph? No, but then Audyssey's "after" graph is not taken from an actual reading, it is only a predicted reponse based on an average, which doesn't accurately represent any one particular measurement point.

I was searching through my archived e-mails between Chris K. (Audyssey) and I, because I'm sure that I asked him directly what Audyssey was doing in the time domain. I couldn't find the correspondence, but I seem to remember him saying something to the effect that Audyssey derives the FR filters by measuring the test tones in the time domain, but nothing about specifically targeting time related issues like decay. I still think that their claims of "working in the time domain" are potentially misleading.

In the end, Audyssey and PEQ are simply two different means to the same end. Yes, Audyssey is technically more precise, but when that precision isn't needed (because you are correcting an average), it's more marketing hype and specsmanship than true value.

One should also keep in mind that the principles behind Audyssey are only one perspective on what needs to be corrected and how it should be corrected. There are others. For instance Meridian believes that you should target long decay and not FR.

BTW - there is no official training by Audyssey for pro calibrators, and I don't believe that they even have any sort of qualifications that must be met. But then they really don't need to. Other than adjusting the target curve (if desired), Audyssey Pro calibration is handled pretty the same way as the consumer version. It just has better resoltion, a better mic, and more flexibility.

I'm not so sure that the accuracy of Audyssey's filters is really needed (except for bass frequencies) and it seems to me that for the mids and highs that broader filters may be the better solution.

Hi,

Audyssey of course devotes more filters to the subwoofer channel than the satellite channels to conserve DSP resources.


Why wouldn't you want an exact inverse of the FR? Because, it's an exact fix for an averaged reading. If Audyssey applies an exact inverse of a dip that isn't there in the spot you are listening in, then it has altered the FR in a way that is no longer accurate to your current seat.

For instance, lets say Audyssey measures a series of peaks at 600, 650, 700, 750, 800, and 900Hz after calculating the average response from multiple mic positions. Move the mic (or your head) 6" in any direction and the center point for any peak measured will move a few Hz. Since you have a moving target, why not just set a broad filter that pulls down the range of 600 - 900Hz. The overall trend will be better and you haven't caused peaks where there were dips and vice versa.


Also from the Audyssey website:

http://www.audyssey.com/images/technology/calculatedsystem.gif


How does MultEQ measure your room differently?
MultEQ looks at patterns in the time domain responses and classifies them into clusters based on the similarities in those patterns, typically in 3-5 groups. A representative response is created from each cluster, and a final response is then created from grouping the representatives. That response is then used to create the EQ filter. It is a complicated process that is based on the complex mathematics of pattern recognition and fuzzy logic. But there is nothing fuzzy about the results. If you want to know more, we offer copies of selected publications we have written for technical conferences.

Time and Frequency correction:

1.The time domain is where many of the problems reside. Parametric and graphic equalizers can only correct for the frequency response and do so in a very coarse manner because they have limited resolution (bands).
2.Further, whether they have fixed or adjustable bands it does not matter because bands cause phase problems that most people hear as "ringing" or "smearing." This is why, after thirty plus years of trying this method most people don't like the results and turn it off.



So the point is that Audyssey is not merely averaging the frequency responses at multiple locations, it is looking for common acoustical issues in multiple locations and weighing more heavily those common problems. This is a more sophisicated and targeted approach that using a broad filter.


I was searching through my archived e-mails between Chris K. (Audyssey) and I, because I'm sure that I asked him directly what Audyssey was doing in the time domain. I couldn't find the correspondence, but I seem to remember him saying something to the effect that Audyssey derives the FR filters by measuring the test tones in the time domain, but nothing about specifically targeting time related issues like decay. I still think that their claims of "working in the time domain" are potentially misleading.


Here's is an article by Kal Rubinson in Stereophile magazine. Notice the improvement in spectral decay. It doesn't look like misleading claims to me and this isn't Audyssey marketing, but an objective third party. :D

Music in the Round #23 (http://stereophile.com/musicintheround/307mitr/index1.html)


http://stereophile.com/images/archivesart/307Mirfig1.jpg
Fig.1 Right-channel speaker, in-room cumulative spectral decay plot, 20–200Hz, without equalization, measured with ETF 5. (10dB/vertical div., 800ms time window)


http://stereophile.com/images/archivesart/307Mirfig2.jpg
Fig.2 Right-channel speaker, in-room cumulative spectral decay plot, 20–200Hz, with Audyssey equalization, measured with ETF 5. (10dB/vertical div., 800ms time window)


http://stereophile.com/images/archivesart/307Mirfig3.jpg
Fig.3 Subwoofer, in-room cumulative spectral decay plot, 20–200Hz, without equalization, measured with ETF 5. (10dB/vertical div., 800ms time window)


http://stereophile.com/images/archivesart/307Mirfig4.jpg
Fig.4 Subwoofer, in-room cumulative spectral decay plot, 20–200Hz, with Audyssey equalization, measured with ETF 5. (10dB/vertical div., 800ms time window)



One should also keep in mind that the principles behind Audyssey are only one perspective on what needs to be corrected and how it should be corrected. There are others. For instance Meridian believes that you should target long decay and not FR.


Yes, but why not target problems in both the frequency and time domain? Kal's measurement demonstrate that Audyssey is addressing long decay times as well as smoothing frequency response. ;)

Larry

I belong to the crowd who had a chance to compare both Audyssey and that calibrator's work. In my subjective opinion, his results were clearly superior.

That pretty much reflects my own experience with my last HT, with one key difference....I didn't have a subjective opinion, it was clearly, and objectively demonstrated.

The difference between Audyssey and the Lake units was certainly audible, but was also clearly demonstrated with the room response software.

Audyssey asks the right questions, but that's no guarantee that it will generate the right answers.

Here's is an article by Kal Rubison in Stereophile magazine. Notice the improvement in spectral decay. It doesn't look like misleading claims to me and this isn't Audyssey marketing, but an objective third party. :D

Yes, but why not target problems in both the frequency and time domain? Kal's measurement demonstrate that Audyssey is addressing long decay times as well as smoothing frequency response. ;)

Larry

You get the exact same effect with PEQ because FR and decay are inextricably linked.

Here's an interesting post on the matter.

http://www.hometheatershack.com/forums/rew-forum/7135-waterfalls.html

In the end, Audyssey and PEQ are simply two different means to the same end. Yes, Audyssey is technically more precise, but when that precision isn't needed (because you are correcting an average), it's more marketing hype and specsmanship than true value.

This is a misunderstanding of the nature of Audyssey's measurements. Unfortunately, their website is a little bit misleading as well. I can't speak for anybody at Audyssey, you can ask them directly in their official thread over in the amps/receivers forum, but my understanding as they have explained it there and elsewhere is that measuring in a single position is not the proper way to measure response.

Unfortunately their website makes it appear as if they are correcting an average, or attempting to correct for multiple locations, when in my understanding this is not the case. The goal is to characterize as completely as possible what the actual room and system response is in the listening position, and according to their studies the best way to do this is to take a variety of samples closely clustered around the main seating position to get the best characterization of that location. Again, this is the way I understand it only, but taking single measurements can lead to some very anomalous results and unique peaks and dips that are outliers that shouldn't really be corrected for, and taking a variety of clustered measurements tightly around that listening position gives the most accurate characterization of that position by being able to identify what are anomalous measurements that are particularly unique to a very narrow position, versus those that are more universal. It isn't just an average curve from a bunch of measurements. Some of the measurements taken may not be used at all, in my understanding. For instance if you take a whole bunch of measurements that are clustered right around the sweetspot that are all quite similar, and then move the microphone and take a measurement in a corner location with a totally different response curve, I believe that measurement would be ignored entirely.


BTW - there is no official training by Audyssey for pro calibrators, and I don't believe that they even have any sort of qualifications that must be met. But then they really don't need to. Other than adjusting the target curve (if desired), Audyssey Pro calibration is handled pretty the same way as the consumer version. It just has better resoltion, a better mic, and more flexibility.

That's not true either. Audyssey worked with my previous company with training, though as with any "training" it's usually more of a training of salesman. I wasn't present for the training, however, so I'm only commenting second-hand. However, it is a fairly straightforward process.

Was the Audyssey calibration the consumer version performed by the client and the calibrator's version was using the on-board parametric equalizer?

Consumer Version

Was this strictly a subjective listening comparison, or did the calibrator also provide comparative measurements?

Listening comparison. That was why I qualified my statement with "subjective"

Why wouldn't you want an exact inverse of the FR? Because, it's an exact fix for an averaged reading. If Audyssey applies an exact inverse of a dip that isn't there in the spot you are listening in, then it has altered the FR in a way that is no longer accurate to your current seat.

Are the results as ruler flat as Audyssey's "after" graph? No, but then Audyssey's "after" graph is not taken from an actual reading, it is only a predicted reponse based on an average, which doesn't accurately represent any one particular measurement point.

I was searching through my archived e-mails between Chris K. (Audyssey) and I, because I'm sure that I asked him directly what Audyssey was doing in the time domain. I couldn't find the correspondence, but I seem to remember him saying something to the effect that Audyssey derives the FR filters by measuring the test tones in the time domain, but nothing about specifically targeting time related issues like decay. I still think that their claims of "working in the time domain" are potentially misleading.

In the end, Audyssey and PEQ are simply two different means to the same end. Yes, Audyssey is technically more precise, but when that precision isn't needed (because you are correcting an average), it's more marketing hype and specsmanship than true value.


I don't work for Audyssey nor do I believe they are necessarily the best overall solution for time domain correction in all instances. There are other solutions (like the Tact) that have been around longer and do offer more flexibility. (at about 10 times the price). I did, however, for 10 years work for another (now defunct) company that sold the first commercially available digital time domain room correction product. If I were trying to get the VERY BEST digital time domain correction resolution that money can buy, it would not be the Audyssey. But for the market that I believe Audyssey is going after, my hats are off to them. Audyssey Pro, is $700. A two channel Tact unit is about $7000 --- and requires a VERY VERY SIGNIFICANT time investment of the user to get the very best results!!! The basic Audyssey solution that is sold with so many receiver and pre-pros is an outstanding solution for the average guy/girl who just wants his/her system to sound better with minimal effort. And in probably 90% of the cases, that is what happens

That said:

If you want the Audyssey correction to more closely mirror what you get when you use some other device (PEQ) then run all measuring points from the listening position as opposed to their recommend approach of 8 different positions.

I don't know how the Audyssey system works, but one solution to the problem you described of a 6 inch head movement changing the result is for Audyssey to decrease the filter granularity as you move up in frequency response. That is what we did and I would be surprised if they do not do something similar. I can assure you that based upon the way the Audyssey works, the do "work in the time domain" The test tones alone will tell you that. And a Fast Fourier Transform of the time measurement does give you the frequency response.

"You get the exact same effect with PEQ because FR and decay are inextricably linked. "

I thought this was only the case at freq up to a couple hundred Hz in a normal sized room.

You get the exact same effect with PEQ because FR and decay are inextricably linked.

Hi,

To be clear, I made no claim that using a parametric equalizer to knock down peaks won't also aide in reducing modal ringing by simply removing energy from the system.

What I was responding to was your suggestion that Audyssey wasn't also addressing decay times as demonstrated in Kal's unbiased measurements.

Larry

If you want the Audyssey correction to more closely mirror what you get when you use some other device (PEQ) then run all measuring points from the listening position as opposed to their recommend approach of 8 different positions.

No, this is bad advice and will lead to inferior results. Again, it is unfortunate that Audyssey calls them positions, they aren't many (and way more than 8 is possible) listening positions, they are different sampling positions ideally clustered around the main listening position. If you want to achieve a broader correction for a larger area, then you'd sample more broadly. Leaving the microphone in one fixed sweetspot and running all the measurements from that unchanged position is not a good idea at all.

Leaving the microphone in one fixed sweetspot and running all the measurements from that unchanged position is not a good idea at all.

Agreed. Which is exactly what you get when you use a PEQ from a single location!!

Audyssey patent search, 2 I found:
System And Method For Automatic Multiple Listener Room Acoustic Correction With Low Filter Orders (http://www.freepatentsonline.com/y2009/0202082.html)

United States Patent Application 20090202082
A system and a methods for correcting, simultaneously at multiple-listener positions, distortions introduced by the acoustical characteristics includes warping room responses, intelligently weighing the warped room acoustical responses to form a weighted response, a low order spectral fitting to the weighted response, forming a warped filter from the low order spectral fit, and unwarping the warped filter to form the room acoustical correction filter.
What is claimed is:

1. A method for correcting room acoustics at multiple-listener positions, the method comprising: measuring with a microphone a room acoustical response at each listener position in a multiple-listener environment; processing each of the room acoustical response measured at said each listener position to obtain non-uniform resolution of the room acoustical response in an audio frequency domain, wherein the non-uniform resolution results in higher resolution at low frequencies for each of the measured room acoustical response; determining a general response by computing a weighted average of the processed acoustical responses; generating a low order spectral model of the general response; obtaining an acoustic correction filter from the low order spectral model, wherein the acoustic correction filter is the inverse of the low order spectral model; and processing the acoustic correction filter to obtain a room acoustic correction filter with uniform resolution in the audio frequency domain; wherein the room acoustic correction filter corrects the room acoustics at the multiple-listener positions.

2. The method of claim 1, further comprising generating a stimulus signal for measuring the room acoustical response at each of the listener positions.

3. The method of claim 1, wherein the general response is determined by a pattern recognition method.

4. The method of claim 3, wherein the pattern recognition method comprises a method selected from a group consisting of: a hard c-means clustering method, a fuzzy c-means clustering method, and an adaptive learning method.

5. The method of claim 1, wherein the spectral model comprises a model selected from a group consisting of a Linear Predictive Coding (LPC) model and a pole-zero model.

6. The method of claim 1, wherein the processing comprises psycho-acoustically motivated warping.

7. The method of claim 6, wherein the warping is achieved by means of a bilinear conformal map.

8. The method of claim 6, wherein the psycho-acoustically motivated warping is accomplished in the frequency domain.

9. A method for correcting room acoustics at multiple-listener positions, the method comprising: measuring with a microphone a room acoustical response at each listener position in a multiple-listener environment; processing each of the room acoustical response measured at said each listener position to obtain non-uniform resolution of the room acoustical response in an audio frequency domain, wherein the non-uniform resolution results in higher resolution at low frequencies for each of the measured room acoustical response; obtaining minimum-phase response of each of the said processed acoustical responses; determining a general response by computing the weighted average of the minimum-phase processed responses; generating a low order spectral model of the general response; obtaining an acoustic correction filter from the low order spectral model; and processing the acoustic correction filter to obtain a room acoustic correction filter with uniform resolution in the audio frequency domain; wherein the room acoustic correction filter corrects the room acoustics at the multiple-listener positions.

10. The method of claim 9, further comprising generating a stimulus signal for measuring the room acoustical response at each of the listener positions.

11. The method of claim 9, wherein the general response is determined by a pattern recognition method.

12. The method of claim 11, wherein the pattern recognition method comprises a method selected from a group consisting of: a hard c-means clustering method, a fuzzy c-means clustering method, and an adaptive learning method.

13. The method of claim 9, wherein the processing comprises psycho-acoustically motivated warping.

14. The method of claim 13, wherein the warping is achieved by means of a bilinear conformal map.

15. The method of claim 13, wherein the psycho-acoustically motivated warping is accomplished in the frequency domain

16. The method of claim 9, wherein the spectral model comprises a model selected from a group consisting of a Linear Predictive Coding (LPC) model and a pole-zero model.




System and method for automatic multiple listener room acoustic correction with low filter orders (http://www.freepatentsonline.com/7567675.html)

United States Patent 7567675
A system and a method for correcting, simultaneously at multiple-listener positions, distortions introduced by the acoustical characteristics includes warping room responses, intelligently weighing the warped room acoustical responses to form a weighted response, a low order spectral fitting to the weighted response, forming a warped filter from the low order spectral fit, and unwarping the warped filter to form the room acoustical correction filter.
The present application is a continuation-in-part of and claims priority to U.S. application Ser. No. 10/465,644 filed Jun. 20, 2003 which claims the benefit of priority to U.S. Provisional Patent Application 60/390,122 filed Jun. 21, 2002, which are incorporated by reference in their entirety herein.
Claims:
We claim:

1. A method for correcting room acoustics at multiple-listener positions, the method comprising: measuring with a microphone a room acoustical response at each listener position in a multiple-listener environment; warping each of the room acoustical response measured at said each listener position; determining a general response by computing a weighted average of the warped room acoustical responses; generating a low order spectral model of the general response; obtaining a warped acoustic correction filter from the low order spectral model, wherein the warped acoustic correction filter is the inverse of the low order spectral model; and unwarping the warped acoustic correction filter to obtain a room acoustic correction filter; wherein the room acoustic correction filter corrects the room acoustics at the multiple-listener positions.

2. The method of claim 1, further comprising generating a stimulus signal for measuring the room acoustical response at each of the listener positions.

3. The method according to claim 1, wherein the general response is determined by a pattern recognition method.

4. The method of claim 3, wherein the pattern recognition method comprises a method selected from a group consisting of: a hard c-means clustering method, a fuzzy c-means clustering method, and an adaptive learning method.

5. The method according to claim 1, wherein the warping is achieved by means of a bilinear conformal map.

6. The method of claim 1, wherein the spectral model comprises a model selected from a group consisting of a Linear Predictive Coding (LPC) model and a pole-zero model.

7. A method for generating substantially distortion-free audio at multiple-listeners in an environment, the method comprising: measuring with a microphone acoustical characteristics of the environment at each expected listener position in the multiple-listener environment; warping each of the acoustical characteristics measured at said each expected listener position; generating a low order spectral model of each of the warped acoustical characteristics; obtaining a warped acoustic correction filter from the low order spectral model, wherein the warped acoustic correction filter is the inverse of the low order spectral model; unwarping the warped acoustic correction filter to obtain a room acoustic correction filter; filtering an audio signal with the room acoustical correction filter; and transmitting the filtered audio from at least one loudspeaker, wherein the audio signal received at said each expected listener position is substantially free of distortions.

8. The method of claim 7, further comprising determining a general response by a pattern recognition method.

9. The method of claim 8, wherein the pattern recognition method comprises a method selected from a group consisting of: a hard c-means clustering method, a fuzzy c-means clustering method, and an adaptive learning method.

10. The method of claim 7, wherein the warping is achieved by a bilinear conformal map.

11. The method of claim 7, wherein the spectral model comprises a model selected from a group consisting of: a Linear Predictive Coding (LPC) model and a frequency weighted pole-zero model.

12. A system for generating substantially distortion-free audio at multiple-listeners in an environment, the system comprising: a filtering means for performing multiple-listener room acoustic correction, the filtering means formed from: (i) warped room acoustical responses, wherein the room acoustical responses are measured with a microphone at each of an expected listener position in a multiple-listener environment; (ii) a weighted average response of the warped room acoustical responses; (iii) a low order spectral model of the weighted average response; (iv) a warped filter formed from the low order spectral model, wherein the warped filter is the inverse of the low order spectral model; and (v) an unwarped room acoustic correction filter obtained by unwarping the warped filter; wherein an audio signal, filtered by the filtering means comprised of the room acoustic correction filter, is received substantially distortion-free at each of the expected listener positions; and a means for transmitting the audio signal.

13. The system of claim 12, wherein the weighted average response is determined by a pattern recognition means.

14. The system of claim 13, wherein the pattern recognition means comprises a means selected from a group consisting of a hard c-means clustering system, a fuzzy c-means clustering system, and an adaptive learning system.

15. The system of claim 12, wherein the warping is achieved by an all-pass filter chain.

16. The system of claim 12, wherein the spectral model comprises a model selected from a group consisting of a Linear Predictive Coding (LPC) model and a frequency weighted pole-zero model.

17. A method for correcting room acoustics at multiple-listener positions, the method comprising: warping each room acoustical response, said each room acoustical response obtained with a microphone at each expected listener position; clustering each of the warped room acoustical response into at least one cluster, wherein each cluster includes a centroid; forming a general response from the at least one centroid; inverting the general response to obtain an inverse response; obtaining a lower order spectral model of the inverse response; unwarping the lower order spectral model of the inverse response to form the room acoustic correction filter; wherein the room acoustic correction filter corrects the room acoustics at the multiple-listener positions.

18. The method of claim 17, wherein the warping is achieved by a bilinear conformal map.

19. The method of claim 18, wherein the spectral model comprises a frequency weighted pole-zero model.


Or, you can read all 20 with "Kyriakakis, Chris" http://www.freepatentsonline.com/result.html?p=1&edit_alert=&srch=xprtsrch&query_txt=Kyriakakis%2C+Chris&uspat=on&usapp=on&date_range=all&stemming=on&sort=relevance&search=Search

This is a misunderstanding of the nature of Audyssey's measurements. Unfortunately, their website is a little bit misleading as well. I can't speak for anybody at Audyssey, you can ask them directly in their official thread over in the amps/receivers forum, but my understanding as they have explained it there and elsewhere is that measuring in a single position is not the proper way to measure response.

Unfortunately their website makes it appear as if they are correcting an average, or attempting to correct for multiple locations, when in my understanding this is not the case. The goal is to characterize as completely as possible what the actual room and system response is in the listening position, and according to their studies the best way to do this is to take a variety of samples closely clustered around the main seating position to get the best characterization of that location. Again, this is the way I understand it only, but taking single measurements can lead to some very anomalous results and unique peaks and dips that are outliers that shouldn't really be corrected for, and taking a variety of clustered measurements tightly around that listening position gives the most accurate characterization of that position by being able to identify what are anomalous measurements that are particularly unique to a very narrow position, versus those that are more universal. It isn't just an average curve from a bunch of measurements. Some of the measurements taken may not be used at all, in my understanding. For instance if you take a whole bunch of measurements that are clustered right around the sweetspot that are all quite similar, and then move the microphone and take a measurement in a corner location with a totally different response curve, I believe that measurement would be ignored entirely.






You are completely correct. Audyssey does look for similarities between mic positions, I fully agree with the concept and performing all of my Audyssey calibrations with the clustering method. Unfortunately, for reasons beyond me and Chris, the corrected sound was subjectively worse than the untouched signal.

I think colmachine said it best: "Audyssey asks the right questions, but that's no guarantee that it will generate the right answers."

You are completely correct. Audyssey does look for similarities between mic positions, I fully agree with the concept and performing all of my Audyssey calibrations with the clustering method. Unfortunately, for reasons beyond me and Chris, the corrected sound was subjectively worse than the untouched signal.

I think colmachine said it best: "Audyssey asks the right questions, but that's no guarantee that it will generate the right answers."

For note of disclosure, I use Audyssey PRO on my denon 5308, along with Dynamic EQ. I never use dynamic volume. I am extremely pleased with Audyssey and dynamic EQ for film, and I've installed scores of systems with user-enabled Audyssey with exceptional results, though not as good as with PRO. I've also done and heard a handful of audyssey pro systems, and I've been extremely pleased with them as well.

For 2-channel music listening, Audyssey usually remains completely off, which is mainly for CDs and vinyl, and some radio.

For 2-channel music listening, Audyssey usually remains completely off, which is mainly for CDs and vinyl, and some radio.
How come? I find my TacT to do an incredible job on both stereo and 5.1. Indeed the effect for stereo is even stronger.

For note of disclosure, I use Audyssey PRO on my denon 5308, along with Dynamic EQ. I never use dynamic volume. I am extremely pleased with Audyssey and dynamic EQ for film, and I've installed scores of systems with user-enabled Audyssey with exceptional results, though not as good as with PRO. I've also done and heard a handful of audyssey pro systems, and I've been extremely pleased with them as well.

For 2-channel music listening, Audyssey usually remains completely off, which is mainly for CDs and vinyl, and some radio.

Seeing as you're in Seattle, I wish you could have heard Audyssey Pro in action with some of my previous speakers. I think you would have left scratching your head.

I also have the same question as armim. Why only run EQ on movies? I use EQ for everything. Good sound is good sound isn't it?

How come? I find my TacT to do an incredible job on both stereo and 5.1. Indeed the effect for stereo is even stronger.

I'm more of a purist when it comes to 2-channel, and I lose a lot of my imaging with EQ turned on.

I spent a long time listening various ways, and it improves certain things significantly, but it flattens my imaging like crazy. In fact, the times I use it are for critical listening of content I was helping to master, to "hear into" the audio better for analysis, but for me sitting down to music I turn it off.

I also don't really have a good way of testing the audibility of an extra ADC-DAC step, but I go analog into my 5308 for CDs, and I'm a little skeptical of doing that extra step.

Part of it may also be the fact that I've spent years with my current speakers before I got my 5308 with audyssey, and change isn't always welcome. Even when it is a mixed bag, which I feel it is for 2-channel in my system, we latch onto what changes for the worst immediately before we notice what has changed for the better that we didn't previously have.

But for HT use, the improvement really is night and day.

I'm more of a purist when it comes to 2-channel, and I lose a lot of my imaging with EQ turned on.

Hi Chris,

Do you have a theory why a flatter frequency response with less modal ringing would have an adverse impact on imaging?

If you are using Dynamic EQ with certain music content that is mixed at levels hotter than the film reference levels that might result in an over-emphasis in bass when playing below reference. Do you think that too much bass could explain a loss of imaging? If so, perhaps turning off Dynamic EQ but leaving MultEQ on might be the way to go for certain two-channel music.

Thanks.

Larry

Do you have a theory why a flatter frequency response with less modal ringing would have an adverse impact on imaging?I know this will be a "hot button" for some but many long-term audiophiles have become accustomed to a particular rendition of stereo sources based on years of listening in environments where the room acoustics contribute significantly to the overall soundstage and imaging (to say nothing of FR). Indeed, successful stereo often relies more heavily on the room's contribution even though it may be spurious in many ways.

I know this will be a "hot button" for some but many long-term audiophiles have become accustomed to a particular rendition of stereo sources based on years of listening in environments where the room acoustics contribute significantly to the overall soundstage and imaging (to say nothing of FR). Indeed, successful stereo often relies more heavily on the room's contribution even though it may be spurious in many ways.

Hi Kal,

Yes, as I understand it stereo requires a more lively acoustic environment to generate ambience via reflections, whereas multichannel furnishes ambience in the recording and ideally should have a less lively acoustic environment. Therefore, a home theater properly designed for multichannel movies wouldn't be ideal for listening to stereo. For this reason I always listen to multichannel music in my home theater, or upconvert two channel to multichannel.

Would you mind offering your thoughts on the second part of my previous posting?


If you are using Dynamic EQ with certain music content that is mixed at levels hotter than the film reference levels that might result in an over-emphasis in bass when playing below reference. Do you think that too much bass could explain a loss of imaging? If so, perhaps turning off Dynamic EQ but leaving MultEQ on might be the way to go for certain two-channel music.


Is it possible that Dynamic EQ might also effect the imaging of certain music mixes if the bass is over-emphasized?

Thanks.

Larry

Is it possible that Dynamic EQ might also effect the imaging of certain music mixes if the bass is over-emphasized?I suppose so. I have never used DynamicEQ so I have no hands-on experience to relate.

Hi Chris,

Do you have a theory why a flatter frequency response with less modal ringing would have an adverse impact on imaging?

I've thought a little bit about it, but without knowing more precisely exactly what audyssey does in a very intimate way, I can only guess. In my listening, one thing I notice occassionally is instability if a tightly (or should be tightly imaged) singer or instrument plays a wide range of notes, which would make sense since my left and right speakers are individually EQ'd, leading to perhaps different frequency responses between them. For instance, at one frequency I get a little more from the L channel, the image pulls that way, at another I get a little more from the R channel the image pulls right, and this could blur the imaging because different parts of the sound are being imaged in different places.

Which begs the question of why that would be the case, because if audyssey can correct FR with pretty significant precision, you'd think imaging would improve. However, I am listening in a small room, so this may make measurement a little more difficult because it is compensating heavily for the room too, not just the direct sound. I don't really know.

My other guess is that my mains are dynaudios, and I believe they used 1st order crossovers and keep phase aligned (if I recall correctly), which is rather odd I think for speakers. But dynaudio has a very unique and amazing imaging depth that you very very rarely hear from other speakers. I've read some literature on speaker design that many designers don't do this because doing so can sacrifice frequency response which they argue is much more audible than phase. Audyssey likely is improving frequency response at the expense of phase, and this may also impact imaging negatively. But this is a realm above my expertise, so again I really don't know exactly why.

But I do know it is different, and the imaging is very important to me and that's the main thing that is affected negatively.

On the other hand, my surround improvement is vastly improved, partly due to the improvement in imaging because now all the speakers have very similar frequency response. I've only heard this degree of surround sound coherence all the way around in systems that have properly placed speakers that are all identical, which is very rare. My rears are paradigms which don't match my main L/R and there isn't a matching center for my mains, I have a Dynaudio Contour center that is close but without EQ definitely isn't a very tight match, but with audyssey and particularly dynamic EQ turned on to keep the surrounds up in volume as I decrease from reference (I usually listen from -20 to -10 depending on the movie, average say -15), the imaging around me is simply amazing. But surround movies don't really have the depth of soundstage like stereo does, so that loss doesn't really bother me at all anyway because it's not like I'm really losing anything like I am with 2-channel listening.



If you are using Dynamic EQ with certain music content that is mixed at levels hotter than the film reference levels that might result in an over-emphasis in bass when playing below reference. Do you think that too much bass could explain a loss of imaging? If so, perhaps turning off Dynamic EQ but leaving MultEQ on might be the way to go for certain two-channel music.

Thanks.

Larry

To clarify what I said above, I was comparing EQ on, but Dynamic EQ turned off, as you suggested. Dynamic EQ I don't like for music because it is over-emphasized for most music content which is mixed at a much higher average SPL than most movies. I only turn dynamic EQ on for music if I'm doing a party mode or something, or just want to show off my subwoofer.

I know this will be a "hot button" for some but many long-term audiophiles have become accustomed to a particular rendition of stereo sources based on years of listening in environments where the room acoustics contribute significantly to the overall soundstage and imaging (to say nothing of FR). Indeed, successful stereo often relies more heavily on the room's contribution even though it may be spurious in many ways.

I also would agree and freely admit that. Again, this is one of those things where change from something you're very familiar with may be perceived negatively just because it's different, particularly if there is something minor that changes for the negative because we latch onto that first, and take longer to appreciate what may have improved.

Chris Wiggles wrote:
On the other hand, my surround improvement is vastly improved, partly due to the improvement in imaging because now all the speakers have very similar frequency response. I've only heard this degree of surround sound coherence all the way around in systems that have properly placed speakers that are all identical, which is very rare.

BINGO! You just nailed it on the head of what Audyssey does. It essentially changes the speakers response so that they have similar timbre matching and thus a more pleasing presentation. If possible, I always profess to use the exact same speakers for every channel, and one of the reasons that bookshelves are the way to go for any complete speaker package amongst a few other reasons.

This may be the reason you turn Audyssey off Chris while listening to two channel music as the FR of your front two speakers have been changed in order to better match your surrounds if they are not the same exact speakers. Is this a good or bad thing completely depends on individual tastes.

On the other hand, my surround improvement is vastly improved, partly due to the improvement in imaging because now all the speakers have very similar frequency response. I've only heard this degree of surround sound coherence all the way around in systems that have properly placed speakers that are all identical, which is very rare.




BINGO! You just nailed it on the head of what Audyssey does. It essentially changes the speakers response so that they have similar timbre matching and thus a more pleasing presentation. If possible, I always profess to use the exact same speakers for every channel, and one of the reasons that bookshelves are the way to go for any complete speaker package amongst a few other reasons.

This may be the reason you turn Audyssey off Chris while listening to two channel music as the FR of your front two speakers have been changed in order to better match your surrounds if they are not the same exact speakers. Is this a good or bad thing completely depends on individual tastes.

Hi Shawn,

I don't entirely follow. Isn't Audyssey attempting to match the individual speaker frequency responses to the SAME target curve, rather than some combined response of mains to surrounds?

Chris, I hesitate to ask this of an audiophile, with cultivated tastes in how things should sound, but have you considered upconverting your two-channel to multichannel with Audyssey MultEQ on? I don't have a trained ear, but to me, in my home theater, it usually sounds better due to improved envelopment.

Larry

Hi Shawn,

I don't entirely follow. Isn't Audyssey attempting to match the individual speaker frequency responses to the SAME target curve, rather than some combined response of mains to surrounds?

Yes it should target a flat curve, more or less. You can choose when doing PRO what version of the curve you want, and audyssey's curve isn't strictly flat, it has a slight rolloff, and you have choices that they suggest based on room size. With the newer software you can also manually target curves that you adjust as well.

I am not under the understanding that it is making any compromises between the responses of various speakers, but attempting to hit "flat" (perceptually flat according to Audyssey's studies) for all the speakers.


Chris, I hesitate to ask this of an audiophile, with cultivated tastes in how things should sound, but have you considered upconverting your two-channel to multichannel with Audyssey MultEQ on? I don't have a trained ear, but to me, in my home theater, it usually sounds better due to improved envelopment.

Larry

No need to hesitate in the least! No, I've never considered doing that. Yes you're right it leads to vastly improved envelopment, but that's not really my goal with stereo. I've never really been a fan of surround music exactly, and I definitely wouldn't and don't prefer to take stereo and expand it to surround. Yes I've from time to time turned on PLIIx, neo6, ultra 2 etc out of curiosity, but I am kind of a purist when it comes to stereo. The best by far I've ever heard from stereo sources has been left in stereo. But that's just me, and of course it assumes me sitting in the sweetspot. Sometimes if I have people over listening to music sure I'll throw it into PLIIx or the like, but never for serious listening I've never really cared for it.

It's very interesting to turn on PLIIx with vinyl actually, you get some very interesting results sometimes. You get a kind of "whoa that's neat" at first, but again, I always end up going back to plain stereo. I currently don't use a subwoofer either for stereo listening, I've gone back and forth with that choice, but until I perhaps get a better sub, and two of them to even out the response in my room, I've kind of arrived at leaving the sub off with some exceptions.

I don't really have an opinion on whether I would prefer or not prefer deriving a center channel from stereo, and just have 3.0/3.1 across the front, because I would demand that the front center be the same as the front L/R and at the same height, something that isn't true in my system. So I don't really hold an opinion on that, except to say that I'm skeptical but still open to what that might sound like. There is something that is simply stunning about a really good stereo image from just two speakers, something that I've never heard matched, let alone exceeded, from a matrixed surround version played back across more speakers. I'm open to that possibility, but I've just never heard it, and it's not better in my system so until things change or I hear a system that proves me wrong, I leave it in stereo.

Hi Shawn,

I don't entirely follow. Isn't Audyssey attempting to match the individual speaker frequency responses to the SAME target curve, rather than some combined response of mains to surrounds?

Yes it should target a flat curve, more or less. You can choose when doing PRO what version of the curve you want, and audyssey's curve isn't strictly flat, it has a slight rolloff, and you have choices that they suggest based on room size. With the newer software you can also manually target curves that you adjust as well.

I am not under the understanding that it is making any compromises between the responses of various speakers, but attempting to hit "flat" (perceptually flat according to Audyssey's studies) for all the speakers.


Hi Chris,

My point is that once any version of MultEQ is engaged, this includes the Pro version, that it always uses the same curve in attempting to fit the response of each speaker. That curve may have a roll-off, it may be flat, or it may be customized in the case of the Pro version, but it will always be the same target curve for all speakers.

As you acknowledge, as far as I have read, there have not been any descriptions of matching the responses of the mains to match the surrounds. It is modifying the response of all speakers to match the same target curve.

Larry

Hi Chris,

My point is that once any version of MultEQ is engaged, this includes the Pro version, that it always uses the same curve to in attempting to fit the response of each speaker. That curve may have a roll-off, it may be flat, or it may be customized in the case of the Pro version, but it will always be the same target curve for all speakers.

As you acknowledge, as far as I have read, there have not been any descriptions of matching the responses of the mains to match the surrounds. It is modifying the response of all speakers to match the same target curve.

Larry

Right, I understand that we have an understanding that we agree about our understandings. :)

Yes you're right it leads to vastly improved envelopment, but that's not really my goal with stereo. I've never really been a fan of surround music exactly, and I definitely wouldn't and don't prefer to take stereo and expand it to surround. Yes I've from time to time turned on PLIIx, neo6, ultra 2 etc out of curiosity, but I am kind of a purist when it comes to stereo. The best by far I've ever heard from stereo sources has been left in stereo. But that's just me, and of course it assumes me sitting in the sweetspot. Sometimes if I have people over listening to music sure I'll throw it into PLIIx or the like, but never for serious listening I've never really cared for it.

Hi Chris,

Aside from differences in personal preferences and training (I don't have any :o), differences in the liveliness of our rooms might also contribute to our different listening experiences. For example, my home theater measures a low reverberation time. When I listen to stereo content not only do I lose envelopment I also lose a sense of spaciousness, a feel for the size of the room. No doubt your room is livelier and the reflections help furnish that sense of spaciousness.

Larry

This may be the reason you turn Audyssey off Chris while listening to two channel music as the FR of your front two speakers have been changed in order to better match your surrounds if they are not the same exact speakers. I really doubt that this is what is happening. However, it is testable by running an Audyssey calibration on a system with only the front speakers active.

I really doubt that this is what is happening. However, it is testable by running an Audyssey calibration on a system with only the front speakers active.


I respect your opinion Kal, but I think it is happening. Audyssey also corrects above the transition frequency (all those hundreds of little corrections throughout the frequency response spectrum). The only reason to do this, other than a bad speaker, is for timbre matching and dealing with boundary issues. I am not saying this is a bad thing, but when calibrating manually, I generally leave everything above the transition frequency alone, meaning I don't use equalization on anything but the sub(s) and deal with SBIR issues through positioning and treatment unless there are issues with the FR above it or for a few other perhaps preferable aspects.

It is very testable. In order to test it, you would need to run nearfield tests on the speakers before and after the Audyssey cal.

I respect your opinion Kal, but I think it is happening. Audyssey also corrects above the transition frequency (all those hundreds of little corrections throughout the frequency response spectrum). The only reason to do this, other than a bad speaker, is for timbre matching. I am not saying this is a bad thing, but when calibrating manually, I generally leave everything above the transition frequency alone, meaning I don't use equalization on anything but the sub(s) and deal with SBIR issues through positioning and treatment.

It is very testable. In order to test it, you would need to run nearfield tests on the speakers before and after the Audyssey cal.

Hi Shawn,

Yes, Audyssey is making corrections above the transition frequency in an attempt to match each speaker's corrected frequency response to the selected Audyssey target curve. However, this does not mean that the main speaker frequency responses are somehow being morphed to match the surround speaker's response. It just means that all satellite speaker's responses are being adjusted to be closer to a common target curve.

If the result of these high frequency corrections is to achieve closer timbre matching of all satellite speakers why would that be a problem?

Why would your test have to be performed in the nearfield? The listener, Chris in this situation, could have been sitting in either the nearfield or reverberant field in his primary listening location when making these comparisons. When he switches between Audyssey on and off, there will of course be a difference in responses, but that doesn't mean the mains will take on the frequency response of the surrounds. It simply means that that they will be closer to the common target curve. Likewise with all other satellite speakers including the surrounds, they too will be closer to the target curve, hence closer to timbre matching.

So when Chris turns on Audyssey, with either two channel or multichannel content, this results in a flatter frequency response in all active speakers, and conversely turning it off Audyssey results in bumpier responses. But turning Audyssey on doesn't mean the main speakers are now emulating the inherent response of the surrounds, but rather the just the common target curve.

Thanks.

Larry

Can Audessey be used in tandem with DEQX? Or will they interfere with eachother?

Also, are there any stand alone room correction units for people who own processors without Audessey-like correction software?

I do realize that I literally hijacked this thread, so I will stop here before I make the OP really mad!

I posed this question by email to both Auddyssey and Trinnov and both gave me the same answer. They said that it was not recommended to layer eq systems and both said to disengage other eq systems before using their products.

If the result of these high frequency corrections is to achieve closer timbre matching of all satellite speakers why would that be a problem?


It's not, unless you mind the FR of the speaker being changed.


Why would your test have to be performed in the nearfield?


Because you are trying to test whether or not the actual FR of the speaker itself has been changed from its original design. Placing a mic at the listening position will result in boundary effects to work into the FR and that is not what you are testing.

It's not, unless you mind the FR of the speaker being changed.




Because you are trying to test whether or not the actual FR of the speaker itself has been changed from its original design. Placing a mic at the listening position will result in boundary effects to work into the FR and that is not what you are testing.


Well of course it would change. That's the point of EQing. If it's not changing the frequency response, then you've got a box with pretty lights on it and nothing more.

It's not, unless you mind the FR of the speaker being changed.This is an interesting point. What is the FR of the speaker? Is it the manufacturer's advertised or, even, measured FR under anechoic conditions? Is it what you measure/hear with the mic up close to the speaker or is it what you measure/hear at the listening position? In view of the influence of the acoustics of the space on the FR (for good or ill), which is the inviolate reference that must not be changed?

This is an interesting point. What is the FR of the speaker? Is it the manufacturer's advertised or, even, measured FR under anechoic conditions? Is it what you measure/hear with the mic up close to the speaker or is it what you measure/hear at the listening position? In view of the influence of the acoustics of the space on the FR (for good or ill), which is the inviolate reference that must not be changed?

I should think it would be the nearfield environment (that being the response without effects of the room) of the "actual" speaker you have in your room provided there are no serious fequency response problems with the speaker itself would you not agree? Isn't that what we are striving for at the listening position itself minus the room effects and the natural rolloff due to the attenuation of high frequency energy with distance? Isn't that WHY we are calibrating in the first place, and when did I say it it can never be violated? I just said if you don't mind it being changed. Some do, some don't.

DynamicEQ is interesting. Not so much for what it does, but what it means. Remember all those people complaining of lean bass? It turns out the well known characteristic of human hearing where bass needs to be louder at lower listening levels to be perceived the same as it is at reference wasn't included as part of the original Audyssey correction algorithm. DynamicEQ fixes that.

A whole group of naysayers were outright dismissed as bass heads that didn't like reference sound, but the introduction of DynamicEQ means those same people were actually right. They were hearing flaws in Audyssey's implementation (below reference level).

Is it possible something similar is happening with 2 channel? Could some of the "one size fits all" improvements tailored towards multichannel do more harm than good when applied to 2 channel?

For example maybe 2ch sounds better with Audyssey off because their algorithm reels in the decay time. With multichannel, the reverb is captured in the surround channels so removing decay improves clarity. With 2ch reducing the decay by the same amount may be perceived as sucking the life out the presentation making it sound overly dry & clinical.

Or it could be different mixing standards. I find it ironic that Audyssey's "Reference" curve isn't flat. Reference sound is thrown around like it's an industry standard, but it's just Audyssey's proprietary curve they think makes all your music and movies sound "best". You don't like the high frequency content removed from your CD's... well you just don't like reference sound, son.

My impression of the Pro pack is it helps bring Audyssey back in line as a tool instead of magic bullet. Only bad thing about the Pro pack for consumers is it's tied to the MAC address of your processor.

Audyssey will tell you there is NO two channel reference. There are a set of standards for movies that the director may or may not follow and it is that standard to which Audyssey strives.

Having installed hundreds of two channel digital room correction systems (not Audyssey), in 99% of the cases, there was no harm done. In a very few of the cases the improvement may or may not have justified the expense but in every other case, there were major improvements. And the listening environments were from very high end speakers in family rooms (no treatment) to dedicated listening rooms with appropriate treatment.

In virtually all of these cases, the digital signal from the CD transport went to the correction system and then was converted to analog either by our product or an external DAC.

Audyssey or even Audyssey Pro will work perfectly fine for most (and that is a key word) environments but in those cases where it does not work well, there may not be enough flexibility even in Audyssey Pro to make it work. And I can assure you that the flexibility required to adapt the product to those more complex room situations implies product complexity and most folks don't really have the acoustic (or mathematical) knowledge to properly use such a system.

As I have stated previously, there are more flexible products on the market that cost 10 times more.

As to what an appropriate target curve might be, there are a couple of approaches. The one we used that provided the best results was to design a curve that had the general shape of the FR at the listening position without correction --- and with the knowledge (from manufacturers spec) of what the low end extension was.Audyssey determines low end cut off based upon the 3db down spot which, in some cases is NOT the correct low end extension point if there happens to be an anomolie in the room that could create that size null. As an example, all Dunlavy speakers have about a 4db roll-off from 10K to 20K at the listening position (not in an anechoic chamber) and we built that roll-off into our target curve. When we used a flat target curve, the sound would become fatiguing with some material.

Audyssey for the price and functionality is a no brainer. If it doesn't work in your environment, there are other much more expensive solutions.

Rabident...I agree that the DynamicEQ idea was, and is, actually very sound and is based on solid scientific fact.

I found the implementation didn't live up to the promise. I did find Dolby Volume to be clearly superior. There are some caveats to that, the "music" algorithm being one.

I should think it would be the nearfield environment (that being the response without effects of the room) of the "actual" speaker you have in your room provided there are no serious fequency response problems with the speaker itself would you not agree? Isn't that what we are striving for at the listening position itself minus the room effects and the natural rolloff due to the attenuation of high frequency energy with distance? Thanks for taking the bait. :)

That near-field FR is something the listener was not supposed to hear and that we never listen to. (Well, some might.) Nearfield responses may not benefit from the proper blending of multiple drivers. Speakers are designed to be listened to in an acoustical environment which, with consequent difficulty, is not fixed or defined. I am not trying to be critical you or of any particular response but to raise for discussion the question of what is our goal. For me, it is the correct FR at my listening position, taking into account all the equipment, speakers and room effects, such that it results in as accurate as possible recreation of the original event. IMHO, nothing after the actual performance is completely free of human manipulation and, therefore, is not sacrosanct.

Isn't that WHY we are calibrating in the first place, and when did I say it it can never be violated? I just said if you don't mind it being changed. Some do, some don't.Sure but I do not see why there should be an issue at all.

For me, it is the correct FR at my listening position, taking into account all the equipment, speakers and room effects, such that it results in as accurate as possible recreation of the original event. IMHO, nothing after the actual performance is completely free of human manipulation and, therefore, is not sacrosanct.



What do you believe the target curve to be to achieve "correct FR" at the listening position? How do you know that the target you chose will give you an "accurate recreation of the original event" if there is no standard? When measuring, how much wieght is put on direct vs reflected sound?

I posed this question by email to both Auddyssey and Trinnov and both gave me the same answer. They said that it was not recommended to layer eq systems and both said to disengage other eq systems before using their products.

Would you mind copy-pasting that response in the thread?

"It turns out the well known characteristic of human hearing where bass needs to be louder at lower listening levels to be perceived the same as it is at reference wasn't included as part of the original Audyssey correction algorithm. DynamicEQ fixes that.

A whole group of naysayers were outright dismissed as bass heads that didn't like reference sound, but the introduction of DynamicEQ means those same people were actually right. They were hearing flaws in Audyssey's implementation (below reference level)."

That doesn't wash - there was no DynamicEQ before they used Audyssey, nor does it seem to be a complaint w/other EQ systems.

What do you believe the target curve to be to achieve "correct FR" at the listening position? How do you know that the target you chose will give you an "accurate recreation of the original event" if there is no standard? There are several answers. First, the target curve depends on the presumptions of the mastering engineers and, while there is some consensus with movies, there is none with music recordings. Second, in view of that, I can only try to make the response through my system flat (or, for movies, a bit rolled off) to minimize its influence. Third, your very questions support my argument that those who think that changing the FR is a violation of principle are ignoring all the variable influences encumbered by a home system/room.

When measuring, how much wieght is put on direct vs reflected sound?That is hard to encompass in a single measurement and is greatly influenced by the speaker design. I like wide-dispersion speakers in an acoustically treated room but I have no quantitative information to offer you.

There are several answers.


That is hard to encompass in a single measurement...

Ah, well, there it is. Unlike the fiction of Neo in The Matrix or other fantasized universal solutions in movies, literature, and religion about "The One," there is no unity of perfection in the studio or sound reproduction realm. In fact, the idea that a single standard of perfection can be attained through science and technology is rendered moot by the very fundamental, unscientific, mystic roots of faith (or gullibility) necessary for perfection to exist as a concept.

We have amazing tools that are more sophisticated with each successive generation that bring us both closer to the ideal and also to the realization that the ideal is far too relative to ever be unified. Neither my ideal nor my judgement/belief on perfection will ever be the same as someone else's, and vice-versa.

In fact, for some folks, the more the masses embrace something as being ideal (often meaning "good enough"), the more those folks who want to be "special" or who have very high expectations seek to distance themselves and their aspirations from those of the masses. That's one reason why snake oil products enjoy such celebrity in some segments of the audio world.

It's all compromise in the end. "How much compromise is enough?" is the real question. Let me pull some figures out of my @$$ here.

50% think MP3s players, car radios, DVDs, TVs, and all-in-one stereos are good enough.
25% think CDs, DVDs, flat panel TVs, surround receivers, and HTIB are good enough.
10% think CDs. DVDs, 720p flat panels displays, high end receivers with auto EQ, and big box retail surround speaker systems are good enough.
5% think SACD, DVD-A, Bluray, 1080p flat panels, pre/pros with auto EQ, separate amps, bass traps, big name brand front speakers and bookshelf surrounds are good enough.
5% think vinyl and stereo (no video) are good enough.
3% think SACD, DVD-A, Bluray, 1080p flat panels, pre/pros with auto EQ, separate amps, some passive room treatment (drapes, carpet, maybe bass traps) big name brand front speakers and name brand bookshelf surrounds are good enough for HT ,and vinyl and stereo are good enough for music.
1% don't listen to anything.
.75% think that complete, custom-designed, multi-channel, high end solutions in dedicated and treated rooms are good enough.
.25% are on forums like these, shockingly close to the realization that it may never be good enough.

So, I'd postulate that Audessey PRO and other fairly sophisticated approaches far exceed the expectations of 90%+ of the available audience, and that it will never meet the expectations of at least 1%.

As for me, I'm PDH with my little Synthesis® set up and calibration, and I believe that I'd be thrilled with a great Audessey PRO set up, too.

For those who aren't yet satisfied, you just have to step up and pay for the next more expensive solution and keep spending until you're broke or happy (or both). If you find it, I don't want to come over and hear it, thus spoiling my current stasis.:p

Well, okay, maybe just to listen once... :D

DynamicEQ is interesting. Not so much for what it does, but what it means. Remember all those people complaining of lean bass? It turns out the well known characteristic of human hearing where bass needs to be louder at lower listening levels to be perceived the same as it is at reference wasn't included as part of the original Audyssey correction algorithm. DynamicEQ fixes that.

A whole group of naysayers were outright dismissed as bass heads that didn't like reference sound, but the introduction of DynamicEQ means those same people were actually right. They were hearing flaws in Audyssey's implementation (below reference level).


I used to wonder why a some audiophiles that did use equalization preferred a house curve that wasn't flat, but rather boosted at both ends of the frequency spectrum, the so-called smiley curve. Then it occured to me that, via trial and error, they were implementing a crude static version of Dynamic EQ when listening at levels lower than the mixing environment. In other words, the house curve was approximately perceptually flat when hearing's decreased sensitivity to lows and highs are compensated for via equalization.

The issue of equalizing to a mostly flat target curve (with or without a slight roll-off), but listening at levels different from the mixing levels, is not a flaw only in Audyssey correction, it is a flaw in virtually all equalization methods. Audyssey I believe is unique in that it boosts bass, highs and surround levels to compensate for loss in sensitivity at lower listening levels. Further, it does it dynamically by monitoring in real-time the actual levels of the content adjusting for the quieter passages, something that conventional equalization simply can not do.


Is it possible something similar is happening with 2 channel? Could some of the "one size fits all" improvements tailored towards multichannel do more harm than good when applied to 2 channel?

For example maybe 2ch sounds better with Audyssey off because their algorithm reels in the decay time. With multichannel, the reverb is captured in the surround channels so removing decay improves clarity. With 2ch reducing the decay by the same amount may be perceived as sucking the life out the presentation making it sound overly dry & clinical.


I'm inclined to believe Kal's previous explanation:


I know this will be a "hot button" for some but many long-term audiophiles have become accustomed to a particular rendition of stereo sources based on years of listening in environments where the room acoustics contribute significantly to the overall soundstage and imaging (to say nothing of FR). Indeed, successful stereo often relies more heavily on the room's contribution even though it may be spurious in many ways.


As I mentioned earlier, if a room is acoustically designed for listening to multichannel content and two-channel content is played, it will sound dry, but that is not the fault of equalization, that is the fault of playing the wrong content in the wrong room. ;)


Or it could be different mixing standards. I find it ironic that Audyssey's "Reference" curve isn't flat. Reference sound is thrown around like it's an industry standard, but it's just Audyssey's proprietary curve they think makes all your music and movies sound "best". You don't like the high frequency content removed from your CD's... well you just don't like reference sound, son.

With regard to why Audyssey has a target curve that rolls off, it is because their research shows that the majority of home listening environments is in the reverberant field not the near field. Therefore, for film standards , the roll-off is needed to tame brightness caused by the differences between the recording and listening acoustic environments. For folks who are seated in the near field at home, the Flat target curve may be more appropriate. However, a great deal of music is mixed much hotter than film standard reference, so we may find that regardless of whether we are seated in the near field, gentle high frequency roll-offs may still be needed to listen at moderate levels at home in a typically small room without fatigue.

Audyssey makes no claim that for CDs, mixed to no specific standard, that Dynamic EQ will make them sound better. It may or it may not. Its not that Audyssey claims the customer doesn't like reference, there is no reference for Dynamic EQ to adhere to for music. If it sounds bad, they advise to turn off Dynamic EQ. Some new models permit different levels of Dynamic EQ to compensate for music mixes that are hotter than film standards. Another workaround for devices that don't permit adjustments is to decrease the input source levels. This cause the user to manually turn up the master volume to get to the same comfortable listening level, which in turn causes Dynamic EQ to reduce its corrections thinking it is closer to reference.


My impression of the Pro pack is it helps bring Audyssey back in line as a tool instead of magic bullet.

Yes, for example, each room and primary listening location within the room, is likely to need a different amount of roll-off, than the standard Audyssey Reference curve, that is, if it needs a roll-off at all. So the Pro software offers a bit more adjustability to more closely tailor the roll-off to the particular acoustic environment.

Larry

Audyssey will tell you there is NO two channel reference. There are a set of standards for movies that the director may or may not follow and it is that standard to which Audyssey strives.

Having installed hundreds of two channel digital room correction systems (not Audyssey), in 99% of the cases, there was no harm done. In a very few of the cases the improvement may or may not have justified the expense but in every other case, there were major improvements. And the listening environments were from very high end speakers in family rooms (no treatment) to dedicated listening rooms with appropriate treatment.

In virtually all of these cases, the digital signal from the CD transport went to the correction system and then was converted to analog either by our product or an external DAC.

Audyssey or even Audyssey Pro will work perfectly fine for most (and that is a key word) environments but in those cases where it does not work well, there may not be enough flexibility even in Audyssey Pro to make it work. And I can assure you that the flexibility required to adapt the product to those more complex room situations implies product complexity and most folks don't really have the acoustic (or mathematical) knowledge to properly use such a system.

As I have stated previously, there are more flexible products on the market that cost 10 times more.

As to what an appropriate target curve might be, there are a couple of approaches. The one we used that provided the best results was to design a curve that had the general shape of the FR at the listening position without correction --- and with the knowledge (from manufacturers spec) of what the low end extension was.Audyssey determines low end cut off based upon the 3db down spot which, in some cases is NOT the correct low end extension point if there happens to be an anomolie in the room that could create that size null. As an example, all Dunlavy speakers have about a 4db roll-off from 10K to 20K at the listening position (not in an anechoic chamber) and we built that roll-off into our target curve. When we used a flat target curve, the sound would become fatiguing with some material.

Audyssey for the price and functionality is a no brainer. If it doesn't work in your environment, there are other much more expensive solutions.

Hi,

You make a lot of excellent points.

I have skimmed over some Audyssey Pro literature and it appears they offer a limited number of target curves and they can only be adjusted by +/- 3 dB.

Yes, Audyssey admits that theoretically if a modal peak has the right characteristics it might fool MultEQ into calculating an erroneous -3 dB point, but this is a rare occassion.

Just one point, in the current devices Audyssey offers a suite of software, MultEQ in its various forms, including the Pro version, Dynamic EQ and Dynamic Volume. Many folks feel that Dynamic EQ/Volume is as important to their enjoyment as MultEQ. So even if another equalization approach is more flexible, it will not offer features that dynamically adjusts the equalization to yield a perceptually flat response when listening a levels lower than the mixing environment.

Larry

...and they can only be adjusted by +/- 3 dB.That's not really true. According to the manual (pages 31-32), once cannot set a point that is more than +/- 3dB from 0dB but, as in the example shown, one can move a point on one of the set curves that is outside that window to any point within the window.

That's not really true. According to the manual (pages 31-32), once cannot set a point that is more than +/- 3dB from 0dB but, as in the example shown, one can move a point on one of the set curves that is outside that window to any point within the window.

Hi Kal,

Thanks for taking the time to explain this to me in the Audyssey thread (http://www.avsforum.com/avs-vb/showpost.php?p=17344140&postcount=19527). :D

However, I suspect some of the professionals here may disagree that being limited to +/- 3 dB on the entire Flat target curve and to +/- 3 dB in the portion below ~12 kHz on the Reference target curve offers enough flexibility for them. ;)

Thanks again for the clarification.

Larry

Hi Kal,

Thanks for taking the time to explain this to me in the Audyssey thread (http://www.avsforum.com/avs-vb/showpost.php?p=17344140&postcount=19527). :D

However, I suspect some of the professionals here may disagree that being limited to +/- 3 dB on the entire Flat target curve and to +/- 3 dB in the portion below ~12 kHz on the Reference target curve offers enough flexibility for them. ;)

Thanks again for the clarification.

LarryMebbe. However, this is not a correction curve but a target and I have no situations in which I would need more than that 6dB range.

BTW - Do any of you Audyssey Pro users know what the vertical scale is in the graphs below?

http://i80.photobucket.com/albums/j193/ptwalker/measurements/MAGSAudysseygraph.jpg

Is it 3 or 5dB between each horizontal line?

Thanks for taking the bait. :)

What bait???? You have to remember Kal that speakers are generally flat in the near-field with jut a slight boost in the higher frequencies. So, wouldn't you naturally be able to recreate the mixed recording just as the engineer intended it (not necessarily the actual recording) provided you removed the effects of the room? What are you then left with? I can see this is going nowhere really fast.:rolleyes:

That doesn't wash - there was no DynamicEQ before they used Audyssey, nor does it seem to be a complaint w/other EQ systems.

Other EQs have enough flexibility to deal with the problem, at least for a target volume level. If listening less than reference level, bass can be manually boosted to compensate.

Gen 1 Audyssey came later, initially to most users as a closed, proprietary, all or nothing option. Calibration was done to reference levels and didn't allow any further tweaking. Users coming from other EQ's with manually applied low frequency compensation experienced bass suck-out with Audyssey.

Gen 2 Audyssey's introduction of DynEQ acknowledges the problem and provides a solution.

Other EQs have enough flexibility to deal with the problem, at least for a target volume level. If listening less than reference level, bass can be manually boosted to compensate.


Hi,

Yes, there is an issue, but as I mentioned earlier, the problem associated with listening below reference levels is not strictly an Audyssey problem.

When listening at below reference level yes, hearing is less sensitive to bass, but it is also less sensitive to treble and to sounds coming from the rear. Furthermore, the levels of recordings do not stay at the same volume throughout, they can vary greatly. Other EQs that allow the user to manually boost the bass (or the treble) by a fixed amount, are incomplete static solutions that do not provide a perceptually flat response as the content's levels change, nor do they dynamically adjust the levels of sounds coming from the rear.

Larry

Audyssey also corrects above the transition frequency (all those hundreds of little corrections throughout the frequency response spectrum). The only reason to do this, other than a bad speaker, is for timbre matching and dealing with boundary issues. I am not saying this is a bad thing, but when calibrating manually, I generally leave everything above the transition frequency alone, meaning I don't use equalization on anything but the sub(s) and deal with SBIR issues through positioning and treatment unless there are issues with the FR above it or for a few other perhaps preferable aspects.

Hi Shawn,

On the advisability of equalizing above the transition frequency I doubt we'll ever reach a consensus, but I think its healthy to exchange differing views. :D

Here are some remarks by Audyssey on the subject.

"Official" Audyssey thread, posting #19358. (http://www.avsforum.com/avs-vb/showpost.php?p=17336944&postcount=19458)

"Official" Audyssey thread, posting #19469. (http://www.avsforum.com/avs-vb/showpost.php?p=17338131&postcount=19469)

"Official" Audyssey thread, posting #19480. (http://www.avsforum.com/avs-vb/showpost.php?p=17340451&postcount=19480)

"Official" Audyssey thread, posting #19536. (http://www.avsforum.com/avs-vb/showpost.php?p=17344862&postcount=19536)

Larry

How come? I find my TacT to do an incredible job on both stereo and 5.1. Indeed the effect for stereo is even stronger.

Hi Amir,

I'm not familar with the TacT.

Does it equalize above the transition frequency?

I would value your opinion on the pros and cons of equalizing above the transition frequency with these sophisticated software approaches in contrast to manual parametric equalization.

Thanks.

Larry

BTW - Do any of you Audyssey Pro users know what the vertical scale is in the graphs below?

Is it 3 or 5dB between each horizontal line?

I had the same question and asked Audyssey. The answer is none of the above. The graphs are only a general indication of the before and after and not detailed and specific plots. That said, they are still informative.

Does it equalize above the transition frequency?

I would value your opinion on the pros and cons of equalizing above the transition frequency with these sophisticated software approaches in contrast to manual parametric equalization.
TacT provides full-frequency equalization, though the user can choose to equalize only frequencies below any chosen frequency. It equalizes to a flat FR, but has 12 user-selectable "house" curves, as well as the ability to tweak these curves to suit individual preference.

In my case, I'm using a 2ch-only version in a dedicated stereo setup so I can't comment on 5.1. I gave up a fairly high-end analog preamp for the TacT, and I've been extremely impressed with the results.

Edited: User can choose to EQ only frequencies below a chosen point, but may not choose to EQ only frequenceis above a chosen point.

Hi Amir,

I'm not familar with the TacT.

Does it equalize above the transition frequency?

I would value your opinion on the pros and cons of equalizing above the transition frequency with these sophisticated software approaches in contrast to manual parametric equalization.

Thanks.

Larry

There are two Tact product types. The 2.0 is a two channel (no sub) digital room correction product (which is what I own) that operates in the time domain. Their other room correction products (2 channel and 10 channel) fix (it is NOT frequency based EQ) above the transition point. They all use time analysis (at 1 or more measuring spots) and you can set crossover, slope, etc and they will time align the subs to the mains.

It has a LOT of flexibility as it should given it (the two channel version with sub support) costs almost 10 times the cost of Audyssey Pro. A user must have some reasonable knowledge of room acoustics so as not to blow up drivers, burn crossover, etc (I have done both when I first learned to us it).

It is for that reason that the Audyssey Pro solution is so wonderful for the masses as it may be impossible to actually "break" anything and, for most people, get great results.

TacT provides full-frequency equalization, though the user can choose to equalize only frequencies above or below any chosen frequency. It equalizes to a flat FR, but has 12 user-selectable "house" curves, as well as the ability to tweak these curves to suit individual preference.

In my case, I'm using a 2ch-only version in a dedicated stereo setup so I can't comment on 5.1. I gave up a fairly high-end analog preamp for the TacT, and I've been extremely impressed with the results.

Hi,

Thanks for the response.

So you equalize the full frequency range and are impressed with the results?

Larry

There are two Tact product types. The 2.0 is a two channel (no sub) digital room correction product (which is what I own) that operates in the time domain. Their other room correction products (2 channel and 10 channel) fix (it is NOT frequency based EQ) above the transition point. They all use time analysis (at 1 or more measuring spots) and you can set crossover, slope, etc and they will time align the subs to the mains.

Hi,

To clarify, your version doesn't equalize above the transition frequency?

Thanks.

Larry

What bait???? Just keeeding.

You have to remember Kal that speakers are generally flat in the near-field with jut a slight boost in the higher frequencies. So, wouldn't you naturally be able to recreate the mixed recording just as the engineer intended it (not necessarily the actual recording) provided you removed the effects of the room? What are you then left with? I can see this is going nowhere really fast.:rolleyes:Depends on how you are using the term near-field. As used for listening, it is within the point where room effects begin to dominate. That does not mean that they are not affecting the sound or the measurements. For measurement, it is so close that room effects are not significant. However, the speaker was likely not designed to be listened to at that distance. So, I agree that this is going nowhere and that was my intent. There is no way to use these arguments to justify abjuring EQ.

Hi,

Thanks for the response.

So you equalize the full frequency range and are impressed with the results?

Larry
Yes I EQ the full FR using their 2.2XP "Stereo Processor"/preamp. So impressed I don't think I can go back to a non-EQ product. For completeness sake, I have a dedicated 2ch room which is fairly well treated with conventional absorbtion/diffusion products.

Hi,

To clarify, your version doesn't equalize above the transition frequency?

Thanks.

Larry

There is no "transition point" for my product. Mine is the 2.0. So I correct full spectrum. The 2.2 which allows for correction of a separate sub(s) is what you are referring two and if I were to ever switch to a sub supported two channel system, that is what I would purchase. (I will not use Audyssey for two channel as the analog section of my pre-pro is not good enough for my listening requirements and Audyssey Pro does not let me do some things that I believe are mandatory for a true high-end two channel listening environment.)

Let me tell you about a huge potential advantage of correcting the entire frequency spectrum in a two channel environment. We all know about fixing bass problem caused by room boundaries but even in a perfectly symmetrical room, the chance of the left and right impulse response (and hence FR) being identical are pretty slim (in my experience of doing hundreds of rooms --very close to ZERO). To the degree the difference increase, the benefits of correction in the mids and highs increases. What you will get with room correction done correctly is (among other things) increased image focus and stability. Assume the uncorrected room response at, say 2K is +1.5 db on the left channel and -3 db on the right channel (not at all unusual). As musical content containing that frequency plays, the image will gravitate between left and right (maybe very slightly but clearly perceptible) and hence cause image shift. By correcting that FR deviation, the image becomes more stable. Trust me that this is not theoretical but does, in fact, occur. (Maybe the selection of 2K is not accurate but the point remains). Obviously in a multichannel system, this specific benefit somewhat is reduced but not completely.

I have been in a few rooms where room correction above the bass (and in one example including the bass) did absolutely nothing. But that is the exception.

There is no "transition point" for my product.

Hi,

Thanks for your very informative response.

With regard to the transition frequency, I was referring to the Schroeder Transition frequency for your room.

f = 2000*(sqrt T/V)

Where T is the reverberation time in seconds, and V is the room volume in cubic meters.

As I understand it, below the transition frequency acoustics are dominated by relatively isolated, orderly room modes while above the frequency acoustics are dominated by complex overlapping room modes and reflections.

I don't have a good grasp of the details, but apparently the controversy about the suitability of equalization above the transition frequency seems to hinge on the fact that wave acoustics (standing wave models) used to predict and correct modal responses at low frequencies doesn't work at high frequencies where the modes are not isolated. So another approach to room correction is needed in place of conventional equalization.


Obviously in a multichannel system, this specific benefit somewhat is reduced but not completely.

Perhaps its not so obvious to me. ;)

I would think that in a multichannel system there would be a much greater likelihood that the timbre of the rear speaker array differed from the front speaker array. Much more so than any difference likely to exist between "identical" paired two-channel mains. Therefore, equalization that matches the timbre of multichannel systems, front to back might be extremely beneficial. The only offsetting factor would be that our forward facing hearing sensitivity is greater than our rear, so we are more likely to hear timbre mismatches up front.

Larry

BTW - Do any of you Audyssey Pro users know what the vertical scale is in the graphs below?

Sorry if I missed it, but what is that notch in upper midrange / lower treble that appears in the post-Audyssey charts? Is that a BBC dip?

Sorry if I missed it, but what is that notch in upper midrange / lower treble that appears in the post-Audyssey charts? Is that a BBC dip?

Hi,

Here's an excerpt from some Audyssey Pro documentation.


It is recommended that you engage the Midrange Compensation option for your first listening test (this is selected by default). Midrange compensation is sometimes necessary to correct for the directivity differences that often occur in that frequency range due to the crossover circuitry or horn-loaded speakers. You can later return to this screen and disengage this option if you prefer.


Larry

Other EQs that allow the user to manually boost the bass (or the treble) by a fixed amount, are incomplete static solutions that do not provide a perceptually flat response as the content's levels change, nor do they dynamically adjust the levels of sounds coming from the rear.

Right, but 1st generation Audyssey has the same short comings. Some people complained bass was too low at low volume.

Audyssey publicly dismissed it as preference, while privately working on DynEQ to address the problem.

Right, but 1st generation Audyssey has the same short comings.

Hi,

I believe I said that. :D

Yes, there is an issue, but as I mentioned earlier, the problem associated with listening below reference levels is not strictly an Audyssey problem.



The issue of equalizing to a mostly flat target curve (with or without a slight roll-off), but listening at levels different from the mixing levels, is not a flaw only in Audyssey correction, it is a flaw in virtually all equalization methods.

Some people complained bass was too low at low volume.

Audyssey publicly dismissed it as preference,...


It was a preference issue. Folks preferred to have a perceptually flat response. Too bad the early Audyssey products, and current conventional equalization methods don't provide it. ;)

Larry

Right, but 1st generation Audyssey has the same short comings. Some people complained bass was too low at low volume.

Audyssey publicly dismissed it as preference, while privately working on DynEQ to address the problem.

I think your characterization of things is a little bit strange here. DynEQ was not to address shortcomings of a first-generation product. Audyssey still works exactly the same. Dynamic EQ is a totally different goal and can be enabled/disabled separately from basic Audyssey EQ.

I think your characterization of things is a little bit strange here. DynEQ was not to address shortcomings of a first-generation product. Audyssey still works exactly the same. Dynamic EQ is a totally different goal and can be enabled/disabled separately from basic Audyssey EQ.

Hi Chris,

I agree that the characterization of Audyssey's motives for developing Dynamic EQ seems to have an undo conspiratorial tone. However, don't you agree that it is technically correct that the earlier Audyssey products without Dynamic EQ (along with other conventional equalization techniques) do not provide a perceptually flat response when listening below the levels the content was mixed?

Larry

Hi,

Here's an excerpt from some Audyssey Pro documentation.



Larry

Thank you. It is indeed a BBC dip! Pretty cool. You should be able to set the frequency that it occurs in case your speakers do not cross over at that region.

Thank you. It is indeed a BBC dip! Pretty cool. You should be able to set the frequency that it occurs in case your speakers do not cross over at that region.

Hi,

Makes sense, but I didn't see that level of adjustability documented in the 3.0 version of the software.

Larry

Thank you. It is indeed a BBC dip! Pretty cool. You should be able to set the frequency that it occurs in case your speakers do not cross over at that region.

Yeah, it is wise of them to have included such a feature, but as you said, it needs to be adjustable, which is isn't. However, you could start with a flat target curve and just add a dip where you want it with the new software that allows the user to manipulate the target curve.

Yeah, it is wise of them to have included such a feature, but as you said, it needs to be adjustable, which is isn't. However, you could start with a flat target curve and just add a dip where you want it with the new software that allows the user to manipulate the target curve.

Hi Tim,

Yes, as long you don't need a dip of more than 3 dB below the target curve that should work.

Larry

Yeah, it is wise of them to have included such a feature, but as you said, it needs to be adjustable, which is isn't. However, you could start with a flat target curve and just add a dip where you want it with the new software that allows the user to manipulate the target curve.

I need to do more reading on the new features...pretty behind on Audyssey.

There was a good discussion about the BBC dip on the Orion forum. Some concluded that it worked well with some recordings and not with others, but generally, the recording sounds more laid back. I suspect my PL100s could greatly benefit from such a dip.

Some people suggest that the BBC dip isn't a speaker correction tool, but a recording correction tool. Therefore, maybe placing the dip at 2k is by design, even if the crossover does not occur in that region.

Here is an excellent summary:
http://orion.quicksytes.com/viewtopic.php?f=1&t=1321&hilit=bbc+dip&start=15#p9811

There was a good discussion about the BBC dip on the Orion forum. Some concluded that it worked well with some recordings and not with others, but generally, the recording sounds more laid back.That would seem to apply to any fixed deviation from flat unless related specifically to a particular speaker and/or system.

I suspect my PL100s could greatly benefit from such a dip.Cannot hurt to try.

...don't you agree that it is technically correct that the earlier Audyssey products without Dynamic EQ (along with other conventional equalization techniques) do not provide a perceptually flat response when listening below the levels the content was mixed?That was happening anyway: lower the volume and the bass seems to drop faster than the rest of the frequencies. Has nothing to do with earlier Audyssey products, since it would happen with or without room correction.

Loudness compensation has been with us for decades. On your previous surround processor, low frequencies were dynamically contoured (using ISO 226 equiloudness curves) based on how far below reference the volume level was. On the next model up, the dynamic loudness compensation could be used along side room correction. Heck, my '80s vintage Yamaha pre-amp had a variable (not static) loudness knob that could be manually adjusted depending on listening level.

Chris is correct about MultEQ and Dynamic EQ being separate technologies, each with a different goal. The former addresses the room's unwanted contributions while the latter compensates specifically for our human hearing. No overlap, except the brand name.

Heck, my '80s vintage Yamaha pre-amp had a variable (not static) loudness knob that could be manually adjusted depending on listening level.

Hi Sanjay,

Yeah, but did the knob twirl itself automatically depending on loud and soft passages of the content? :D

Larry

Heck, my '80s vintage Yamaha pre-amp had a variable (not static) loudness knob that could be manually adjusted depending on listening level.Yeah, but did the knob twirl itself automatically depending on loud and soft passages of the content? :DEr, no. I used the term "manually" in this context to actually mean... well, manually.

Chris is correct about MultEQ and Dynamic EQ being separate technologies, each with a different goal. The former addresses the room's unwanted contributions while the latter compensates specifically for our human hearing. No overlap, except the brand name.

Now it's marketed that way, but when MultEQ was the only technology in their portfolio it was positioned with the goal of producing the reference sound. The desire for more bass wasn't characterized as the desire for "perceptual flatness". They dismissed it as preference for exaggerated bass.

It's only now with the addition of DynEQ that true correction (for humans) is possible. We see that the complaints of lack of bass for some were valid. It was not simply a preference for exaggerated bass. It was due the way we hear and the fact that Audyssey didn't take that into account when originally designing their "reference" sound.

I think your characterization of things is a little bit strange here. DynEQ was not to address shortcomings of a first-generation product. Audyssey still works exactly the same. Dynamic EQ is a totally different goal and can be enabled/disabled separately from basic Audyssey EQ.

Chris,

I don't see how you can see it as anything less than a short coming. The original system was designed for humans. i.e. it wasn't a test tool for speakers to "remove the room" and see how they measured. It was bundled into consumer receivers and intended to correct the sound people heard. The problem was MultEQ was only accurate at reference volume because of the way we perceive sound. DynEQ adds another layer that further adjusts the sound so Audyssey is perceived accurate for all volume levels.

That to me is a system level shortcoming in their first gen technology that was addressed later with DynEQ.

Now it's marketed that way, but when MultEQ was the only technology in their portfolio it was positioned with the goal of producing the reference sound. Which it largely achieved as long as you realize that "reference" includes, also, the playback level.

I agree that the characterization of Audyssey's motives for developing Dynamic EQ seems to have an undo conspiratorial tone.

The tone goes back to my only underlying issue with Audyssey at this point, and that's how "reference vs preference" is thrown around. I understand preference can play a part, but Audyssey and supporters use the phrase to unilaterally dismiss anyone that has a complaint about their current generation offering.

Which it largely achieved as long as you realize that "reference" includes, also, the playback level.

That realization came to me with the introduction of DynEQ. Did you realize when you bought your 9.8 that the Audyssey correction was only going to sound accurate at reference volume level?

That realization came to me with the introduction of DynEQ. Did you realize when you bought your 9.8 that the Audyssey correction was only going to sound accurate at reference volume level?Yes, of course. In fact, when Audyssey announced and demonstrated DEQ, I told Chris I didn't need it. :D

I don't see how you can see it as anything less than a short coming.Because it has absolutely nothing to do with Audyssey room correction. Lowering the volume level will cause a perceptual loss of bass. Always has. Always will. How did this aspect of our human hearing suddenly become a shortcoming of Audyssey room correction? It was bundled into consumer receivers and intended to correct the sound people heard.No, you're starting from a false premise. Audyssey never claimed that their room correction system was "intended" to address anything more than the room's unwanted contributions. That to me is a system level shortcoming in their first gen technology that was addressed later with DynEQ.The only thing that happened "later" is that Audyssey came out with their own brand of loudness compensation. So did THX and Dolby. Were they too addressing shortcomings of their earlier gen technologies? Which ones?

The tone goes back to my only underlying issue with Audyssey at this point, and that's how "reference vs preference" is thrown around. I understand preference can play a part, but Audyssey and supporters use the phrase to unilaterally dismiss anyone that has a complaint about their current generation offering.

Hi,

Fine, I get it. You personally don't like how you feel Audyssey marketed their first generation products. Frankly, I don't see how that has any bearing whatsoever on the topic of this thread.

Larry

I am old enough to remember Tom Holman discussing the loudness compensation circuit in the Advent 300 receiver from 1976.He also designed a really nice phono stage for the same model. Based on that history it is surprising Dynamic EQ or modern day loudness compensation wasn't included from day one with MultiEQ.

Here's a couple of observations by enthusiasts that did their own Audyssey Pro calibration:

After a new round of calibration with Audyssey Pro (left and right set at 50hz) the system is complete and the results are incredible. The vocals and imaging are drastically improved, so much so I can't stand listening with out Audyssey running. I'm an audio guy first and have done all my listening in two channel stereo. I know that if i can get it right with two channel, surround sound will be great as well.

The sound improvements have also extended to the high frequency bringing back sparkle and life to the system. Frankly I'm surprised that I got such a dramatic improvement considering the quality of my main speakers to begin with. Without Audyssey the vocals have an irritating glare and sound veiled. With Audyssey on,the vocals open up and separate from the other instruments, rock solid imaging. Bass differences were more subtle, i noticed a smoothing out of upper bass and noticed deeper extension on the low stuff. Brings me back to the old days of audiophiling, moving speakers, trying various forms of isolation, room treatments etc to get the system to open up and sing. The changes I have experienced are not minor tweaks in resolution they have altered the character of the system. The harmonics of the whole system is sweeter, warmer and detailed.

peter

I got my baptism with Audyssey Pro 3.2 spending a day to study up, install, run, calibrate, and experiement with an Onkyo 886. I've always been a strong advocate of MultEQxt since its inception, and been impressed at how it has evolved. It had gotten so good, that I had become a little skeptical as to how Pro could bring much else to the party. Man, was I wrong.

I used ten measurement positions across/around a four seat sofa in our rather large room. Fate had blessed me with subs being co-located in a good acoustic position, so MultEQ had already done a nice job there. I do have a small acoustic node that can't be fixed with EQ, but it is so minor that I cannot hear it, and the bass sounds great. It was what Pro did for the other channels that made such a big improvement. The level of spectral balancing made a huge improvement in the 360 degree sound field, surround "richness" and realism, and front imaging made a huge improvement. The surrounds particularly weave together with each other and with the front channels much smoother. Timbre matching of all the speakers is pretty amazing.

The only peculiar thing that occured was that the side surrounds' crossover recommendation was 40Hz, which they certainly cannot deliver. I changed that to 100Hz. Speaker distances were amazingly precise.

Being able to save each position's measurement data is terrific. Although the room is pretty large, the seating is only 14' from the LCRs, so Curve 1 was the best choice. I experimented with the mid-range correction filter, but on listening, I learned that the Klipsch THX U2 speakers don't need it. The curve editor is very cool. With a little imagination you can make some subtle changes that may fit your subjective tastes. I experimented with a bit of extra beef in the LF, creating a curve filter that makes a nice smooth curve that transitioned through the crossover region nicely, then rejoins the (amazingly smooth, flat) lower mid-range. It is great how you can build a filter and apply to each or all speakers, and save the filter file, as well as others you may create and want to load at a later time.

Here's another very nice surprise. I fired up a good RTA, and with pink noise, could easily verify what Audyssey Pro indicated would be the resulting FR. Even the mid-range correction dip was quite visible when I had it on. So, a skeptic I am no more. There is clearly a lot more going on in Audyssey Pro than I anticipated.

Great job, Chris and the gang at Audyssey!

Larry

"The curve editor is very cool. With a little imagination you can make some subtle changes that may fit your subjective tastes. I experimented with a bit of extra beef in the LF"

Does anyone know the range of the curve editor? Like what if you wanted a house curve that boosted 8 dB from 100 to 20 Hz?

"The curve editor is very cool. With a little imagination you can make some subtle changes that may fit your subjective tastes. I experimented with a bit of extra beef in the LF"

Does anyone know the range of the curve editor? Like what if you wanted a house curve that boosted 8 dB from 100 to 20 Hz?Not possible. Adjustments are restricted to the range +/-3dB from flat.

"Adjustments are restricted to the range +/-3dB from flat."

To be clear, are we talking about the so-called user selectable target curve?

Don't Audyssey's own selection of curve choices vary by more than that?

"Adjustments are restricted to the range +/-3dB from flat."

To be clear, are we talking about the so-called user selectable target curve?

Don't Audyssey's own selection of curve choices vary by more than that?

Hi Noah,

MultEQ Pro permits the choice of three curves:

High Frequency Roll-off 1 curve introduces a slight roll-off at high frequencies that accounts for the balance between direct and reflected sound for small to medium size rooms (room volume less than 2500 cu. ft.)

High Frequency Roll-off 2 curve introduces a slightly greater roll-off at high frequencies that restores the balance between direct and reflected sound for medium to large size rooms (room volume between 2500 and 5000 cu. ft.)

SMPTE 202M curve is an international standard for the high frequency roll-off applied in a typical 500-seat movie theater. It is appropriate for professional mixing spaces and dubbing stages that must be calibrated for film sound postproduction. It can also be used in extremely large playback spaces (room volume greater than 5000 cu. ft.)

These three curves are customizable within +/- 3 dB of the flat portion of the curve.

Larry

"Adjustments are restricted to the range +/-3dB from flat."

To be clear, are we talking about the so-called user selectable target curve?

Don't Audyssey's own selection of curve choices vary by more than that?Yes. You can grab and move a point from any of the provided target curves and put it anywhere in the acceptable range (or close to it). Thus, one can take a point at the upper end of the HF roll-off and move it up as far as +3dB from reference (flat). So, you can move it more than 3dB as long as it is in the acceptable range.

OK, thanks.

Sure seems paltry, hardly enough to provide real alternatives to what Audyssey thinks we should have.

I know they espouses adhering to a standard, but as they themselves have often said, the only source that adheres to one is movies, leaving all audio sources subject to the whim and incompetence of their masterers.

OK, thanks.

Sure seems paltry, hardly enough to provide real alternatives to what Audyssey thinks we should have.

I know they espouses adhering to a standard, but as they themselves have often said, the only source that adheres to one is movies, leaving all audio sources subject to the whim and incompetence of their masterers.

Hi Noah,

Regardless of how much adjustments Audyssey permitted, I doubt that customizing target curves to deal with incompetent mixing would prove to be a practical approach. ;) Don't you think that using the tone controls would be a lot easier and probably just as effective to help make a poor mixes more palatable?

However, customizing target curves to deal with different acoustics in the listening environment is a practical application of this feature. If the provided target curves are not suitable, I think that +/- 3 dB adjustment might be sufficient to come up with customized roll-offs to suit most rooms.

Larry

"Don't you think that using the tone controls would be a lot easier and probably just as effective to help make a poor mixes more palatable?"

Easier, sure, but how could it possibly be as effective?

"Don't you think that using the tone controls would be a lot easier and probably just as effective to help make a poor mixes more palatable?"

Easier, sure, but how could it possibly be as effective?

Different curves from Audyssey are technically superior sure, but when you consider that with all the recordings we own and play, each one has their own problems, the number of custom curves that Audyssey provides may be insufficient to deal with each of these tracks. And then you may want to deviate more than 3dB.

I know a lot of people who have computer based front ends for their music simply make a bunch of different EQ presets. It's not simply 2-band tone control either...sometimes up to 15 bands.

Anyway, Audyssey is better, but sometimes for particularly hot recordings, tone controls are useful.

"Don't you think that using the tone controls would be a lot easier and probably just as effective to help make a poor mixes more palatable?"

Easier, sure, but how could it possibly be as effective?

Hi Noah,

So is your solution to construct a customer target curve for every different poorly mastered piece of content, in a futile attempt to correct it?

As I alluded to earlier, high frequency roll-offs are usually needed to comfortably make the transition from the mixing acoustic environment to the listening environment. The standard roll-offs on the canned target curves are based on an educated guess of the statistical average of the intended home environment. The ability to craft a custom roll-off can be helpful to refine that roll-off to your particular situation. It really only needs to be done only once.

Larry

"So is your solution to construct a customer target curve for every different poorly mastered piece of content, in a futile attempt to correct it?"

No.

I have quite a few CD's that have a similar overly bright bass-shy sound that was helped some w/tone controls but not enough.

Kal,

I just re-read your review of the MultEQ Pro3.0 s/w ( http://stereophile.com/musicintheround/music_in_the_round_35/index1.html for those interested) and have a question.

You say "I was able to change the curve with the most gentle rolloff (down by about 8dB at 20kHz) to a curve with a slower slope that was laser-straight o 20kHz."

How was 8 dB of boost at 20 kHz possible?

Though I just noticed that you say "straight", not "flat" - so the 3 dB limit would mean you got it to -5 dB at 20 kHz?

Kal,

I just re-read your review of the MultEQ Pro3.0 s/w ( http://stereophile.com/musicintheround/music_in_the_round_35/index1.html for those interested) and have a question.

You say "I was able to change the curve with the most gentle rolloff (down by about 8dB at 20kHz) to a curve with a slower slope that was laser-straight o 20kHz."

How was 8 dB of boost at 20 kHz possible?

Though I just noticed that you say "straight", not "flat" - so the 3 dB limit would mean you got it to -5 dB at 20 kHz?
I said straight and I meant straight and flat! Read all the posts again. The limitation is that one cannot set it beyond +/-3dB of 0dB. Thus, I was able to move the -8dB point to 0 (8dB change) but I could also, if I wanted to, to move it to +3dB (11dB change).

"The limitation is that one cannot set it beyond +/-3dB of 0dB."

Sorry to be dense, but I looked back, and it's still not clear.

Is 0 dB referenced to absolute response, or to one of the Audyssey curves?

"The limitation is that one cannot set it beyond +/-3dB of 0dB."

Sorry to be dense, but I looked back, and it's still not clear.

Is 0 dB referenced to absolute response, or to one of the Audyssey curves?"Adjustments are restricted to the range +/-3dB from flat. "

I said straight and I meant straight and flat! Read all the posts again. The limitation is that one cannot set it beyond +/-3dB of 0dB. Thus, I was able to move the -8dB point to 0 (8dB change) but I could also, if I wanted to, to move it to +3dB (11dB change).

So how does Audyssey keep the user from blowing a tweeter if they chose lift the HF response that much?

So how does Audyssey keep the user from blowing a tweeter if they chose lift the HF response that much?3dB?

3dB?

No, 8 to 11dB as stated in the post I quoted.

No, 8 to 11dB as stated in the post I quoted.C'mon. If the original target is down 8dB, that is an attenuation 8dB. So raising the target curve to 0, a change of 8dB, means that there's neither a boost or a cut in the target curve. Raising the target curve up to +3dB, now a change of 11dB, results in a boost from reference of only 3dB.

Now, since we are talking about target curves, the actual amount of boost or cut from the raw speaker/room response will, as always, depend on what the uncorrected response is.

C'mon. If the original target is down 8dB, that is an attenuation 8dB. So raising the target curve to 0, a change of 8dB, means that there's neither a boost or a cut in the target curve. Raising the target curve up to +3dB, now a change of 11dB, results in a boost from reference of only 3dB.

Now, since we are talking about target curves, the actual amount of boost or cut from the raw speaker/room response will, as always, depend on what the uncorrected response is.

But Audyssey's treble target is not an attenuation of the signal. I've seen enough uncorrected speaker responses to know that Audyssey's treble target is pretty close to the natural roll-off you get from a 1 inch tweeter at a 8 to 12 foot seat in a typical room. So to achieve flat treble at the seat, Audyssey would have to significantly boost the output of the tweeter, hence my question.

But Audyssey's treble target is not an attenuation of the signal. No, it represents a potential attenuation from flat. I've seen enough uncorrected speaker responses to know that Audyssey's treble target is pretty close to the natural roll-off you get from a 1 inch tweeter at a 8 to 12 foot seat in a typical room. So, that would mean that the default Audyssey target would neither boost nor cut the uncorrected output?

So to achieve flat treble at the seat, Audyssey would have to significantly boost the output of the tweeter, hence my question. I guess that depends what you mean by "significantly." 5-8dB is significant but not troubling, particularly at HF.

So, that would mean that the default Audyssey target would neither boost nor cut the uncorrected output?

Correct.

I guess that depends what you mean by "significantly." 5-8dB is significant but not troubling, particularly at HF.

I would also consider 5-8dB significant, and it seems to be common wisdom that one shouldn't boost the treble too much. I'm not saying that know one way or another. What have you read that leads you to believe that an 8dB boost is not a problem for HF?

I would also consider 5-8dB significant, and it seems to be common wisdom that one shouldn't boost the treble too much. I'm not saying that know one way or another. What have you read that leads you to believe that an 8dB boost is not a problem for HF?I would be more concerned with LF or low MF boosts as they would affect amplifier headroom. There's generally so little energy in the 15-20kHz range that a boost there is not a challenge. In practice, I have heard nothing troubling.

I agree that there isn't a lot of content in the upper treble, but I would still be concerned with an 8dB boost if listening at or near reference levels for the sake of the tweeter's voice coil, not so much the amp power required.

Do you listen with flat treble response at high volumes?

I guess I have always liked to set up a reference balance and then listen to the recordings as they are.With Audyssey (NAD version)I find all of my recordings sound better than before.
Since music is an emotional communication, I suppose one might attempt to enhance that with eq or tone control adjustments specifically for each recording,but it would seems unlikely to be the artists intention.IMHO.

"Thus, I was able to move the -8dB point to 0 (8dB change) but I could also, if I wanted to, to move it to +3dB (11dB change)."

Either I was blind yesterday or you did an edit soon enough to not have it indicated as such.

Anyway, thanks for the clarification.

Do you listen with flat treble response at high volumes?I listen with treble flat almost all the time at what ever level I choose. However, I should say that, while I can listen to music at pretty much any level I choose, I usually do not listen to movies at reference level.

Since music is an emotional communication, I suppose one might attempt to enhance that with eq or tone control adjustments specifically for each recording,but it would seems unlikely to be the artists intention.IMHO.

Because there are no standards for musical recordings, there is no way of knowing if what you are hearing is as the artist intended, even if your system is accurate.

More thoughts on this subject here:

http://www.avsforum.com/avs-vb/showthread.php?t=1162044

I listen with treble flat almost all the time at what ever level I choose. However, I should say that, while I can listen to music at pretty much any level I choose, I usually do not listen to movies at reference level.

So you forgo the roll-off in the treble when using Audyssey?

I've tried that, but the sound becomes unnaturally etched to these ears. I EQ to flat up to about 15K and don't touch it above that point. I *think* this is because to get flat treble as measured at the LP, you have to put a rise in the direct response to compensate for the lack of reflected energy. So what I hear is lifted treble even though the measurement is flat.

So you forgo the roll-off in the treble when using Audyssey? Generally.

I've tried that, but the sound becomes unnaturally etched to these ears. I EQ to flat up to about 15K and don't touch it above that point. I *think* this is because to get flat treble as measured at the LP, you have to put a rise in the direct response to compensate for the lack of reflected energy. So what I hear is lifted treble even though the measurement is flat.That is a reasonable analysis but it depends greatly on the dispersion characteristics of the tweeter and on many room properties. That said, I suspect that, if I did as you suggest, I might not hear much of a difference given the fall-off in my hearing in that range.

Hi guys

Some time passed since last time I read this discussion and I see that there are still some misinformation or misunderstanding. I am very surprised that no one from Audyssey explains here how their system works.
I will try, but first please consider that I am NOT an “Audyssey expert”, I am trained by them, I attend almost every training and demo they provide in MA or CEDIA and I calibrate receivers equipped with “Audyssey Pro” option, so I have some experiences in this field. (BTW it is Audyssey not Audessey)

So:
1. “Audyssey MultEQ XT” removes distortion caused by the interaction of sound from loudspeakers within the room.
2. “Audyssey” provides 2000+ band equalizer for every audio channel.
3. “Audyssey” calibration must be done properly. If not – result may be really bad.
4. “Audyssey” offers couple of options. One of option – which nobody mentioned here - is “Audyssey Flat”. “Audyssey Flat” is designed for music. “Audyssey MultEQ XT” is designed for movies.

Audyssey created three target sound options (for movies)

http://www.avsforum.com/avs-vb/attachment.php?attachmentid=155757&stc=1&d=1256223864

All of them are adjustable.
Midrange compensation can be turned off and roll of can be adjusted per every channel.

Why they created “Target Curves”?
Movies mixed for large theaters. In a smaller room at home, Audyssey applies a gentle rolloff for high frequencies and midrange compensation, when you select MultEQ XT. However, if the movie already had the high-frequency corrected for homes, it is better to use Audyssey Flat mode.
If you play music, Audyssey Flat is the only proper option.


Chris. Try to listen to your stereo with “Audyssey flat” for several hours and then turn it off, please. I like to know your opinion.

As audioguy wrote: “Audyssey for the price and functionality is a no brainer. If it doesn't work in your environment, there are other much more expensive solutions.”

Best Regards

I think Audyssey flat is also recommended for highly treated (room treatment) HT rooms.

Mike

I think Audyssey flat is also recommended for highly treated (room treatment) HT rooms.

MikeI would recommend Audyssey flat if I am in well treated room. So far I calibrated receivers installed in living rooms.

BTW two days ago I was on "For love of music event" where I listened to Snell A7 Illusion speakers driven by McIntosh preamplifier and power amplifier.
Music was amazing! I'm not sure that Audyssey or other equalization could make important improvement. Maybe just small one. Yes, room is well treated.

Regards

I am very surprised that no one from Audyssey explains here how their system works.

Hi,

Thanks for your input.

Prof. Chris Kyriakakis, is Chief Technology Officer for Audyssey and he is an extremely active poster over at the "Offical" Audyssey thread (http://www.avsforum.com/avs-vb/showthread.php?t=795421). Due to commitments on his time it is not practical for him to follow multiple threads, but folks can post MultEQ Pro questions at that thread and he will respond.


4. “Audyssey” offers couple of options. One of option – which nobody mentioned here - is “Audyssey Flat”. “Audyssey Flat” is designed for music. “Audyssey MultEQ XT” is designed for movies.
Audyssey created three target sound options (for movies)

http://www.avsforum.com/avs-vb/attachment.php?attachmentid=155757&stc=1&d=1256223864

All of them are adjustable.
Midrange compensation can be turned off and roll of can be adjusted per every channel.

Why they created “Target Curves”?
Movies mixed for large theaters. In a smaller room at home, Audyssey applies a gentle rolloff for high frequencies and midrange compensation, when you select MultEQ XT. However, if the movie already had the high-frequency corrected for homes, it is better to use Audyssey Flat mode.
If you play music, Audyssey Flat is the only proper option.


I don't think it is technically correct to state that MultEQ XT is specifically designed exclusively for movies. As you state it was designed to correct for the influence of the room on sounds, and that objective is not limited to just movies. In other words, MultEQ XT is designed to smooth the frequency response for both movies and music.

The high frequency roll-offs of the target curves are designed to take into consideration differences in the acoustics between the recording environment and the listening environment. That is, the balance of direct sounds and reflected sounds is usually different between the mixing and listening environments and at higher frequencies, where speakers become more directive, it is usually necessary to roll-off the highs to avoid overly bright sounds in typical home setups. Whereas it is true that only movies have published standards for mixing, nevertheless, many music mixes which are frequently mixed hotter than movies, may also benefit from a high frequency roll-off.

Obviously, the third target curve, SMPTE 202M, is specifically designed for commercial movie theaters or large venues, but the other two target curves may be useful for music depending on the room. That is, lively home listening environments may benefit from a roll-off even when listening to music.

The reason for developing a Flat target curve wasn't so much to deal with music, but rather to accommodate THX listening modes which introduce their own approach to high frequency roll-offs, THX Re-EQ. The Flat target curve also accommodates those situations where the listener is seated in the near field. The following is an excerpt from the Audyssey Pro user's guide discussing the issue.


In addition to the user selected curve, MultEQ uses a Flat curve to create equalization filters that correct the response to flat from the low frequency cutoff point that MultEQ Pro has determined to the upper frequency limit of the tweeters. This setting must be used in a THX system in order to allow THX re-equalization to operate as it was intended. It is also recommended for very small or highly treated rooms in which the listener is seated close to the loudspeakers.


Larry

After following this thread I'm considering the Audyssey Sound Equalizer with my Theta CBIII. There is no dealer within a few hours and I would appreciate any lead to a dealer that would ship the balanced unit with Pro kit at a discount. Please PM if you have a suggestion.

Has anyone tried the standalone unit w/o an Audyssey equipped processor?

DaveN, I use a Balanced Pro unit. Works well. (The only limitation vs. new processors is that you don't get the Dynamic EQ or Dynamic Volume functionality.)

In my system it does a terrific job of improving the tonal and time alignment of the mains and surrounds. I bypass it for the left and right mains when I am doing pure two channel listening, but find it makes a material improvement for surround and sub intensive applications.

What "Movies" do you use as your "reference" ?
There's a lot of talk about calibrating but what do the "Calibrators" use as their reference and how do they know how it should sound?

If you haven't already taken a look, Sean Olive over at Harman performed some testing (http://seanolive.blogspot.com/2009/11/subjective-and-objective-evaluation-of.html) with different processors and target curves where he discovered that the most favored curve was a fairly inclined slant from gained bass frequencies to attenuated high frequencies. Interesting findings and a good first effort.

Correlation Between Subjective and Objective Measurements


In-room acoustical measurements were made at the six listening seats using a proprietary 12-channel audio measurement system developed by the Harman R&D Group. Slides 23 and 24 show the amplitude response of the different room corrections spatially averaged for the six seats (slide 23), and at the primary listening seat (slide 24). The measurements are plotted from top to bottom in descending order of preference, each vertically offset to more clearly delineate the differences. A few observations can be made:


1. The six-seat spatially averaged curves (slide 23) of the room corrections do not explain listeners' room correction preferences as well as the spatially averaged curves taken at the primary seat (slide 24). This makes perfect sense since all of the listening was done in the primary listening seat.
2. Looking at slide 24, the most preferred room corrections produced the smoothest, most extended amplitude responses measured at the primary listening seat. The largest measured differences among the different room corrections occur below 100 Hz and around 2 kHz where the loudspeaker had a significant hole in its sound power response. The room corrections that were able to fill in this sound power dip received higher preference and spectral balance ratings.
3. A flat in-room target response is clearly not the optimal target curve for room equalization. The preferred room corrections have a target response that has a smooth downward slope with increasing frequency. This tells us that listeners prefer a certain amount of natural room gain. Removing the rom gain, makes the reproduced music sound unnatural, and too thin, according to these listeners. This also makes perfect sense since the recording was likely mixed in room where the room gain was also not removed; therefore, to remove it from the consumers' listening room would destroy spectral balance of the music as intended by the artist.


Slide 24:
http://docs.google.com/gview?srcid=0B97zTRsdcJTfY2U4ODhiZmUtNDEyNC00ZDcyLWEzZTAtMGJ iODQ1ZTUxMGQ4&pid=explorer&docid=6acfbb12ff81eb2c6a886a651d8b41c8%7Cbfd2f1cc1e50e5153b0 3304f616bcfda&a=bi&pagenumber=24&w=800

Very interesting. A logarithmic 10db drop from about 20hz to 20K Hz. I'm going to try it since I think I can create that curve with Audyssey Pro.

If you haven't already taken a look, Sean Olive over at Harman performed some testing (http://seanolive.blogspot.com/2009/11/subjective-and-objective-evaluation-of.html) with different processors and target curves where he discovered that the most favored curve was a fairly inclined slant from gained bass frequencies to attenuated high frequencies. Interesting findings and a good first effort.



Slide 24:
http://docs.google.com/gview?srcid=0B97zTRsdcJTfY2U4ODhiZmUtNDEyNC00ZDcyLWEzZTAtMGJ iODQ1ZTUxMGQ4&pid=explorer&docid=6acfbb12ff81eb2c6a886a651d8b41c8%7Cbfd2f1cc1e50e5153b0 3304f616bcfda&a=bi&pagenumber=24&w=800

Hi,

Thanks for the link.

Dr. Olive's listening test method states that the responses were normalized to within .1 dB, but I wonder what the test volume was in relation to the levels the content was mixed.

We know that the sensitivity of human hearing decreases in the bass and treble regions as the levels are reduced. Therefore, if the listening room acoustic environment was the same as the mixing acoustic environment, it would require boosting the bass and to a lesser degree the treble to obtain a perceptually flat frequency response when the listening levels are lower than the mixing levels. How much a boost is required for bass and treble for a perceptually flat response will differ depending on how much below the mixing level the content was played. This non-linear hearing phenomena is obvious in reviewing the various equal loudness curves.

However, rarely are the acoustic environments between the mixing room and home listening room the same. In typical home listening environments as the frequency increases speakers become more directive and a roll-off of the high frequencies is usually needed to tame brightness. This needed reduction in the highs probably is greater than the need to slightly boost them when listening at lower levels.

Whereas the Harman listening room does have a bit of acoustic treatments, it nevertheless has a large portion of exposed wood flooring and it is the size of many home listening rooms. Therefore, even though it may have acoustics which are not as lively as most home listening rooms it still has more in common with them than the recording studios. Therefore I believe that a high frequency roll-off is also appropriate for listening in such a room.

So putting this together we may infer that if listening at below the mixing level a listener would prefer a sizable boost in the bass to make up for the decrease in hearing sensitivity and a gentle roll-off in the treble to deal with acoustic differences between the mixing and home listening environments. It is not surprising that a smooth frequency response is preferred connecting this boost in lows to the roll-off in highs.

It should be noted that Audyssey correction now includes a feature called Dynamic EQ that deals with correcting the frequency response to accommodate a reduction in listening levels and it does so in real-time measuring the actual changes in levels throughout the content. Although the Audyssey target curve is flat below 4 kHz, this Dynamic EQ feature alters the response effectively inserting a tilt in the frequency response as the volume is reduced.




In addressing audioguy's remark it may not be necessary, or even helpful, to construct a static tilted target curve in Audyssey Pro if one intends to engage the Dynamic EQ feature. Dynamic EQ will dynamically adjust the "tilt" in real-time to accommodate the actual listening levels. Any static tilt is only going to be effective at one particular listening level.

Larry

The author started a thread on this

http://www.avsforum.com/avs-vb/showthread.php?t=1192916