Audyssey killing my HT sound?

[Hetfield]

The important thing is that you know EXACTLY what sounds awesome to you and you don't need anyone telling you what is the definition of "amazing great sound".

What? You didn't hear those dips and peaks around 3kHz and 10kHz?

No, I can't say I have. I guess I like the non flat perfect sound experts talk about.

Sent from my Pixel using Tapatalk

 

[PENG]

I am not sure which corrected response @Matthew J Poes is referring to with, "The Harman work did, in fact, show that the corrected response was preferred."

You are right, but keep in mind "Corrected" response, and "flat" response are:

a) Both terms are relative, so you have to consider the "to what extent"..
b) The two are not mutually exclusive, you can have corrected but not that flat, or flat and yet not even corrected.

If you read the conclusions, it did cover the following:
https://drive.google.com/file/d/0B97zTRsdcJTfY2U4ODhiZmUtNDEyNC00ZDcyLWEzZTAtMGJiODQ1ZTUxMGQ4/view?hl=en

- Less preferred products had less smooth and extended in-room frequency responses; this was associated with more negative comments related to lack of bass (thin),brightness, and coloration
- Flat in-room response is not the optimal target response (program may be a nuisance variable)

Audyssey got bashed on forums to no end as more people learnt about "The Harman work..". Harman did not identify any of them, but most people guessed RC6, the worse performer, was Audyssey. Fair or not, that probably helped Audyssey in the end, as it might have forced them to become more practically minded in developing their XT version as well as the Editor App. Note: I am not saying RC6 is them, in theory, no one knows for sure, that except Harman, who, imo did the right thing not identifying any.

The fact is, owners of recent D&M models that come with Audyssey XT32 Sub EQ, and if they use the App, they can now do the following:

- disable the notorious BBC dip.
- limit the correction/eq range to below 300 Hz (or whatever the user prefers).
- reshape the target curve if they are willing to take the time and experiment.

Combined with the much higher filter resolution and the more powerful DSP engines, the negative findings by Harman on their product that was most likely the XT version at best may no longer apply, though I am sure it still won't come close to RC1 and RC2, i.e. Harman's own lol..

I wish Harman would do a follow up using similar or further improved protocol and still include the top 5 (based on user base) using their most up-to-date versions. Or hopefully Matthew can do something similar even if in much reduced scope as he likely has much less resource than Harman.

 

[Matthew J Poes]

That's not true. The Harman How to Listen software trains you identify peaks and dips in an ever narrowing band of frequencies and can be used on any computer, with any speakers.

I don't mean to condemn, but I think more people could learn to listen, better, if they try to train their ears. And I highly recommend this program to anyone!

But also to give perspective, while some are still debating the effectiveness of Audyssey, (LOL!) Harman has created an AI system to address room/cabin induced problems as well as a sound cancelling system that effectively creates electronic sound proofing for any room/car. The car version seems to be more well published, as it uses the car's sensors to also anticipate bumps and microphones to fine tune the cancellation it emits through speakers - anti nodes combine with the anticipated outside noise and cancel each other out.

I am not sure which corrected response @Matthew J Poes is referring to with, "The Harman work did, in fact, show that the corrected response was preferred."

One important point is that the Harman systems also work with the available anechoic data for any of their loudspeakers - you cannot take full advantage of room curve data, without knowing what the speaker sends out, free of reflection.

But at the same time, an SPL meter is one of the simplest tools to balance the output of each speaker at each listening location. I think the consumer could get receiver's etc. a lot cheaper if we weren't stuck paying for that Audyssey sticker...

That’s what I was referring too. That the ARCOS corrected response in their test scores better than no correction. In other words, it isn’t just about no harm but about improvement. That can’t be said for all room correction systems.

 

[Matthew J Poes]

You are right, but keep in mind "Corrected" response, and "flat" response are:

a) Both terms are relative, so you have to consider the "to what extent"..
b) The two are not mutually exclusive, you can have corrected but not that flat, or flat and yet not even corrected.

If you read the conclusions, it did cover the following:
https://drive.google.com/file/d/0B97zTRsdcJTfY2U4ODhiZmUtNDEyNC00ZDcyLWEzZTAtMGJiODQ1ZTUxMGQ4/view?hl=en

- Less preferred products had less smooth and extended in-room frequency responses; this was associated with more negative comments related to lack of bass (thin),brightness, and coloration
- Flat in-room response is not the optimal target response (program may be a nuisance variable)

Audyssey got bashed on forums to no end as more people learnt about "The Harman work..". Harman did not identify any of them, but most people guessed RC6, the worse performer, was Audyssey. Fair or not, that probably helped Audyssey in the end, as it might have forced them to become more practically minded in developing their XT version as well as the Editor App. Note: I am not saying RC6 is them, in theory, no one knows for sure, that except Harman, who, imo did the right thing not identifying any.

The fact is, owners of recent D&M models that come with Audyssey XT32 Sub EQ, and if they use the App, they can now do the following:

- disable the notorious BBC dip.
- limit the correction/eq range to below 300 Hz (or whatever the user prefers).
- reshape the target curve if they are willing to take the time and experiment.

Combined with the much higher filter resolution and the more powerful DSP engines, the negative findings by Harman on their product that was most likely the XT version at best may no longer apply, though I am sure it still won't come close to RC1 and RC2, i.e. Harman's own lol..

I wish Harman would do a follow up using similar or further improved protocol and still include the top 5 (based on user base) using their most up-to-date versions. Or hopefully Matthew can do something similar even if in much reduced scope as he likely has much less resource than Harman.

I’ve actually been working on developing a simple way to do MUSHRA testing. Once i figure that out I could do a test every bit as sophisticated as what Harman did. Because all of these systems are based on preconditioning the music via a transfer function of some kind, I can actually capture that transfer function and use it to encode special musical tracks. I can easily use these in ABX testing but that is of little merit. MUSHRA is what is needed and that is a bit harder. I really wish someone would create a free or cheap iOS app that can do it. It’s not a very complicated idea.

Having said that, if you guys want any kind of review of room correction any time soon, I need to avoid deep dive analysis. That takes time to do right and inevitably turns articles into niches few want to bother reading.

My guess is that Harman and most other groups are uninterested in this research right now. Room correction has notably lagged as of late. When is the last time you saw a major update to Audyssey? DIRAC is way behind on their development work and I believe it is because they are more focused on other markets with bigger profits (smart phones, VR, etc.). A lot of great products never made it to the mainstream market and have similarly stagnated (Trinnov). A quick search of the literature shows the research never stopped, we’ve learned a lot since these systems came to market. Where are the new products. MIMO is the likely next big thing and yet there are none on the consumer home market. It isn’t even the biggest new area, there are already technologies that can surpass the capability of MIMO for which the only products are matlab programs.

There are currently a number of concepts that are also very disconnected and I think shouldn’t be. Measurements using 3D impulse responses along with at positions in the near field of each speaker as well as at the listening position could all be utilized to learn about the different types of problems and apply a smarter correction. This could also be used to capture the rooms acoustics and allow multichannel systems to essentially change the acoustics of the room. Think true room correction. Nothing like this is on the market for consumers and it could be.

 

[M Code]

There are currently a number of concepts that are also very disconnected and I think shouldn’t be. Measurements using 3D impulse responses along with at positions in the near field of each speaker as well as at the listening position could all be utilized to learn about the different types of problems and apply a smarter correction. This could also be used to capture the rooms acoustics and allow multichannel systems to essentially change the acoustics of the room. Think true room correction. Nothing like this is on the market for consumers and it could be.

Are U aware that the Room EQ SW used in certain JBL Synthesis and higher end HK AVRs used a near-field measurement of each loudspeaker before running the far-field sweep... I don't recall all of the respective brand/models that have this capability and to my ears does an incredible job. Note that the 1st release of this SW was over 10 years ago. The down-side is that it requires significant DSP resources which usually limits what the SW can do depending upon the product's price-point. The other positive point I find for Harman is that they design and build products for each part of an advanced system from source, amplifier and loudspeaker be it a consumer or pro system. Though much of their in-house R&D is confidential, for each part I have been exposed to is very impressive. And finally there are many well-respected personnel in our audio biz, Dr. Toole & Dr.Olive are (2) are the best known...

Just my $0.02....

 

[Matthew J Poes]

Are U aware that the Room EQ SW used in certain JBL Synthesis and higher end HK AVRs used a near-field measurement of each loudspeaker before running the far-field sweep... I don't recall all of the respective brand/models that have this capability and to my ears does an incredible job. Note that the 1st release of this SW was over 10 years ago. The down-side is that it requires significant DSP resources which usually limits what the SW can do depending upon the product's price-point. The other positive point I find for Harman is that they design and build products for each part of an advanced system from source, amplifier and loudspeaker be it a consumer or pro system. Though much of their in-house R&D is confidential, for each part I have been exposed to is very impressive. And finally there are many well-respected personnel in our audio biz, Dr. Toole & Dr.Olive are (2) are the best known...

Just my $0.02....

Yes I know these guys well. What is currently used in Synthesis is not an automatic system for end user implementation though. It also can’t be as I understand it.

I didn’t think that what Harman included in HK’s was quite the same. I thought it was more closely related to what they released in that little jbl synthesis subwoofer eq box that didn’t seem to ever go anywhere. I used that system once and didn’t get the best results. I also recall reviews not being the most positive about it.

I wasn’t knocking on Harman’s approach but as of right now their approach is only used in very expensive custom install gear and only works as intended with JBl synthesis speakers. It’s a great approach, but not practical for wide dissemination. Even Harman has adopted DIRAC into their other products from companies like Lexicon. While I’m told the next implementation of that series will be on custom architecture instead of the current Arcam architecture, I assume they will continue with Dirac.

What Harman is doing, as far as I can tell, is still a basic SIMO approach. It’s a very sophisticated and well executed version, but it isn’t taking advantage of what I mentioned.

The near field measurements they use I am sure are being used to assess the speakers direct response and detect room induced anomalies. It may also be used to assess interference effects but to do that efficiently you need wavelets typically and I am fairly positive they are not using wavelet eq. Given that Todd Welti told me he hated learning about Wavelets (the math feels counterintuitive and they appear acausal) I feel it unlikely something they explored much.

 

[Matthew J Poes]

https://hometheaterhifi.com/reviews/amplifier/integrated-amplifiers/harman-kardon-hk-990-stereo-integrated-amplifier-with-digital-room-correction-and-dual-subwoofer-bass-management-part-i/

This actually did a good job of delving into what I recall being the problems with the HK system. The author indicates what I recalled being told, that the EQ used nearfield simply to estimate the anechoic response of the speaker. The author noted the problems with this approach. If you could accurately measure the anechoic response of a speaker in a room by simply measuring it from one meter away we would all be doing it. @shadyJ job would be much easier, as would mine. Such an approach only works to a point and still tells us nothing about the polar response of the speaker. It’s a half measure. My understanding is that ARCOS actually makes use of a database of speaker characterization data for Synthesis speakers. That is the right way to do this.

My comment about near-field was related to something else. Such an approach allows an accurate estimation of boundary interference effects. Research into wavelet analysis and processing lead researchers at BOHMER to develop a method to estimate SBIR with a nearfield meadurent and correct it with a kind of wavelet eq. Nobody else is doing this right now that I am aware of. They believe that correcting this is the only thing that matters however and I’m not sure I agree. Their processor is not really automatic either. The philosophy they take is that speaker optimization starts with manual correction of the speakers anechoic response. Then boundary interference effects are corrected. Room modes are ignored.

This is something I think can and should be incorporated into more measurement systems. While I can’t and haven’t replicated Bohmer’s process, when I professionally setup a system I do in fact use near field measurements to separate different distortions. An Algo can do this too. It’s soemthint I think should be done.

HK’s eq also does something in a “dumb” way that I think could be done in a “smart” way. Either speaker companies need to test the dynamic limits of their systems and provide a standardized data file that can be used in the EQ to establish its dynamic limits over a given bandwidth or the system needs to measure and estimate it. It would then allow for more intelligent eq and application of limiters. Nothing ruins a movie or song more than a system distorting. No reason this can’t be addressed in the processor. For the most part EQ should not be extending the bass of a system unless the processor knows the speakers can handle it. Most modern subs are already eqed to their limit. Most mains simply can’t handle it.

 

[Hetfield]

Toy Story 3 happened to be on this morning before work and I watched 15 minutes of it on a low volume and it sounded really good. So I think I'm in the clear with my sound being back where I want it.

Sent from my Pixel using Tapatalk

 

[PENG]

Toy Story 3 happened to be on this morning before work and I watched 15 minutes of it on a low volume and it sounded really good. So I think I'm in the clear with my sound being back where I want it.

Sent from my Pixel using Tapatalk

Did you use DEQ at low volume? You can have Audyssey off as yo prefer, but still have DEQ on. You can change the offset to your taste. It is better than just boosting the sub level because it is "dynamic", hence the effects vary with volume.

 

[Hetfield]

Did you use DEQ at low volume? You can have Audyssey off as yo prefer, but still have DEQ on. You can change the offset to your taste. It is better than just boosting the sub level because it is "dynamic", hence the effects vary with volume.

No, but how the heck do you do that? Sounds interesting.

Sent from my Pixel using Tapatalk

 

[lovinthehd]

Did you use DEQ at low volume? You can have Audyssey off as yo prefer, but still have DEQ on. You can change the offset to your taste. It is better than just boosting the sub level because it is "dynamic", hence the effects vary with volume.

Only in my old Onkyo can I use DEQ without Audyssey on, my Denons don't allow that. Or do newer Denons allow that (DEQ without Audyssey)?

 

[panteragstk]

I learned yesterday that Audyssey was actually making my room sound worse. I had wondered why it didn't really have the bass impact that it did in my old house. Well, see this post and you'll find out what I'm talking about.

What I learned is that we can't just trust what Audyssey is doing. It has to be measured after the fact. My system was fine without Audyssey at all, but the flat curve was excellent as well. I had reference enabled which was garbage. Once I used the Audyssey App my reference curve is great. Removed that 2khz dip as well. Theater sounds like it should now.

 

[PENG]

No, but how the heck do you do that? Sounds interesting.

Sent from my Pixel using Tapatalk

Sorry I was wrong. Audyssey has to be "on" in order to use DEQ. As ADTG indicated he used to have DEQ on and Audyssey set to bypass, but that only means Audyssey is effectively "off" for the left/right channels.

Again, my apology..

 

[PENG]

Only in my old Onkyo can I use DEQ without Audyssey on, my Denons don't allow that. Or do newer Denons allow that (DEQ without Audyssey)?

Apparently not, see my post #73.

 

[Grandzoltar]

Most speakers have that dip already built into the crossover. As you can see in the pic. I allow audessy to correct up to the natural dip at about 2200hz.

 

[Grandzoltar]

Other big named room correction like arc only correct up to 5000hz. Food for thought.

 

[PENG]

I learned yesterday that Audyssey was actually making my room sound worse. I had wondered why it didn't really have the bass impact that it did in my old house. Well, see this post and you'll find out what I'm talking about.

What I learned is that we can't just trust what Audyssey is doing. It has to be measured after the fact. My system was fine without Audyssey at all, but the flat curve was excellent as well. I had reference enabled which was garbage. Once I used the Audyssey App my reference curve is great. Removed that 2khz dip as well. Theater sounds like it should now.

If you use the App a while back, be sure to get the latest update that has removed the below 20 Hz roll off. After that you need to upload the target curve again in order to enjoy or at least feel the deep bass if your sub(s) can go much lower than 20 Hz. I am surprise you prefer the flat curve, apparently most people don't.

 

[PENG]

Most speakers have that dip already built into the crossover. As you can see in the pic. I allow audessy to correct up to the natural dip at about 2200hz.

You should try to disable that 2 kHz dip setting. Attempting to correct above the room transition point, the so called Schroeder frequency is a hit and miss thing so you may want to try limiting it to 300 Hz to hear for yourself. If you like it better and want to go even lower, it is easy to do with the App.

 

[Hetfield]

Dude let me tell you my system sounds so good now! Very happy, very happy. Audyssey off and happy again.
Now please tell me because I'm a big dummy. Why when Audyssey was on and I had that dynamic volume on my volume was like -40 and now it's -15 are about the same loud? Meaning why do I have to turn the amp up higher than I did before when now I'm getting full range or no roll off? Be gentle with me, as I said I'm a big dummy.

Sent from my Pixel using Tapatalk

 

[lovinthehd]

Dude let me tell you my system sounds so good now! Very happy, very happy. Audyssey off and happy again.
Now please tell me because I'm a big dummy. Why when Audyssey was on and I had that dynamic volume on my volume was like -40 and now it's -15 are about the same loud? Meaning why do I have to turn the amp up higher than I did before when now I'm getting full range or no roll off? Be gentle with me, as I said I'm a big dummy.

Sent from my Pixel using Tapatalk

I'd guess they're not actually the same level. -40 with DEQ vs -15 without Audyssey/DEQ on my systems are drastic differences in volume. I can see at -40 with some bass/treble boost from the eq that it might take a few dB higher with the master volume to get the same levels of bass/treble, but....