Home Theater Forum and Systems banner

1 - 20 of 27 Posts

·
Premium Member
Joined
·
12 Posts
Discussion Starter #1
Hello,

This is my first post on this forum, so I hope I have followed all of the posting rules. Sorry my first post is so long, but I have tried to include all of the necessary information.

I would appreciate some help with a problem I am having with dialog intelligibility in my home theater.

I recently bought a Denon X4200W AVR and have calibrated my theater using the Audyssey XT32 that was included with the AVR. The new AVR is used in my theater with a 7.1 setup.

In general, I am enjoying the Denon and have used it for playback of both music and Blu-ray movies. However, I have noticed the vocals in music and the dialog in movies are often slightly muffled and, at times, are difficult to understand with the Audyssey calibration turned on. With Audyssey turned off, the system playback, as a whole, is not as pleasing; but, the dialog clarity improves.

To tackle this problem, I first searched this forum and others for posts dealing with “intelligibility.” The following steps are my take on the advice given to others with this issue.

First, I used this problem as an excuse to buy a new center speaker (an RBH 661C/R). The RBH 661 is recommended by RBH to use with the RBH 61C/R’s I use for front left and right speakers. Overall, the new center speaker sounds better than the one it replaced, but the vocal/dialog issue is still degraded with Audyssey engaged.

I finally decided it was time to break out my UMIK-1 and use REW (Ver. 15.14) to try to find the cause of the intelligibility problem. Details of my test technique are included at the bottom of this post.

This graph is for the center speaker with all other speakers turned off (1/6-Octave smoothing). The purple trace is the recorded response with Audyssey turned on (Dynamic EQ turned off) and the green trace is with the Audyssey turned off.

CENTER SPEAKER AUDESSY ON & OFF 4_15_2016.jpg

Here is a graph (actually a photo of my plasma TV) presented by the AVR of the Audyssey calibration of the center speaker:

center speaker (1 of 1).jpg


My interpretation of these graphs is that the Audyssey calibration greatly reduced the mode at about 130Hz and smoothed out the null at about 300Hz.

However, the REW graph with Audyssey turned on, compared to the calibration turned off, seems to have significantly reduced the SPL between about 1.3kHz and 2.7kHz. My research found that this band (1.3 to 2.7kHz) is right in the middle of the intelligibility frequency spectrum of the human voice!

In case this will be of value, here is a graph of my left front speaker with Audyssey turned on and off (1/6-octave smoothing; Purple is Audyssey off).

LEFT SPEAKER  AUDYSSEY ON & OFF 4_16_2016.jpg

Here is a graph of my right front speaker with Audyssey turned on and off (1/6-Octave smoothing; Green is Audyssey off).

RIGHT FRONT PEAKER AUDYSSEY ON & OFF 4_16_2016.jpg

So, my first question is if I am analyzing these graphs correctly (i.e., Audyssey’s calibration is degrading the vocals). Or do the graphs show another anomaly that can account for the speech-clarity issue?

If I have found the problem, is there any alternative other than manually EQing the room?

DATA GENERATING TECHNIQUE

I located the UMIK-1 at my primary listening position, a leather reclining chair, using a mic boom and tripod. To reduce some of the reflections from the reclining chair, I covered it with a thick blanket. The mic was positioned pointing toward the ceiling and the calibration file for 90 degrees was used in REW. Each measurement in the graph is the average of 3 mic positions – all at my usual ear level, but the first measurement was located at the center position, then the second measurement was 6 inches to the left (parallel to the head rest) and the third measurement was 6 inches to the right of the center position (also parallel to the head rest). Using REW, I averaged the 3 readings and then displayed them with 1/6-octave smoothing.

Thanks for your help.[/SIZE]
 

·
Premium Member
Joined
·
15,054 Posts
What centre channel are you using? May we see a picture of your center channel speaker location as well as the front of the room? Placement and type of center channel speaker can dramatically change what you hear.
 

·
Premium Member
Joined
·
12 Posts
Discussion Starter #3
I took a few quick photos of the front of my home theater.

These two wide-shots show my RBH SX-61/R left and right speakers, my RBH 661C/R center speaker and two Velodyne DD-18 subwoofers. You can also see the two corner bass traps, and the panels for absorption of reflections. Not shown is the absorption panel mounted on the ceiling between the center speaker and the primary listening position.

Photo of front of theater (1 of 1).jpg

Home Theater (1 of 1).jpg


Here is a close-up of the center speaker, under my Panasonic plasma TV, and between my subwoofers. As you can see in the photos, I didn't have much choice in the location of the center speaker.

Photo of Center Speaker (1 of 1).jpg


I hope these are useful.
 

·
Super Moderator
Joined
·
9,234 Posts

First, welcome to the Forum!

As I see it, there are two primary issues affecting vocal clarity. One is the loss in the 2 kHz area (as you noted). The other is the roll-out of the high frequencies, where the sibilants are.

For the latter, you might try taking measurements with a 0° calibration file and the mic pointed directly at the speaker. Since you have absorption on the ceiling, it’s possible the lack of overhead reflections arriving at the mic are causing the high end portion of the graph to look worse than it really is.

There are a couple of things you can try. If you’re running Audyssey with the normal recommendation of analyzing multiple positions (i.e. putting the mic in various locations all over the room), try instead analyzing only at the primary listening position, or in locations only a few inches varying from that. Our very own AudiocRaver has performed evaluations and written some pieces showing that single-mic analysis typically gets the best results.

Alternately, bypassing Audyssey for manual EQ is a good option, assuming your receiver has built-in equalization that is flexible enough.

Regards,
Wayne

 

·
Plain ole user
Joined
·
11,121 Posts
I agree with Wayne (both of them) with regard to using only measurements at the primary listening position. I would also try moving the center, left, and right speakers out toward the listening position.
 

·
Premium Member
Joined
·
12 Posts
Discussion Starter #7
Thanks for all of the suggestions.

Regarding my procedure for Audyssey measurements, I was fortunate to have read and followed the directions of Wayne Myers (“AudiocRaver”) in making the Audyssey tests in my first post. (I even took his advice and bought a Bosch laser distance meter—a great gadget!) I have reread his extensive posts and followed his procedure in the reconfiguration described below.

The REW measurements and photos I have previously posted were based on the following speaker configuration:

OLD:

Left and Right Front speakers: (1) Distance from front of each tweeter to front wall: 2’ 7”; (2) Distance from front of each tweeter to side wall: 2’4”; (3) Distance from center of Left Front tweeter to Right Front tweeter: 8’6”; (4) Distance from center of tweeter for both Left and Right Front speakers to Primary Listening Position (“PLP”): 8’6”.

Center speaker: (1) Distance from front of tweeter to front wall: 1’10”; (2) Distance from center of tweeter to both side walls: 6’8”; (3) Distance from center of tweeter to Primary Listening Position (“PLP”): 8’1”.

Today, I moved the Left, Center and Right speakers further away from the front wall. I also moved the Left and Right speakers further away from the side walls. Here are their new measurements:

NEW:

Left and Right Front speakers: (1) Distance from front of each tweeter to front wall: 3’ 0”; (2) Distance from front of each tweeter to side wall: 2’9”; (3) Distance from center of Left Front tweeter to Right Front tweeter: 8’1”; (4) Distance from center of tweeter for both Left and Right Front speakers to PLP: 8’1”.

Center speaker: (1) Distance from front of tweeter to front wall: 2’8”; (2) Distance from center of tweeter to both side walls: 6’8”; (3) Distance from center of tweeter to Primary Listening Position (“PLP”): 7’3”.

Here is a graph of the Center speaker in the new position with Audyssey off (green) and on (gold)(The measurement procedure was the same as in my first post:

NEW CENTER  AUDYSSEY ON & OFF 4_17_2016.jpg


Audyssey is clearly decreasing the major excursions from a smooth response, but is still putting a "hole" in the important vocal frequencies.

Here is a graph of the Center speaker in the new position, with the mic aimed at the speaker, with Audyssey on (green)and off (red):

NEW CENTER MIC HORIZ AUD ON & OFF 4_17_2016.jpg

Comparing the traces for the vertical vs the horizontal mic orientation, there seems to be better response in the upper frequencies. Wayne suggests this may be due to the reflection absorption panel attached to the ceiling between the center speaker and the PLP. But, I'm not sure what to make of this -- is the absorption panel helping or hurting?

This is the measured response (mic back to vertical) of the Left speaker with Audyssey on and off:

NEW FR LEFT  AUD ON AND OFF 4_17_2016.jpg

Here is a graph of the response (mic still vertical) of the Right speaker with Audyssey on and off:

NEW FRONT RIGHT AUDYSSEY ON & OFF 4_17_2016.jpg

This graph is a comparison of the Center speaker's response, with Audyssey ON in both traces, with the difference being the speaker location. Clearly, moving the speaker further into the room has flattened the response in the 200Hz -- 1kHz range and from 2.5 -- 5kHz.

CENTER NEW VS OLD POSITION 4_17_2016.jpg

I welcome additional comments and suggestions. However, it seems moving the front and center speakers further into the room may provide additional improvement.
 

·
Premium Member
Joined
·
12 Posts
Discussion Starter #9
Have you tried the dialogue enhancing feature on the Denon ?
Thanks for the suggestion.

Yes, I have tried that feature. According to the menu, it increases the dialogue by 2.5 dB. So, when the enhance feature is engaged, the distorted audio is louder than when it is not engaged. I am able to understand slightly more of the vocals, but the distortion is just louder.

However, I am hopeful about a technique I discovered tonight that is available on my Denon AVR. I can copy the Audyssey EQ settings (or at least the Audyssey level settings) to a manual EQ. Then, I can adjust the manual EQ levels from -6 to +6 (I'm not sure if these numbers are dB or just relative settings). After increasing the levels in the 1.5 to 2.5kHz range, the vocals sound better. Tomorrow, I will run some REW scans with the higher manual EQ levels and see if the measurements correspond to my perception.
 

·
Registered
Joined
·
12 Posts
If I may chime in on this....other owner of Dennons have had similar complaints on other sites about their center channels as well (as have Marantz owners). Aside from tweaking with EQ (which I am a fan of), you may try adding another speaker from above the screen to have dual center channels. Alternatively, you may want to look at a different center channel altogether that will still blend well with your mains but has a different tweeter design for more intelligible voice output. Both of these solutions worked for me and others facing similar issues. I am now running 2 centers and one of them a different speaker design than the rest. Best of luck and welcome to the Shack.
 

·
Banned
Joined
·
4,838 Posts
I skimmed through the previous posts, beg pardon if someone else already suggested this.

The calibration mic pattern you have used probably ended up showing Audyssey a peak in the 1.5 to 2.5 kHz range, and Audyssey has compensated for it with the dip you are hearing and measuring.

Precisely what calibration pattern are you using with Audyssey?
 

·
Premium Member
Joined
·
1,562 Posts
> Audyssey 'Reference' setting includes a 3dB dip around 2kHz. Audyssey 'Flat' does not have this dip.
> The Denon center 'channel level' can be increased as needed manually if the 'dialog level adjust' is not enough.

I didn't see these other 2 options mentioned above, but I may have missed them.
 

·
Premium Member
Joined
·
12 Posts
Discussion Starter #13
....other owner of Dennons have had similar complaints on other sites about their center channels as well (as have Marantz owners). .
Thank you so much for this information. Since I have been compulsively testing, changing speaker locations, and modifying acoustical treatment in my home theater for days, I can’t tell you how relieved I am to know that perhaps my problem is in the Denon AVR.

Frankly, I have always assumed the problems that are so evident in the vocal playback, as confirmed by my many REW scans, were my fault (i.e., poor speaker selection, poor speaker location, improper acoustical treatment, improper test technique, etc., etc.)

Before I consider buying yet another center speaker, either to replace my new one, or to supplement my current speaker, I am going to try to further isolate the problem.

Perhaps my problem is in the audio amplifier section of the AVR. To check this, I am going to connect the “pre/out” signal from the AVR’s center channel to a separate Rotel amp. After adjusting the levels, and re-running the Audyssey program, it will be very interesting to see if the problem is improved.

Thanks again, and stay tuned ….
 

·
Premium Member
Joined
·
12 Posts
Discussion Starter #14
> Audyssey 'Reference' setting includes a 3dB dip around 2kHz. Audyssey 'Flat' does not have this dip.

Thank you for the reminder about Flat vs Reference. I will make sure to always use the Flat Audssey setting.

> The Denon center 'channel level' can be increased as needed manually if the 'dialog level adjust' is not enough.

I didn't see these other 2 options mentioned above, but I may have missed them.
That's a great idea to use the "channel level" to supplement the "dialog adjust." I tried that technique last night while watching a movie. Unfortunately, it made the vocal distortion more pronounced. With no "dialog adjust" or increased "chanel level," I had to reduce the volume just to understand what was being said!

If you get a chance, please read my next post where I propose a new theory for this distortion problem.
 

·
Premium Member
Joined
·
12 Posts
Discussion Starter #15
I skimmed through the previous posts, beg pardon if someone else already suggested this.

The calibration mic pattern you have used probably ended up showing Audyssey a peak in the 1.5 to 2.5 kHz range, and Audyssey has compensated for it with the dip you are hearing and measuring.

Precisely what calibration pattern are you using with Audyssey?
Thank you for taking an interest in my problem. I really appreciate all of the work you put into your post on how to properly calibrate using Audyssey.

I will be more specific in my description of the mic positions used during Audyssey MultEQ XT32 calibration. I am using the following pattern (taken from your post):

"8-Point Basic Setup Mic Pattern - MultEQ XT32, MultEQ XT, MultEQ (6 pts). Use for step 2 of this Process. All dimensions are relative to PLP Center (PLPC):
1. PLPC
2. 3 inches forward
3. 3 inches up
4. 3 inches up & 3 inches forward
5. 3 inches left
6. 3 inches right
7. 6 inches left
8. 6 inches right"

Please take a look at the following REW scans showing the Center speaker with Audyssey (Flat) On and Off. The way I interpret these results is the relocation of the center channel further into the room has virtually eliminated the cut in the 1.5kHz to 2.5kHz range. However, please notice the roll-off of the frequency response from about 2kHz upward. While Audyssey improves this roll-off, it is still far from ideal.

TH CENTER SPEAKER MEASUREMENTS 4_21_2016.jpg

This post is getting long, so I will start another post with my new theory of the cause of my distorted audio coming from my Center speaker.
 

·
Registered
Joined
·
969 Posts
Is it muddy sound or just low volume?

If it's low volume
Another thing to consider is that Audyssey may have reduced the level of your center speaker compared to the other speakers in your system. Wayne's point about running Audyssey from just one listening position and moving the mic only slightly has some merit. I have an older Denon 4311ci and I do find the center channel's volume is low after running Audyssey. If I break out my SPL meter my fronts tend to be in and around 75dB and my center is 73dB from my main seating position. Going around the speakers and bumping up/down the channel levels makes a huge difference.

If it's muddy sound:
The 2D graphs don't take in to account the time domain, ie: reverberation or transient decay time (RT60). Looking at the pictures of the room there is a lot of absorbers on your front wall but it's hard to tell how thick they are. So I have lots of questions.
How thick are the absorbers and what material are they made of?
Do you also have them on the first reflection points on your side walls?
What other room treatments do you have?
Is your room in your basement?
If so:
Do you have a subfloor with insulation or is it carpet with underpad over concrete?
Are your walls stuffed with insulation from floor to ceiling?

What I'm thinking is that your RT60 over the frequency spectrum may be uneven. High at some frequencies and low at others. You might also have reflected sound reaching your ears from first or second reflection points (but not at all frequencies) smearing the sound. If you do a waterfall graph in REW it will give you a better understanding.
 

·
Premium Member
Joined
·
12 Posts
Discussion Starter #18
For days now, I have been racking my brain, and my back muscles, in trying to eliminate the distorted audio I am hearing from my new center speaker connected to my new Denon AVR. I tried to watch a new Blu-ray movie last night, but I couldn’t get into the film because the vocals from the center speaker were so distorted – and the volume of the AVR was set at about -20dB!

Due to some great advice I received on this forum, I have improved the frequency response of the Center channel in a previously depressed range from about 1.5kHz to 2.5kHz. However, with the Audyssey calibration On or Off, the frequency response I am measuring at my PLP rolls-off almost linearly from about 2kHz on up. (B/T/W, Wayne noticed this in a very early post on this thread, but I failed to fully appreciate his insight.)

My first thought was my Umic-1 was not responding appropriately in the higher frequencies. Fortunately, I have a second Umic-1 and was able to compare the two. While I was very impressed by how closely they tracked in response to the same audio signal, my theory was shot down. I concluded there is nothing wrong with the mic I am using.

My next thought was to suspect the acoustic treatment I have in my home theater. Until yesterday afternoon, I had a 4’ by 8’ acoustic absorption panel at the first reflection points on both side walls and on the ceiling. I also had much thicker absorption across the back wall. With some significant effort, and some back sprain, those panels are now removed. But, removing those panels had no effect on the high frequency roll-off, as can be seen in the following REW scans:

TH CENTER SPEAKER MEASUREMENTS 4_21_2016.jpg

My next idea was to apply some manual EQ to the system to supplement what the Audyssey calibration was doing. Instead of starting with a fresh manual equalizer with all of the settings at 0dB, I was able to use the copy feature of my AVR to transfer the Audyssey EQ boost/cut levels to the manual equalizer. My plan was to modify the manual EQ to try to have a more flat frequency response in the mid and high frequency ranges.
When I opened the manual EQ, with the copied Audyssey settings, I was very surprised to see Audyssey was already boosting the frequencies from 2kHz on up. Here is a photo of the manual EQ from the Denon AVR as displayed on my plasma TV:

EQ SCREEN (1 of 1).jpg

And, here is a close up view of the same image:

EQ SCREEN CLOSE-UP (1 of 1).jpg

As you can see, the Audyssey is already boosting the following frequency bands by the maximum of 6dB:2--16kHz. It has been my understanding that applying an EQ, especially in the mid and high frequencies, should be limited to cutting, not boosting (although, I have read a few authors who don’t think a judicious boost of up to 3dB is that bad). If my math is correct, a 3dB boost requires a doubling of audio output power, so 6d6B requires a factor of 4 increase in power.

So, my current theory is some vocal distortion is caused by the diminished mid and high frequency levels. But, I think asking the AVR to boost a broad band of frequencies by 6dB may be resulting in clipping by the AVR’s amp (especially if there is a large demand for power by other channels, and the center channel vocals are loud). Perhaps it is clipping that is the major cause of the distortion.

My next step is to take the Center channel pre-out signal and run it through a separate audio power amplifier. I will post again once this has been done.

Thanks for sticking with me through my own “odyssey.” As usual, your thoughts, ideas and suggestions are most welcome.
 

·
Premium Member
Joined
·
1,562 Posts
Possibly the CC tweeter is bad? You could place the mic about 18" in front of the CC. If the HF response still rolls off >5dB with Audyssey turned off then the tweeter is bad.


You can also look at your current CC data on the REW 'distortion' chart to see if the distortion is really high at some frequencies. If you post an .mdat file of the CC we could take a look also. Maybe there is a hint there.
 

·
Premium Member
Joined
·
12 Posts
Discussion Starter #20
If you don't mind can you let me know what type of stand are you using for the center speaker ?
Thanks for your interest. Then center speaker stand is Sonus model number SFC18.

Please let me know if you have any other questions.
 
1 - 20 of 27 Posts
Top