My equalization/room look ok?

[thewire]

I did some testing at various volume settings on my receiver up until the setting was at 0. It has been setup using the auto setup and a digital SPL meter. I re-calibrated the SPL meter with REW for the final test that is the longest one shown. Any advise, questions or room treatment suggestions would be really great. I could also manually adjust my equalizer settings if you see where I should do that. I didn't use any filters. A hush box fan was running nearby above the mic and my computer was behind the couch in the back of the room. When I played nothing it kept reading the background noise at around -20db. If it looks wrong I will do it again but I would have to bring my computer in there again. I plan to once I figure out how to do a waterfall. I copied and pasted my notes and images bellow.















GEAR:

RADIO SHACK DIGITAL SPL METER (STEREO HEADPHONE JACK LEFT OUTPUT)
PIONEER VSX1012 THX SELECT RECIEVER 90 WATT x 6 AT TWEETERS
LIBERTY THX/MONSTER/BETTERCABLES CABLES
5 MARANTZ MA50 THX (EARLY GENERATION) MONO AMPS (+1 ON RESERVE)
DENON 5200 THX ULTRA 2 CHANNEL AMP

CRYSTAL AUDIO 12THX SUB
6 POLK RTI70 FLOOR STANDING
POLK CSI40 CENTER CHANNEL (TWEETER MOD)

ROOM:

CONCRETE/FRAMING/1 LAYER DRYWALL
13.5 X 20 X 7'4"
OUTDOOR CARPET NO PADDING SPEAKER PLATFORMS, STAGE, FLOOR, RISERS, HUSH BOX
REAR LEFT & RIGHT REAR SPEAKER PLATFORMS ARE BASS TRAPS
CEILUME CEILING TILES, GENERIC CEILING TILES @ LEFT AND RIGHT WALLS DE COUPLED WITH CARPET
NO MORE THAN 12" OF GENERIC TILE WIDTH (WIDTH VARIES)
ADJACENT ONE CLOSET & LOBBY WITH SEALED SOLID CORE DOOR
AC VENT /WALL HEATER/HUSH BOX FAN
TWO ROWS OF THREE SEATS

PROJECTION NEC 6PG+ CRT, TOSHIBA HD A30, FOUCS ENHANCEMENTS CENTERSTAGE 2,
106" DALITE HIGH POWER



MANUAL ROOM TEST RESULTS:


31HZ-35HZ @ 85DB CEILING [MODERATE DISTORTION ISSUE]
70HZ-80HZ @ 101DB INTERNAL CEILING STRUCTURE [MINOR DISTORTION ISSUE]
45HZ DOOR [MAJOR DISTORTION ISSUE] (CORRECTABLE BY PULLING TIGHTLY SHUT)
APROX 47HZ @ 93DB INTERNAL BACK WALL [MINOR DISTORTION ISSUE]
57HZ-60HZ @ 98DB CEILING [MINOR DISTORTION ISSUE]
60HZ-63HZ @ 98DB CEILING [MODERATE DISTORTION ISSUE]
63HZ-67HZ @ 98DB [MINOR DISTORTION ISSUE]

MODERATE DISTORTION ISSUES MAY BE AUDIBLE
DURING MUSIC/MOVIES (OTHER FREQ MAY MASK THE ABILITY TO HEAR IT)


MORE:

USB VOLUME IS AT 0 - (BUT VARIED) EQUALIZER SETTINGS ARE SET AT ALL ADJUST- SETUP IS AUTO SETTING
ALL DB READING/FREQ RANGE ARE APROX AND WERE DONE USING THE SPL METER,
SPL METER MIC, AND REW (CONTINUED SEE TXT FILES AND IMAGES)

OCT 14, 2007

[Otto]

Hi there,

Your room looks great! But can you change your plot ranges? Usually, we're primarily interested in sub response because it's the most accurate to measure and what you'll be addressing with any parametric EQ. Can you change the axes to 15-200 Hz on the horizontal and 45 to 104 dB on the vertical?

Thanks and looks good!

[thewire]

I reloaded my saved files figured out how to change the view and replaced the images. Sorry about that.

[thewire]

the waterfall



room treatments? Change my crossover? Turn my sub up or down?

[brucek]

Tremendous amount of noise in your impulse response. What does the impulse response look like when you measure a simple loopback cable? This will tell us if the computer is capable of a proper reading.

Always set the graphs to logarithmic - not linear.

Do a response with the sub only and the mains shut off. That way we can see if it tracks a decent target.

Always test with the receiver in stereo with no effects or surround modes on.

brucek

[thewire]

brucek wrote:

Tremendous amount of noise in your impulse response. What does the impulse response look like when you measure a simple loopback cable? This will tell us if the computer is capable of a proper reading.

Always set the graphs to logarithmic - not linear.

Do a response with the sub only and the mains shut off. That way we can see if it tracks a decent target.

Always test with the receiver in stereo with no effects or surround modes on.

brucek

Something to do with my receiver and True HD audio made the power in my house go out twice possibly. The first time I had just switched audio to True HD. The second time I was turning down the volume about 10 seconds into the movie. Could be some sorta strange coincidence also. When I turned it on it had gone to a different input selection. I think I need to reset my 3 breakers (two dedicated circuits) to the theater every month much like we are supposed to for the hot tub. Or I am just loosing my marbles. Now that I know the left output works for the mic I went back and re-calibrated using that. The before results were read using the left with the mic but my auto-setup was done using the right on accident. I used the right because it would not give me a ambient noise error. I ignored it this time and the results were much more sensitive. It set my speakers to large. I will redo the test tonight and watch some movies to compare the difference.

Also. Yes I had it set to dolby digital not stereo. I will try with the mains off as well. Crossover is at 50HZ. I will also set the graph to logarithmic as soon as I figure that out. the button pictured was not shown in my version and I had downloaded REW many months ago so I will download the latest version as well. Thank you for the tips.

[thewire]

Using the tips I got I made corrections to the layout and made further notes. I also corrected the ceiling height. These results are now done both using the same input as with the setup. I did not turn off my mains because a 50hz crossover would limit these results. See the notes for the color key. With these results I hope to make improvements.



all_energy_time.jpg


all_impulses2.jpg


all_mesured2.jpg


average.jpg


filter_adjust2.jpg


scope.jpg


waterfalll_standard2.jpg

- notes

GEAR:

RADIO SHACK DIGITAL SPL METER (STEREO HEADPHONE JACK LEFT OUTPUT)
PIONEER VSX1012 THX SELECT RECIEVER 90 WATT x 6 AT TWEETERS
LIBERTY THX/MONSTER/BETTERCABLES CABLES
5 MARANTZ MA50 THX (EARLY GENERATION) MONO AMPS (+1 ON RESERVE)
DENON 5200 THX ULTRA 2 CHANNEL AMP

CRYSTAL AUDIO 12THX SUB
6 POLK RTI70 FLOOR STANDING
POLK CSI40 CENTER CHANNEL (TWEETER MOD)

ROOM:

CONCRETE/FRAMING/1 LAYER DRYWALL
13.5 X 20 X 7'6"
OUTDOOR CARPET NO PADDING SPEAKER PLATFORMS, STAGE, FLOOR, RISERS, HUSH BOX
REAR LEFT & RIGHT REAR SPEAKER PLATFORMS ARE BASS TRAPS
CEILUME CEILING TILES, GENERIC CEILING TILES @ LEFT AND RIGHT WALLS DE COUPLED WITH CARPET
NO MORE THAN 12" OF GENERIC TILE WIDTH (WIDTH VARIES)
ADJACENT ONE CLOSET & LOBBY WITH SEALED SOLID CORE DOOR
AC VENT /WALL HEATER/HUSH BOX FAN
TWO ROWS OF THREE SEATS

PROJECTION NEC 6PG+ CRT, TOSHIBA HD A30, FOUCS ENHANCEMENTS CENTERSTAGE 2,
106" DALITE HIGH POWER



MANUAL ROOM TEST RESULTS:


31HZ-35HZ @ 85DB CEILING [MODERATE DISTORTION ISSUE]
70HZ-80HZ @ 101DB INTERNAL CEILING STRUCTURE [MINOR DISTORTION ISSUE]
45HZ DOOR [MAJOR DISTORTION ISSUE] (CORRECTABLE BY PULLING TIGHTLY SHUT)
APROX 47HZ @ 93DB INTERNAL BACK WALL [MINOR DISTORTION ISSUE]
57HZ-60HZ @ 98DB CEILING [MINOR DISTORTION ISSUE]
60HZ-63HZ @ 98DB CEILING [MODERATE DISTORTION ISSUE]
63HZ-67HZ @ 98DB [MINOR DISTORTION ISSUE]

MODERATE DISTORTION issues may be audible to the human ear
during music/movies (other frequences may mask the ability for them to be heard)


MORE:

USB volume @ 0 - Equalizer settings are set to all adjust - Setup is auto-setting
All db reading/frequency ranges are approximate adn done using a digital SPL meter,
it's mic, and REW version 4.0 (See images)

COLOR REFERENCE:

Red

-12db @ receiver
-.04db high @ check level
SPL meter @ 75db uncorrected

Green

-18 @ receever
-0.3db high @ check level
78db @ SPL meter uncorrected

Blue

-0db @ receiver
-0.1db high @ check level
SPL meter @ 87db uncorrected

Purple is average


OCT 15, 2007

- end notes

[JohnM]

The scope plot shows heavy clipping on the input signal - didn't REW give a warning about that? It only shows the last measurement taken (was that with the 0dB volume?) so others may have been OK.

[brucek]

I did not turn off my mains because a 50hz crossover would limit these results.

I don't really understand that. If you shut off the mains and do a sweep to 200Hz with the sub only it will separate where problems lie between the mains or sub. Since the sub is the only device to receive equalization, it's the starting point. Do a sub only sweep and use logartithmic graphs (press the little ICON button at the top right hand side of REW called Freq Axis).

Once you have equalized the sub, then add the mains and work on the crossover transition problems that may arise. That is likely the problem dip at ~60Hz in your plots. Adjust the sub phase control to reduce that peak. It is likely a result also of having the mains set to large. Usually small is a better choice.

As John says, you're clipping your signal into the soundcard. The results are not useful until you solve that level problem.
Look at the SCOPE tab and you'll see the clipping after you've taken a measure.
In the SCOPE tab, the smooth sweep is the output of the soundcard and the dark red envelope is the input to the soundcard with the Y axis is the percentage of digital full scale.

Here's a SCOPE of a normal sweep. See the dark red does not hit the light red.

SCOPE full no clip.jpg

If I expand the vertical scale I see a nice input waveform of input and output signal of REW.

not clipped.jpg

Here is what it would look like with the signal clipped - not good.

clipped.jpg


In the impulse response the harmonic distortion components appear as additional impulses at negative time. You have a lot of distortion showing. Do a cable only loopback measure to see if your computer and soundcard are contributing to this problem.......

brucek

[thewire]

brucek wrote:

I don't really understand that. If you shut off the mains and do a sweep to 200Hz with the sub only it will separate where problems lie between the mains or sub. Since the sub is the only device to receive equalization, it's the starting point. Do a sub only sweep and use logartithmic graphs (press the little ICON button at the top right hand side of REW called Freq Axis).

Once you have equalized the sub, then add the mains and work on the crossover transition problems that may arise. That is likely the problem dip at ~60Hz in your plots. Adjust the sub phase control to reduce that peak. It is likely a result also of having the mains set to large. Usually small is a better choice.

As John says, you're clipping your signal into the soundcard. The results are not useful until you solve that level problem.
Look at the SCOPE tab and you'll see the clipping after you've taken a measure.
In the SCOPE tab, the smooth sweep is the output of the soundcard and the dark red envelope is the input to the soundcard with the Y axis is the percentage of digital full scale.

Here's a SCOPE of a normal sweep. See the dark red does not hit the light red.

Attachment 4585

If I expand the vertical scale I see a nice input waveform of input and output signal of REW.

Attachment 4584

Here is what it would look like with the signal clipped - not good.

Attachment 4586


In the impulse response the harmonic distortion components appear as additional impulses at negative time. You have a lot of distortion showing. Do a cable only loopback measure to see if your computer and soundcard are contributing to this problem.......

brucek

It seems then perhaps I would be better off giving my soundcard a try instead of USB. I thought the best idea was to get a basic understanding of how well a job my receiver is doing at equalization on it's own first before I make manual corrections. I could turn my volume down to like -50 and the clipping would stop.

cnet wrote:

Sporting the latest revision of Pioneer's autosetup program, MCACC (Multi-Channel Acoustic Calibration), the Pioneer VSX-72TXV balances the satellite speakers' volume levels, measures speaker-to-listener distances, determines the various sizes of the speakers, and equalizes the frequency response of the speakers in your room. This Advanced MCACC also applies "time axis" calibrations that distinguish between the direct sound of the speakers and the room reverberation to produce a more accurate sound. Using the MCACC system is quite simple. Just plug in the supplied microphone, place it at your main listening position, bring up the setup menu, and press the Enter button on the remote. A few minutes later, the Pioneer VSX-72TXV will be tuned to your speakers and room.

One problem is that I have digital SPL meter as a mic and not the supplied one. Another problem is my amplifiers are so much more powerful than the internal amps (90w vs 125w) because I also use a bi-wiring and separates configuration to provide the best quality audio possible. It is over 400ft of wire hence the name I choose "thewire". It sets every volume level to -10db except the sub no matter what mic I have tried. The method it applies to my sub is "bass management" not equalization. This is adjustable using what is called Bass peak level.

cnet wrote:

The next level of bass management came along in 1990 with the Lucasfilm THX standards, which required each speaker to produce 105dB of bass output at the listening position to match the film-studio experience. Realizing that a speaker capable of producing such a prodigious amount of bass would be impractical in all but the rarest of rooms, THX opted for a basic form of bass management to reduce the requirements for speakers. The bass from all the main channels was redirected to one or more subwoofers, allowing the speaker system to produce plenty of bass-kickage in the room with main speakers the size of most "bookshelf" speakers.

The final development of the full concept for bass management occurred when Dolby Digital and DTS discrete 5.1-channel digital formats were released to the consumer market circa 1995. Both formats included an additional channel, called LFE (Low Frequency Effects) dedicated to high-volume bass sound effects that required a playback system capable of producing 115dB of bass at the listening position. To optimize the reproduction of low frequencies for these new formats, bass management would include:

1. A test system to determine the frequency extension and headroom of every speaker (including subwoofers) in a sound system;

2. A frequency-agile crossover filter for each speaker to ensure that it received only what it was able to reproduce;

3. An intelligent redirection and resource-pooling system to optimize the bass headroom of the entire system;

4. A bass peak-level limiter for each speaker to prevent bass from overloading it;

5. A room-correction system to optimize the bass output of a system by analyzing a room's standing-wave resonance and boundary-reflection patterns and applying correction filters where appropriate.

Unfortunately, bass management in its fully developed form never made it to the consumer market in 1995, and is still not available today. Instead, less sophisticated versions that include parts and pieces of the master plan were implemented in AV controllers, with the version required by THX for its licensees being the most common.

The Basics of Bass Management
What we find today in most AV controllers is a form of bass management that includes a highpass (low-cut) filter for each main speaker (front left & right, center, side left & right, and rear left & right if applicable), a corresponding lowpass (high-cut) filter for the subwoofer output, a redirection function for bass below the main speakers' highpass filter frequency, and a bass peak-level limiter.

Unless there are extenuating circumstances, experts agree that all main speakers should be highpass filtered in rooms where the length and width dimensions are less than 40 feet. Standing-wave resonances in such rooms make speakers impossible to place for both smooth bass response and imaging accuracy. Applying highpass filters to the speakers frees them up to be placed for imaging accuracy while not adversely affecting their bass, which is played by subwoofers that can be positioned for smoothest bass response. The highpass filters also release the main speakers from the intense responsibility of playing bass at high volumes, as mentioned above.

The main-speaker highpass filters and subwoofer-output lowpass filters should be set according to two factors: room resonances and the bass-reproduction capabilities of the main speakers (see Fig. 1). Ideally, the user would tell the controller to apply a filter to each speaker at the best frequency for that speaker. In the real world, controllers often limit users to either a global variable frequency that affects all speakers or a fixed frequency for all speakers. Controllers with a fixed frequency for all speakers offer the option of Large or Small speakers, with the Large setting removing the filter and the Small setting applying the filter. If the controller offers frequency-agile filters, then the filter frequency is selectable in addition to the Large/Small setting.

You want me to use my soundcard? Should I use a less quality mic that will not pick up background noise or clip for my auto-setup? Should I buy the actual setup mic? Should I turn of my equalizer and test that way? I could do this the same as people are used to if needed. I will take a photo to better show what I am talking about and post it in a few minutes.

[thewire]

The photo of my incorrectly setup system. Please help me use REW the right way.




Edit: Please note that I have saved the faq page to my computer now for reference while I am using the program. Also note that I can not adjust my USB input volume. Please see my previous post for my questions. Also read my "notes" section. Thank you.

[thewire]

JohnM wrote:

The scope plot shows heavy clipping on the input signal - didn't REW give a warning about that? It only shows the last measurement taken (was that with the 0dB volume?) so others may have been OK.

Yes the db measurement was the clipping result for each setting. ta

[brucek]

I don't really know if your Auto EQ system is correct or not. I assume it works fine. These auto eq systems usually target wide bandwidth corrections (and rightly so) and at frequencies usually above subwoofer problem areas. I see the lowest correction in your photos is 63Hz.

I'm not sure what you're trying to accomplish. The response graphs you have posted are fine other than a dip at 60Hz. As I said you could probably correct a lot of that with sub phase or a bit of sub distance setting.

REW is usually used to correct subwoofer resonance problems. It is used to measure subs and mains speakers (not surrounds or centers). Use it only with the receiver in stereo. It matters not what power your amplifiers are. You set the level at your listening position to 75dB and set REW up for 75dB targets.

brucek

[Doug Plumb]

To check if your EQ is OK take a new set of measurements with slightly different mic locations and see if your EQ improves those.

I think EQ is best applied between about 100 Hz and 5 KHz. You can really improve the mid band response of almost any speaker if you know what you are doing.

The manual for RPlusD downloadable from the link explains this in the chapter on Practical Measurements.

[thewire]

brucek wrote:

I don't really know if your Auto EQ system is correct or not. I assume it works fine. These auto eq systems usually target wide bandwidth corrections (and rightly so) and at frequencies usually above subwoofer problem areas. I see the lowest correction in your photos is 63Hz.

I'm not sure what you're trying to accomplish. The response graphs you have posted are fine other than a dip at 60Hz. As I said you could probably correct a lot of that with sub phase or a bit of sub distance setting.

REW is usually used to correct subwoofer resonance problems. It is used to measure subs and mains speakers (not surrounds or centers). Use it only with the receiver in stereo. It matters not what power your amplifiers are. You set the level at your listening position to 75dB and set REW up for 75dB targets.

brucek

I want to accomplish something in the range of my adjustments. -10 for every speaker is wrong. -10 is as low as it can be adjusted. That tells me it could be wrong for the equalizer I thought.

I want to use REW to know what is wrong.

So my auto-setup works? REW will do better than my db meter?

I posted the above because people are confused.

[thewire]

Doug Plumb wrote:

To check if your EQ is OK take a new set of measurements with slightly different mic locations and see if your EQ improves those.

I think EQ is best applied between about 100 Hz and 5 KHz. You can really improve the mid band response of almost any speaker if you know what you are doing.

The manual for RPlusD downloadable from the link explains this in the chapter on Practical Measurements.

Thank you.

[Doug Plumb]

You are welcome

[brucek]

I think EQ is best applied between about 100 Hz and 5 KHz. You can really improve the mid band response of almost any speaker if you know what you are doing.

Equalization corrections in this area are extremely localized and so should only involve very low Q filtering. Most people agree that room treatment is the answer for the 100hz to 500Hz range. Wide filters (such as tone control) can help at higher frequencies.

REW will do better than my db meter?

Oh for sure. Even though REW is mostly used for setting up parametric equalizers in the 20Hz to 100Hz range (where equalization is most effective), you can certainly take a response measurement up to ~5KHz with a Radio Shack meter. This will give you an idea if the auto eq system has done its job with its low Q filtering.

As far as the -10db levels that your auto eq system is providing - I simply have no idea. Sorry. I'm sure you can over-ride and set them back to zero....

brucek

[Doug Plumb]

Equalization corrections in this area are extremely localized and so should only involve very low Q filtering. Most people agree that room treatment is the answer for the 100hz to 500Hz range. Wide filters (such as tone control) can help at higher frequencies.

This is speaker EQing I am talking about and it can be done and the differences in many cases are obvious even for top end speakers.

At lower frequencies you can EQ out some sharp resonances such as I had at around 140 Hz, it gave the room an unusual sound be once that peak was smoothed out the sound really improved - in some measurements you can't see the peak and get a dip for applying the EQ but this is OK because sharp dips are barely audible.

EQing speakers can be done but for frequencies above about 1000 Hz. This is room independent.

[thewire]

brucek-

Yes I can manually adjust settings like the speaker level without compromising others. Based on what I have been able to post do you think room treatments may be a good idea?

Doug -

I will give that program a demo sometime. Perhaps what I need is to input my mic into the computer and calibrate it to the correct volume using REW somehow and then output the mic input of my soundcard into the mic input of the receiver. Would this work? Would it harm my receiver?