Audyssey MultEQ Comparison To Manual Audio Calibration

 

Sencore Audio Consultant – Audio Tool For Manual Audio Setup And Evaluation

 

I recently purchased a Marantz receiver with Audyssey MultEQ built-in.  I thought it would be interesting to compare the results from manually setting up this receiver using professional tools to Audyssey since it is so popular.

Observations And Comments

My general impression of the setup that resulted from running Audyssey was that it sounded harsh in comparison to the manual calibration which sounded natural when listening to normal program material.  This reminds me of many comments I hear and read about people’s home theater systems sounding harsh.  I suspect many of these problems could be removed by setting up the preamp or receiver properly using a manual method.

Audyssey missed the phase (wiring) of both the front speakers and the center speaker.  This is very troubling.  That type of error will cause a loss of bass at the crossover and will result in speakers not integrating well at the higher frequencies as well.  These speakers combine to form a system and each must be in phase with the others to work together properly.  I have found that this is not an uncommon thing to find in systems that were setup with automated audio calibration systems.  Unfortunately, it is important that you have some method to determine what the absolute phase of your speakers and subwoofers really are.  I find a very high percentage of speaker and subwoofer products are shipped with the red and the black terminals or the 0 degree phase setting being incorrect.

The differences in channel distances from the manual and automatic calibration were not great enough to alter the results dramatically.  However, I have seen cases where Audyssey misses phase and will then have radically incorrect distances for the subwoofer.

Channel levels from Audyssey did not come close to reference levels at 0 db on the volume control.  The error was -5.5db for the center channel which is the reference channel for movies since it is the source of dialog.  5.5 db louder requires 260% more power to the speakers.

I am sure if I provided Audyssey with the two subwoofers not setup properly it would have been easy to obtain very poor results for the LFE channel and bass management.  The increased error in the frequency response from Audyssey is disappointing since it degraded the subwoofer performance from what it would have been without any equalization being applied in the receiver.

It should also be noted that these automated receiver setup systems will do nothing to fix a vast array of problems that are external to the receiver.  These problems include room acoustics, speaker/subwoofer placement, subwoofer settings, failed drivers, loose connections, loose components, poor wiring and rattles.  These issues are just as important as the settings in the receiver, but are frequently missed without careful attention to the system.  For this test all of these problems were solved for the receiver, but in the average home they may not be.

Test Notes

System:

Main Speakers: Hales Revelation 3 (three way with a 10″ woofer)

Center Speaker: McIntosh CS350 (three way with two 6.5″ woofers)

Surrounds: Klipsch RS42 (two way with two 4″ woofers)

Subwoofers: Revel B15 (15″ driver) and James Audio EMB-1000 (10″ driver)

Receiver: Marantz NR1402 (not normally in this system)

The room is acoustically treated to around 230ms RT60 at 500Hz with six seats.  All of the speakers were positioned and wired properly before the test started.  All seating positions in the room were measured with Audyssey.  The primary seat was measured as the primary in Audyssey and used for the manual calibration and comparisons.  Both subwoofers were originally set manually and are equalized and balanced independently of the receiver.  No settings were altered on the subwoofers between Audyssey and the manual receiver adjustments.  All Audyssey measurements were made at 0 db to avoid issues with level based equalization altering the results.

Speaker Phase Results

Audyssey detected both of the front and the center speakers as being out of phase.  All of the results from Audyssey were conducted with the phase of these speakers swapped as indicated by Audyssey.  Manual adjustment and an absolute phase meter showed Audyssey’s request to change the phase of these speakers was wrong.  The manual calibration was done with the front three speakers in phase instead of 180 degrees out of phase as set by Audyssey.

Speaker Size Results

Audyssey set the front speaker to large and all others to small.  Manually all speakers were set to small.  Audyssey not crossing over the low frequencies to the subwoofer for the front channels increased distortion and resulted in less low frequency output for the front channels.  It also did not take advantage of the substantial power savings on the receiver by using the amplifiers in the subwoofers for the front speakers instead of the receiver.

Bass Setting Results

Subwoofer Mode was set to LFE and the LPF for LFE was set to 120Hz for both Audyssey and manually.  These are the proper settings for these parameters in this system.

Distance Results

Audyssey        Manual

Left Front                9.5ft               9.4ft

Front Right              9.3ft               9.1ft

Center                     8.8ft               8.4ft

Subwoofer               9.3ft               8.0ft

Surround L               8.6ft               8.9ft

Surround R               8.3ft              8.0ft

Channel Level Results

With the receiver set at 0 db the output volume level from the speakers should be at reference level.  The following table shows the deviation for each channel as measured by my reference microphone and preamp.  This measurement was made using the Goldline Audio Toolkit disc limited bandwidth pink noise which is very accurate.  The data shows the channels were all low and error from the correct level at the prime seat was as much as 2.0 db from the center channel signal where dialog sets the reference.   A 2 db increase in sound requires 60% more power to the speakers.   The channel imbalance could easily be the result of averaging multiple seats and may be desirable if the goal is to more evenly balance the levels in the theater, but it will not be the best setup for the prime seat.

       Error Measured From Prime Seat

Front L                                        -5.0 db

Center                                         -5.5db

Front R                                        -3.0db

Surround R                                  -4.2db

Surround L                                  -4.7db

Subwoofer                                   -3.0db

Crossover Frequency Results

The crossovers were set to 80Hz for the center and surround speakers by Audyssey.  Manually the crossover for the center and surrounds were set to 80Hz while the front channel was set to 40Hz.  Audyssey not crossing over the low frequencies to the subwoofer for the front channels increased distortion and resulted in less low frequency output for the front channels.

Audyssey Settings

When Audyssey was complete the MultEQ was set to Audyssey, Dynamic EQ On, Reference Offset 0db and Dynamic Volume to off.

Equalization Results

The following data shows how the LFE channel equalization was not as flat at the primary seat as the system was without Audyssey engaged.  The subwoofers had previously been optimized and equalized and this was left in place for Audyssey.  The nominal level of the LFE signal without Audyssey deviated significantly more between 50Hz and 60Hz than without Audyssey.

This is the LFE channel with a 20Hz to 80Hz test signal from Goldline with Audyssey processing active.

This is the LFE channel with a 20Hz to 80Hz test signal from Goldline without Audyssey.

 

The average frequency response as measured using the Avia five channel pink noise signal showed that the average frequency response with manual equalization was also superior to Audyssey.

 

Average system frequency response using Avia asynchronous wide band pink noise and SMPTE 222M house curve with only manual equalization in the receiver.

 

Average system frequency response using Avia asynchronous wide band pink noise and SMPTE 222M house curve with Audyssey engaged.

 

 

W. Jeff Meier

ISF and THX Certified Home Theater Consultant

15 Comments
TV

This seems wildly inaccurate and self-serving.
-The subwoofer distance is irrelevant, it is simply used to adjust the phase.
-The rest of the speakers were very close assuming your measurement was somehow more accurate.
Also, many 3 way speakers run 1 driver intentionally out of phase to create width to the soundstage. This will always cause issues with Audyssey.
-The harsh sound thing is an opinion and some people don’t like the EQ sound, it eliminates the room. That could be the reasoning for calling it unnatural.
-You can manually adjust all the speakers to small after if for some reason it shows up as large. Almost anyone with a decent subwoofer should do this.

You mostly just addressed, Volume Leveling for each speaker and Speaker Phase. What about how Audyssey goes through each Speaker’s response and helps to tame peaks and valleys in the response, this makes a particular difference in the bass.

W. Jeff Meier

My observations were based on about $7,000 worth of measurement equipment along with experience from post production and home theater rigs. There is nothing wildly inaccurate about my observations. They were taken with great care using the highest quality equipment I can find. The equipment I am using, for example, is what was used to setup the audio post production for the movie Transformers. These are also very consistent with what I have seen in hundreds of Audyssey setups in homes around the country.

The subwoofer distance is completely relevant. Look at the large frequency dip at 80Hz with Audyssey engaged versus the settings I input. That is the result of poor phasing from using the wrong distance and speaker phase.

The speakers in my system do not run drivers out of phase. I have an absolute phase meter and test material that I have used to verify this for each driver. Following Audyssey suggestions lead to the large frequency dip at 80Hz shown in the attached plots.

It is a common misperception that EQ is eliminating the room effects. My room meets the Dolby certification for room acoustics. There is nothing wrong in this room to eliminate. Poor room acoustics in general smear the sound with reflections from odd directions it is not primarily a frequency problem.

I did equalize the system manually. I just did not show those parameters out of brevity. In fact the subwoofer EQ was active in Audyssey and they just made things worse. The other channel manual EQ was in the receiver and is not active when Audyssey is active. The plots show my EQ was far superior to Audyssey assuming an SP495 and an Earthworks M30 using a calibrated compensation curve are accurate.

I am absolutely in favor of equalization. What I find with Audyssey is that is does it poorly as shown in this example along with a loss of bit depth or coarsening of the data leading to a grainy sound quality.

I struggle to see how you reached your conclusions after reading this article. I suspect you have been deluded by a very effective marketing campaign by Audyssey.

I see no other way to educate people on this unless they do this themselves or hire me. If that is self-serving in your opinion so be it. In my opinion I am trying to be informative.

Drew

I am wildly lost on how to apply a top end calibration of a warm Marantz NR1402 hooked to 10 to 20 k of speakers in a room with acoustic treatments with the Audyssey equalization turned off. Am I missing something?

I do agree a manual set up is much better but that is my preference. I used a contractor tape measure and my 7 year old kid and 50 USD volume meter from radio shack at 80 db reference.

I tried two systems.

Legacy Marantz Sr7003 with direct coax inputs from Panasonic bdt500 bluray with with burrbrown decoder to def tech 8060st, 8060hd, and 8080bp.

Results were far superior to modern denon receiver with audyssey xt with Sony HDMI bluray.

I used same crossover and life settings. Almost not required with 8080bp rears though.

I too played transformers, but referred to an expert, yeah the kid again.

So in short the 75 receiver on eBay with the black Friday sale on the Panasonic bdt500 was better than the 900 receiver with the modern HDMI out bluray for 350.

I am now an audio fool like you. More expense with less result.

I have to say the vt50 series plasmas from Panasonic are the exception. I’m keeping legacy maramtz with direct line out till it dies.

D Bone

I also think Audyssey XT sounds harsh in my room when compared to setting the manual EQ. XT seems to lose detail at the expense of making the mid-range too prominent. I don’t have near the experience you have, so what I ended up doing was running several 8pt calibrations all with differing mic placements, then I averaged all of that data by using the “Copy Audyssey Flat” function to load the EQ so I could see what was happening and then I input it to the manual EQ section.

From there is used my ear to make subtle changes, and the result is a far less harsh, more detailed sound with less exaggerated bass. Audyssey did get the SPLs and delays right for me, so I guess it’s not totally useless.

zoldar

excellen article. You are demonstrating with your measureents what I have been hearing and saying for years. I too felt adsy multeq sounded harsh an had a tipped up sound in the high mid region.

I now use an amp with adsy 32xt multi eq. and it gets it right for movies for me. but for muti-channel music i leave it of and opt for a manual eq using rew and a mic stand to measure output. I ended up with a cross of about 70 hz (i have a massive 15 db 63 khz room mode) and some eq in the bass.

could you do this review with a higher end receiver? also, pushing a marantz with a tiny power output to 0 DB? isnt that pushing it too far?

W. Jeff Meier

I work with many of the higher end receivers. I am not gathering detailed before and after settings data when I am working. I just fix what I can. The flaws with Audyssey and these others are always significant as shown above. The consistently superior solution is to avoid these automated tools and do the audio calibration manually.

0 dB is going to typically deliver rated power from a product if the channel levels are set to 0 dB. 0 dB is typically used in these applications to indicate reference sound levels on the volume. It may overdrive the amplifier in some situations when more channels are driven, but it will depend on the source material and equipment.

zoldar

I get your point. I did some measurements to see how well adsy 32xt did integrate the sub for 2.1 Stereo DSP mode. That was a disaster chart with huge suckouts en still some rome modes present.

What I don’t get is if adsy has a greater degree of db range to correct with than the manual EQ. Manual EQ gives me a range +-6db on several frequencies. But If I set the cross to 40hz like adsy does, I don’t have enough DB range on the manual EQ to sufficiently flatten the room modes.

If I start with a cross of 70-80KHZ it is possible to get a good response without touching the high and mids, which adsy forces upon you.

I’m goint to read some more of your guidance on this site as I think you are in the right track that automated systems don’t get it right.

I also have a question: with what program do I generate a tone sweep out of each channel(r,l,c,surr, surr back and LFE)? With REW software I only seem to get from the left or right channel only.

W. Jeff Meier

I am certain I am right about this. I recently designed and setup a screening room for Focus Features. The head of post production said he never heard such good sound or saw that accurate a picture even at Dolby.

I can say I have never seen a consumer get it right with REW. I have little confidence in that approach. I have audio setup tips on this site you might find helpful.

I am not a fan of using programs to generate test tones. I believe it is best to use a disc that goes through the whole signal path. I use Goldline, AVIA, THX and my own disc for this. The software, mics, preamp and house curve are also key. My gear is shown here https://www.accucalav.com/thx-audio-calibration/.

Funky Monkey

There is a correct way to set up Audyssey. Once the auto calibration is complete, DO use an SPL meter or app to equalise the volume levels for each channel.
Do set up the speakers to a crossover other than that detected if the system sets your speakers to large, if you have a subwoofer.
And lastly, if you have a “dumb” subwoofer like mine which is a decade old, DO change the level of that manually using a song you are familiar with that has a suitable bass line.
Cannot comment on phase.
Also, follow your manual’s instructions for microphone placement during calibration, including the order of seating positions. Some positions that are recommended are not even seating positions.
Just cos something sounds harsh does not make it inaccurate. It may be unfavourable, but that’s just personal taste.I like a bright sparkly sound. To a degree.

W. Jeff Meier

I do not find that to get the best results from a receiver. I would read the page on setting up audio on this site for details on this.

Franco

I’m far too lazy or poor to afford setting my system up manually with top notch calibration equipment. I have Audyssey MutlEq using six positions, and with Dynamic EQ on and Dynamic Volume at Medium. I was running films for a while without the filters, and I recently activated the Audyssey target curve in my Onkyo Receiver to ‘Movie’. One thing I immediately noticed on a film I’d previously watched without the filtering turned on, is that the panning and directionality of sounds in the satellite speakers are now much more defined, nuanced, and precise than they were before. Panning in particular sounds far more precise and detailed than before, and the LFE’s from my two subwoofers blend in far better with the mid-range than they did with Audyssey off. Plus, spoken dialogue from the centre speaker is much less murky sounding and much more clearer than before.

I’m not challenging the supremacy of manually calibration if you are committed, have some knowledge and money to calibrate your equipment, but I personally think Audyssey has done a fine job of calibrating my system in my room. The only alterations I made were to altering the centre and front left and right speakers from 90Hz for the centre and 80Hz for the left and right to 120Hz.

W. Jeff Meier

All I can say is I rework Audyssey calibrations many times a month with great success. I never have a client who wants to go back. It may have improved your situation, but it did not get anywhere close to what is possible. Most people have very low expectations for how their system can sound.

John Rolin

Too bad you are selling something. Otherwise, I would find you believable.

Leave a Reply

Your email address will not be published.