Auto speaker setup: A less than perfect solution?
You may be able to make much better sound without auto setup.
(Credit: Steve Guttenberg)Auto speaker setup and calibration is a popular feature on almost every receiver and a lot of home-theater-in-a-box systems.
Sure, it sounds like a peachy idea, but the accuracy of auto setup is hardly a sure thing; and at their worst, auto setup systems sound worse than no setup at all.
Ideally, the setup system automatically determines speaker sizes (large or small), measures speaker-to-listener distances, sets the volume levels of all of the speakers, determines the proper subwoofer volume level, checks that all the speaker wires' "+" and "-" connections are properly oriented at the speaker and receiver ends, and calculates the subwoofer-to-speaker crossover point. Some receivers also employ EQ (equalization) curves to correct for speaker and room acoustic anomalies.
What's not to like? Well, it the auto setup worked perfectly, nothing.
But they're mostly flawed: Subwoofer calibrations are almost always off. Auto calibration systems boost the sub volume much too high, and overestimate the sub distance to the listener by a factor of two (so a 10 foot distance becomes 20 or more feet).
Worse yet, auto setup systems rarely set the subwoofer-to-satellite speakers crossover frequency to the optimum point. That is, they tend to set the crossover too high, say 150 Hertz, which unnecessarily restricts the speakers' bass response. The speakers might sound better with a lower crossover setting. I recommend 80Hz for all speakers with 4- to 6-inch woofers; 100Hz for 3-inch woofers; and higher settings of 120Hz or 150Hz only for the tiniest speakers.
Accessing the measurement data post auto setup can be tricky on some receivers. Then you really don't know what you have.
Thing is, manual setup isn't all that difficult and will likely be more accurate. And chances are you wouldn't muck up the distances as poorly as the autosetup would. Running the test tones over the speakers and manually adjusting the sound by ear or with a Radio Shack meter isn't so hard to do.
The EQ programs, such as the various Audyssey systems found on Onkyo and Denon receivers; Yamaha's Parametric Room Acoustic Optimizer; and Pioneer's MCACC Multi Channel Acoustic Calibration might improve the sound of your speakers, and you will have to run the autosetup to access the EQ feature. But again, please don't assume the EQ will always sound better than no EQ. I usually prefer the sound with the EQ turned off. But turning it off isn't so easy, it's frequently buried on a hard-to-find menu.
My advice: Perform the manual setup and see if you're satisfied with the sound. If not, go ahead and run the autosetup, and see if it's an improvement. Just be aware it may not be so easy to undo. Consult the owner's manual and see if it offers information about how to reset the receiver to its factory default settings.
If you've done the manual vs. auto setup, tell us what you heard.
Steve Guttenberg is a frequent contributor to magazines and Web sites including Home Entertainment, Playback, and Ultimate AV. He is a member of the CNET Blog Network, and is not an employee of CNET. Disclosure. 
They should all be used at the very least for getting levels and distances matched, and they all work well for those two items. Your comment about them getting the subwoofer distances wrong is lacking in knowledge. The auto-setup may report the physical distance incorrectly, but they are actually reporting the acoustical delay correctly. This acoustical delay is the combination of the physical distance as well as any processing delay in the actual subwoofer. Typically, subwoofers have a low pass filter in them that adds group delay causing the distance reported by the auto-setup to be more than just the physical distance. Thus, to get correct alignment with the other speakers, you need this to be measured. Tell me, how would you do this with a tape measure? It's not going to work.
As for levels, using a traditional method of playing pink noise and using an SPL meter will definitely get you inaccurate results due to room modes and the inaccuracy of the meters when measuring low frequencies only. This is often +/- 6 dB or more off. Thus, the manual method is useless. The correct way to match the subwoofer level to the satellites is to measure the frequency response of each and align such that you get as close to flat across the spectrum of the sat and sub. According to Audyssey on the AVS forums, their algorithm does this automatically through its time and frequency response measurements. I am not sure about Pioneer, Yamaha, and other home-grown solutions. If you are not fortunate enough to have Audyssey, then I would recommend listening to a piece of music you are very familiar with that covers a wide frequency spectrum, particularly down in the low frequencies to match your sub to sat levels, but I would be impressed if you did much better than +/- 3 dB with this method.
The bass management crossover setup with Audyssey has also worked flawlessly in the few systems I have used it in as well as according to most people who have used it. If your bass management system only supports a single crossover for all speakers, then yes, it may likely suggest a higher crossover because your surrounds or center speaker are smaller and don't play down very low. Or maybe your room has a mode that is cause a big suck-out around 80 Hz and thus a higher crossover is recommended. But this is a trade-off between having localized bass or big suck-outs in your bass response because your satellites can't reproduce those frequencies. Which would you rather have? I'd rather be able to hear my bass with some slight localization. And with the Audyssey systems I have used, you can force it to a lower crossover if you'd like this trade-off to be different than its suggestion. Audyssey optimizes the response based on your selection of this.
And EQ is also hugely important to get right. It can make the difference between bad sound and good sound or good sound and great sound. But yes, some auto-setup systems don't get this right. If for instance you measure just a single position in your room and EQ it for that, then I could see the EQ being wrong because your head is not the size of a pin and is not clamped down in a vice to that position. The response of a room changes dramatically from position to position even just a foot or so apart, and therefore it is important to correct over a larger area. Doing single position EQ could make other positions worse and can lead to "unnatural sound" as you describe. Audyssey's MultEQ is by far the best EQ system I have ever used in this regard, however, and has the most intelligence when it comes to using multiple measurements to correct the sound in a room. And I don't care how great and wonderful you think your speakers are, there is no room that is without some sort of acoustical problems at least in the low frequencies. And you can't solve all room problems with acoustical treatment because of the length of the low frequency sound waves. Thus, you need Audyssey MultEQ. And I doubt your speakers are perfect either, in which case Audyssey MultEQ will help make them better.
All said, I think you are doing a huge disservice for the home theater market by suggesting that one should try and setup their home theater without the auto-setup first. It should be the other way around. Do the auto-setup first, especially if your system has Audyssey, and then if there is something you don't like, go in and fix it. Surely if you have any clue how you might make it better, you probably also have the wherewithal to figure out how to navigate through the system's convoluted menus.
Don't worry, if Steve has been keeping up with current developments in room correction, and has invested a credible effort to use the tools that Audyssey provides, I suspect that he already knows the value of Audyssey over trial and error manual methods.
It's just that he's got a blog to run and stirring the pot to generate controvery is simply an easy way to generate traffic to his site. ;-)
It is a good place to start. After setting it up I had to change the sub crossover because my speakers (Monitor Audio RS8 x5 and 15inch home made subs x2) sound better at a lower crossover. It also adjusts the subs volume a little low so I had to boost it a few db. The 3808ci has several eq settings and I use the Audyssey eq for movies only. I like more bass with music so I adjust each speaker individually after the auto setup and make it a second eq setting. For vinyl, I usually don't use the eq at all. Fortunately the remote has one button that switches between eq settings easily.
Having worked with A/V equipment professionally for many years, I was surprised at how well the auto setup sounded on my Denon when I tried it. But like all automatic things, it didn't know my specific preferences and needed tweaking.
My 5.1 set up consists of a pre-Audyssey NAD T-173 pre/processor, mid-80s Adcom amps (modded, one stereo, the other 3 channel), and JMLabs speakers (Daline 6.1 fronts, Chorus 700 series center and rear). A Radio Shack meter has helped calibrate speaker distance, but otherwise only the sub's volume and crossover frequency can be adjusted (60hz works best for me, the Dalines are full range, and of course the NAD piece has to be set to crossover at the same frequency). The only way to flatten the frequency response would be to use room treatments.
The stereo system is more high-end (Melos preamp, BAT monoblocks, A/P Virgo 3 speakers, stereo Vandersteen subs) and I've had extraordinarily good results with a (stereo) Lyngdorf RP-1 processor. The set up is in a difficult room -- irregular shape, hard surfaces, practically no back wall . . . but I gotta use the room I got. There were room-related bass suckouts no matter where I placed the speakers. Test tone measurements with a Radio Shack meter only confirmed that. Room treatments weren't an option (in the living room), so I tried the Lyngdorf piece and whew (!) . . . much greater transparency, incredible 3-D soundstage, wow there's the bass(!), and all-around going further into the music. It goes completely against my "audiophile" grain: taking that precious analog signal, doing an A/D conversion, they processing it like who-knows-what, and then convering it back D/A . . . aaagghh "the horror" . . . but the proof is in the listening.
- by myoda August 4, 2009 3:37 PM PDT
- Steve:
- Like this Reply to this comment
-
(16 Comments)I found the caibration process in the Onkyo 806 to be very accurate. I have a "difficult" room with hardwood floors and pocket doors. Audyssey MultEQ balanced the room - speaker distances were correct.
The process works properly if you use a tripod, or mic stand with a boom attachment - and RTFM. Read the friendly manual. This guy has some great tips:
http://www.audyssey.com/blog/
Thanks