Will someone please explain crossover frequency to me. Here are my current settings after doing an auto calibration: Front R/L = 40Hz, Center = 90Hz, Surrounds = 150Hz, Sub = 80Hz but the manual setting on the sub is 150Hz.

I have read that the manual setting on the sub (EP800) should be set to 150 as this takes the the manual settings out and the receiver settings take over. In my case 80Hz. True?

I have also read that every speaker should be set to 80Hz. True?

I'm confused how crossover frequency works in a practical way. The way I understand it is that if the sub is set to 80 then it will play all frequencies 80 and below but what about all the other speakers? I guess they play all frequencies above there current settings but not below?

Should I leave all my current settings from my auto calibration or change all speakers to 80Hz?

Thanks!

It looks like auto calibration set each speaker to it's audible range. I think that this would work out well, if it sounds good to you. I have mine all set at "Small" which, I think, sets them all to 80 hz. and with a sub covering 20 to 80 hz. It may be that leaving them set at a lower level with a sub, might produce too much bass, which I have experienced.

da-drifter,

I would suggest you do some experimenting. I am not a fan of auto calibration as it has never worked well in my own experience and I have been privy to a number of problem sounding systems that were fixed by disabling it. This is not to say the lots of people are not having good success with them. I have to assume they are as they are quite popular. I would listen to your system for a while as you have it now utilizing what the auto calibration has set and the EP800 x-over set to 150 Hz. Then try shutting off the auto calibration and just setting all the channels to an 80Hz x-over point in your processor and turn the x-over on the EP800 to 80Hz. You will need an SPL meter capable of measuring bass accurately in order to set the sub level. After listening to this for a while determine which you like better. I would then try setting the M80s to a 40Hz x-over setting in your processor. You may find this evens the bass out in your room as the M80s are now quasi acting like they are two more subwoofers in your room between 40 and 80 Hz. After all this experimenting go with what you felt sounded the best.

I will try to be as simple as possible. The crossover frequency for each speakers is the point where the low frequencies are handed off to the sub. It is not just cut off at that frequency, however, there is a slope or gradual handoff to the sub.

What you call the crossover setting for the sub after running setup (80hz) really isn't a crossover setting, it is really what is called a low pass filter setting. Not sure why receiver mfg's are including that these days. Anyway, you can keep that at 80hz, or even bump it up to say 120hz. Even the folks at Audyssey say there is some movie material that goes above 80hz.

Back to the speakers, as I have mentioned in other threads. Audyssey measures each speaker in "your room" and looks for the
-3dB point for each speaker and reports that back to your Denon receiver. Denon desides if a speaker is going to be called large or small, in this case they follow 40hz as a rule. So any speaker above that is called small, and speakers like the 80's can often be set to large because they can produce material down that low.

When you have a sub, you still want to set all your speakers to "small" and adjust the crossover accordingly. I would try bumping up the 80's to say 60-80hz and see what you think. The other speakers look fine to me based on what your saying.

You never want to decrease the crossover setting after running Audyssey, as that gap won't get corrected by their filters.

What you say is true about the freq knob on the sub, you want to get it out of the way, since your receiver is handling the management for each speaker.

Just to clarify a point, changing a crossover frequency for one or more speakers following an auto-calibration done by any of the several systems available doesn't require that the auto-calibration be shut off and a possibly less accurate manual calibration be done. The option to change the crossover frequency can be used and the speaker levels and distances set by the auto-calibration can remain in effect.

What the guy who designed your speakers said!

What the pro-auto calibration people are not telling you is that auto calibration does not “hear” things like the human ear.

The Audyssey calibration normal curve (not the “Flat Curve) rolls off the high frequencies to compensate for an imbalance in high frequencies in movie soundtracks that can occur when playing what was intended for a large movie theater in a smaller room where the reflected vs direct sound can behave differently. There are potential problems with this. First this isn’t always necessary especially when sitting close to the speakers or in a heavily treated room. Also most music sound engineers don’t master for large venue playback. The result can be a noticeable loss in high frequency (above 10kHz) playback especially in music, but also in movies if your setup isn’t susceptible to the imbalance caused by small venue playback. My room is well treated so in both cases the Audyssey Curve kills the ambiance in movies and especially music for me. Made even worse if you are older like me and your high frequency hearing is diminishing. Auto calibration will not take this into account. It’s one of the reasons I can instantly hear and don’t like the sound of the Audyssey curve with most quality music, and some movies because it kills the high frequencies which my ears are already doing.

Audyssey applies midrange compensation, an intentional dip in the 2kHz region, where most tweeter-to-midrange crossovers are. The thought, correct or not, is that this is a weak point in speaker performance which Audyssey attempts to nullify by intentionally not presenting the playback as the sound engineer intended. The idea is to null out the harshness that can be caused by both the tweeters and midrange drivers operating near their respective lower and upper limits.

Auto calibration also doesn’t take into account that you don’t hear things as well from behind as you do from the front. It will set your rear speakers to the same relative level as all the rest which means unless the sound engineer boosted the rear channels to account for that they will not present an equal soundstage with the rest of your system. I generally find a +3dB boost to the rear speakers brings most soundtracks back into balance in my room.

Also, in general applying EQ above 200 Hz is considered problematic. However, in the bass range it can make noticeable improvements which outweigh any problems caused, especially below 50-60Hz were room treatments start to fail, though it’s still AFAIK considered best to balance your bass with multiple sources and room treatments saving EQ for last. The main problem being is that it’s hard to EQ out a null. Even the Audyssey bass manager, SVS AS-EQ1, I have can only boost a null +5dB while it can reduce a peak -20dB.

Another problem is you simply can't create an EQ curve for multiple spaced seating nor can you correct the time domain for all seats. You can do dam good if EQing for one seat however once you move beyond a single seat the EQ curve must be an average of what “best” for all the seats. I found in testing the AS-EQ1 that this could often make one side seat sound and measure better at the expense of my other two seats, one of the ones always being made to sound worse is the center, sweet spot. OTOH if I do subwoofer EQ for only the center seat it comes out as close to perfect as things can get.

Auto calibration systems can be great for people who more or less want a sound system that is essentially “plug-and-play.” However, if you want to get the most out of your system you should experiment with as many different variations as you have time and energy for. Over time as you train your ears by listening to your system you will develop a much better understanding of how to set it up so it sounds best to you, in your room with your equipment. OTOH if you don’t want to be bothered by experimenting or tweaking the auto calibration will likely give you an acceptable sound. But if you don’t explore other options you will never know if you are getting the most out of your system.

Audyssey, or it’s spokesperson on the internet, has written that it comes down to “Reference vs Preference” which IMO is misleading. In several areas I mentioned above Audyssey admits to tweaking the sound in a way that deviates from Reverence. So ultimately you really have an issue of Preference vs Preference and the only why you can determine what you prefer is by experimenting.

Note that I am not Audyssey bashing. I usually use Audyssey for movies because it comes on automatically if I engage my wide speakers. In my experience what I gain using the wides far outweighs what I loose using Audyssey for movies. OTOH I never use it for music, not even the flat curve. I may use Dolby to get mutli-channel and even occasionally kick in the dolby height speakers for music but never Audyssey . . . my preference.

Dean, the original post(and my reply)dealt with crossover frequencies. This relates to the calibration function and has nothing to do with the separate room EQ function.

Hi John. I don’t believe I quoted you or commented on anything you said. I quoted Ian because I was expanding on what he was saying about auto calibration since it often goes hand-in-hand with the topic of room EQ which Randy noted.



Though I should correct myself in that I said Audyssey sets the rear speakers to the same level as the rest when I should have said that the auto calibration feature does this.

About Audyssey killing the high frequencies. Is this a feature you can turn on or off. I know with MCACC on the Pioneers it is. It's called X-CURVE. X-CURVE slowly rolls off the high frequencies. The default setting is off. I've never tried turning it on, but I might give it a go, because at higher volume in both music and movies, the high's can be a tad too much at times.

Yes. Depending on the brand like Denon you can directly select the “Flat” Curve with no roll off. With my Onkyo the only way to get the Flat Curve is to select one of the THX modes and turn off re-eq.

You should give it a try since the roll off can be beneficial in some rooms.

grunt,

Your explanation is excellent. I wish I had written it myself. I am not sure I would want to boost my surround channels by 3 dB but it is a great example of using the manual calibration settings to customize things to your liking and your room. I am a big fan of living with your various experiments for a while to experience different content being played before landing a final decision. Adjusting channel volumes and x-over points is not affecting the tonal balance and therefore not causing the sort of potential damage that things like Auto Room EQ can cause, certainly above 300 Hz. One of these days we will take some actual measurements of what Auto EQ devices are doing to the amplitude response. It would be interesting to see the effects, measured in the anechoic chamber, after the correction had been applied in a room.

The above said I do think JohnK brought up a valid point though. If you don’t have, at a minimum, a decent SPL meter to play around with the manual calibration settings, the auto calibration may be the safest bet.

i use the Audyssey "Flat" curve, since my room is treated and if i didn't use "flat", the high frequencies would be weak and lacking.
"flat" is just right to my ears.

careful setup like this is what permits me to listen to movies at around 0dB without cringing; it's like in the cinema, wide, crisp, clean, smooth and "natural".

i listen to the music stations on Galaxie at around + 2dB; this gives me a realistic level on those stations i listen to.

I have probably asked this before - If I have not, I'm sure someone else has. I have M60's and an EP 500. The crossover on the 500 is all the way up - 150 hz, I think. Crossover on the receiver is set to 80 and is not configurable by speaker. Is there any harm in setting the mains to Large rather than Small?

Any harm setting them to large will be largely dependent on the amplifier, source material, volume (which is largely dependent on your seating positions and personal loudness preferences).

Having the speakers set to large with highly dynamic blu-ray soundtracks could cause stress or compression due to some low-frequency content at loud volumes being sent to the mains or possibly inducing amplifier clipping damaging the tweets. Mid to low frequencies is what could really cause the current demand to ramp up, particularly problematic with a run of the mill AVR (clipping, overheating). Having the speakers set to large will require more power than having them sent to small. Your EP500 with its own dedicated internal amplifier and woofer designed to handle content between 20hz to 80hz is better suited than the mains.

Thanks Ian!

Just to clarify I only boost the rear speakers by +3dB but leave the side surrounds at the same level as the front speakers.

For the untrained ear (meaning most people) I agree that auto calibration is better than nothing if someone doesn’t have an SPL meter. However, in my experience once a person hones their critical listening skills they can get pretty close to the same settings as auto calibration or an SPL meter gives for sound level (distance at some frequencies if you’re really good) by ear alone. Though I totally understand that for most people auto calibration is far more practical.

A perfect example of the ear sometimes being better than auto calibration was when in my apartment I invited over a friend who use to master demo tapes for local bands. When I demoed the opening battle in Master and Commander he immediately said that my subwoofer distance was off. He adjusted it by ear and I ended up with the best sounding crossover from the mains to the sub I have ever had.

One of the reasons I keep bringing up this topic is that I fell prey to thinking that auto-cal and auto-EQ would do all the work for me and because they are machines will be more accurate than my hearing. Only by accident did I realize that speaker/seating placement and eventually things like room treatments played a much greater role in the quality of sound I’m getting from my system than whether or not I was using EQ or the auto calibration settings. I think the best example of this is that despite all the science and experience Axiom puts into their speaker designs you still put things through double blind listening tests to make sure it works in practice.

Another time I only realized by accident what furniture did to in room sound when I added a futon (previously just had a matrass on the floor) and suddenly had all the “life” sucked out of my music. Later I found that by elevating my mains it returned much of the ambiance I had lost when I put that giant sound absorbing futon in the middle of my room. Most people will never experience this because the room they put their speakers in is likely laid out and furnished before adding the speakers.

In more general terms living abroad especially in “third world” countries I learned that human perception is far more powerful than most Westerners (steeped in the “scientific method”) realize. We’re taught to trust our instruments and distrust our senses. I’m a trained interrogator and can say without a doubt that I can detect a lie better than any polygraph machine I’ve seen. There is a reason El Al Air Lines uses trained interviewers as their primary means of detecting terrorists boarding airplanes . . . they work better than baggage and body screening machines. Pacific islanders were able to navigate between islands by being able to feel the reflected swells on their legs hanging in the water coming off islands as far as 30 miles away. The list of seemingly amazing things human perception is capable of is almost endless but most people, especially Westerners disregard them because it’s what our cultures teach us. Yet audio, especially music, is IMO very much about perception. (steps off soap box)

Those measurements of what auto EQ does would be very interesting to see.

Cheers,
Dean

I can attest to this about BD audio. The new Star Wars Blu-rays kept overheating my Onkyo despite the mains being set to small but with a 40 Hz crossover. So they don’t even have to be set to large in my experience. Note that this happened to a receiver that I tested at 100dB for one and a half hours playing very bass heavy trance music.

My main reason for using the auto calibration features, isn't for speaker playing levels. It's for setting the frequency response for each speaker to gain a more balanced tonal match for the room. I've adjusted furniture around the room and haven't re-calibrated. I need to do that.

Apparently it's useful to point out yet again that Drifter wanted help with "Crossover Frequency Confusion". That confusion about crossovers isn't likely to be lessened by instead discussing the separate and much more complex question of room eq. Randy's reply covers the question asked quite well.

Rob, on your crossover related question, it'd be probably be too much to suggest that "harm" would result from setting the M60s "Large", when a sub as good as the EP500 is available. However, by not setting a crossover for the M60s advantage wouldn't be taken of modern bass management which has been paid for in the sub and HT receiver. The capability of the EP500 would be wasted in part by having the M60s play low bass frequencies that the EP500 is designed to do a better job of. Cleaner M60 response above the crossover is one benefit, less load on the amplifier is another.

On the amplifier load point, however, it should be pointed out that this is sometimes exaggerated in the form of claims that bass frequencies are inherently "power hungry". Of course, basic audio technology is that power requirements are related to loudness and unrelated to frequency. For example, playing a 400Hz note at 80dB requires ten times the power that playing a 40Hz note at 70dB does.

The bottom line is to set even large tower speakers with an appropriate crossover, most likely 80Hz.

Thanks for all the responses - I am going to switch over to small on the mains and as "Ivan" suggested to "High Plains Drifter" listen and experiment.

I don't think it is that simple. If you factor in the speakers impedance (assuming a constant phase angle for now), that is directly related to frequency. So what would demand more current? Let's say a speaker has an impedance of 4 ohms at 400hz and 40hz and playing at 80dB. But volume isn't constant either. Movies and music are filled with dynamics. Dynamic peaks in blu-ray movies especially range from the midbass to subsonic frequencies.

But if the object was to state just the lack of correlation between frequency alone and power requirements, it really is that simple. The reason for my comment above was that I've seen some discussions of bass frequencies that look on wavelength as if it was a physical object that had to be "pushed" by the power. So, a 40Hz wavelength of about 28 feet(1130 feet per second/40)would need more power to push it(because it's "bigger") than a 400Hz wavelength of about 2.8 feet. Actually of course, wavelength only represents the distance the wave travels until the next wave cycle starts(again 1130/40 equals 28 feet)and has no relationship whatever to required power.

If a different factor, impedance, is to be discussed, this of course has a relationship with current, but less than is sometimes imagined. Using Ohm's Law(current equals square root of power divided by impedance)in my example above and if 1 watt is used for 400Hz(at say 8 ohms)the current for that is .353 ampere(square root 1/8)and the .1 watt used for 40Hz(at say 4 ohms)results in a lower current of .158 ampere(square root .1/4). If both frequencies were played at 80dB(and both 4 ohms)the current for each would be the same .5 ampere(square root 1/4).

As to dynamic peaks, obviously they require more power, but the same basic principle of audio technology applies: the power is determined by the loudness at the peak and has nothing to do with the frequency at that point.

I set X-CURVE to cut high frequencies only at -1db, and I don't konw if it's only psycological, but I swear that it sounds much better now. The highs aren't nearly as bad as before.