Axiom Home Page Forums General Discussion Technical Questions Speaker sensitivities / calibration question

Newbie question here. I was wondering about this. Since I don't have identical speakers all of the way around and they have different sensitivities, when you listen to something at anything other than the level you calibrated your speakers at, do the speakers' levels get further off the further you get from the reference volume? Is this perhaps some of what Audyssey's Dynamic EQ takes care of? It's a little embarrassing that I don't know the answer to this question after this long in the hobby, so please enlighten me in the past, if you have a time machine. Maybe tattoo the answer on my baby body.

Well, I think reading Audyssey's page on Dynamic EQ did help me understand it a little better. Different speaker sensitivities is one of the things taken into account for the technology. I can grasp why it would be an important compensation resulting in better sound according to most people. I'm leaning towards Audyssey Dynamic EQ in my next processor or receiver, even though I'm still curious about Trinnov.

I'm hoping JohnK will illuminate me on why conventional calibration is enough and that this doesn't really matter.

No, no, JohnK loves electronic calibration. It's Alan that hates it.

Yep, JohnK has positive information about this technology. I love Audyssey as well, my ears don't lie....

Ha ha, yes, but JohnK is an even bigger fan of saving money than he is of Audyssey.

Just don't ever listen at anything other than reference level.

Good call. I like your simple solution.

I had a plan at one point to just wire power amps (with gain controls) directly to the multi-channels outs of my DVD player. Figured the gain trim would allow me to calibrate for individual speakers, and then I'd just play everything at reference level. No need for a pre-amp or any sort of volume control.

Actually if had a setup only for movies, I could probably get away with something like that. But CDs are mastered at a higher "reference" level (or with no reference at all), that I have to turn them down about 12 dB from where I listen to movies.

Charles, when a calibration is done, whether manually or by an auto-calibration system, the adjustments made to the receiver trims for the various speakers compensates for differences in their sensitivities as well as factors such as different locations. No, once the difference in sensitivity has been taken into account, that difference remains the same at higher or lower sound levels than the one at which the calibration took place(typically around 75dB)and the compensation still applies; otherwise there'd be little value in bothering with calibration, since it would only apply to one volume level.

The usefulness of Audyssey Dynamic EQ arises because of the lesser sensitivity of our ears to bass and treble at lower volumes, together with a lessening of surround effects, but this has nothing to do with the difference in speaker sensitivities.

Yay, someone with a real answer!

Thanks, JohnK. So the receiver basically figures out the speaker's sensitivity by the trim coupled with the distance settings? Is that right? Otherwise, I don't see how it works. Calibrating at one volume level wouldn't really show the receiver the speaker's sensitivity, since distance would be a factor in the volume, right? So does it, in fact, use both trim and distance figures to create a curve for the speaker's sensitivity, or am I still missing the boat?

Charles, the volume levels of the various speakers are equalized. It doesn't matter whether the correction is due to a difference in sensitivities, a difference in speaker location, or whatever. The receiver doesn't specifically care what the sensitivities are; it just wants the speakers to be equally loud at the listening position.

Yes, and that's why I'm at a loss as to how it can accurately predict how much juice to send a speaker at any volume other than reference, if it doesn't know the speakers's sensitivity. Is it really a straight line in output, not a curve?

As was mentioned in my first reply, the difference in sensitivities doesn't change with changes in volume levels. There's no need to "predict" anything. Whatever the incoming voltage from the source material is at a given instant in time, it's subjected to the same gain(around 14dB for the preamplification and 28dB for the amplification, increasing the voltage about 120 times)and results in the speaker being audible.

The difference in sensitivities doesn't change. You're right. But if not every speaker in the system has the same sensitivity, then how is the output matched correctly at volumes other than what you calibrated at? I realize I'm probably just being super-thick, so I appreciate your patience.

The trim settings in the receiver vary how much of the voltage coming in from the source material is let through to be preamplified and amplified to each speaker(the receiver "volume" control varies this uniformly to all speakers). For a more sensitive speaker, or one closer to the listening position, or whatever else might make it louder, the trim reduces the incoming voltage in its channel enough so that after the uniform gain that applies to all channels takes effect, it's no longer louder than the other speakers. This takes place at all "volumes", i.e., all levels of voltages coming in from the source material. Since the difference in sensitivities doesn't change with volume, there's no "matching" problem at different volumes.

It just seems like sensitivity and distance should act on a speaker's output to different degrees. The way you explain it makes it seem like the sensitivity and distance affect the volume the exact same amount. Is that right?

CV, it is magic.


That is all.

Finally, someone put it in layman's terms. Ha ha.

No, it doesn't matter whether it's sensitivity, distance or something else, and how much each contributes to the difference; the only factor of significance is sound level at the listening position.

I'm going to accept this since my brain is too tired to continue. I have no doubt you're right, since it would be too big of an oversight for the entire industry if it wasn't right, but I still can't quite see how it doesn't matter how much each contributes if you're listening at volumes other than reference.

No, the receiver (or more accurately, the calibration process) determines the trim based on the speaker's sensitvity and distance.



Again, it's not trying to figure out sensitivity or create any type of sensitivity curve, it's trying to determine the trim setting based on distance and speaker sensitivity.

The amount of juice sent to a speaker is not predtermined or predicted in any fashion whatsoever. The amount of juice sent to a speaker is a result of the speaker's sensitivity and how loud it is being played.

To illustrate why changing volume doesn't affect the calibration, try and think of EQ'ing certain frequencies. When you tweak certain frequencies to suit your tastes, your changes all remain in tact when you move up/down in volume. Same applies to the calibration found in receivers.

That still doesn't explain how the levels for the speakers stay consistent at all volume levels. It just tells me that they won't stay consistent because of differing sensitivities that won't be taken into account. This is making me feel seriously dumb. I'm sure it's easy for anyone who knows how electricity and speakers work.

But the differing sensitvities HAVE been taken into account. That's why when you look at your speaker trim / channel level settings, one speaker may need a +8dB boost, one speaker may only need a +1dB boost, one speaker may have a -3dB cut, etc.

Yes, at the calibration volume those are the appropriate corrections. But if the sensitivities are different, how do those corrections still apply perfectly? That's what I'm asking. Is this totally linear correction really working no matter if it's distance or sensitivity that's calling for those corrections to be made at reference volume? Because if it's not totally linear, then there's an increasing disparity the further you get from reference volume.

Yes; sensitivity is a given number, as is distance. Neither can be changed by changes in sound level.

I think the mist is clearing a bit. I apologize if I was sounding a bit flustered, because I was. It's frustrating when I can't wrap my head around something that should be simple. Thanks, guys.

Those corrections still apply because the sensitivites do not change, regardless of how different they are.



Yes, if prior to calibration your L speaker is 5dB louder than your R speaker (regardless of sensitivity or distance) then L will be 5dB louder than R at any and all volume settings. Same goes for speakers that both at the same level. Once they are corrected to the same level, they will remain on the same level at any and all volume settings.

Thanks. Whew!

See?


Magic.

Damn you, David Blaine!

I know! My receiver was encased in ice when I first got it.

Must've had and ICE amp.

Hope you had an Onkyo to defrost it with!

But CV....

What if the speakers were moving at near the speed of light, trailing an infinity long length of speaker cable, thus according to Einstein and proven by the clock discrepancies in super fast flying GPS satellites around the earth, time would begin to slow down for the speakers. Then what would we hear if we were riding on the speakers?

I'm just saying.............

We would all hear that damn tree falling in that damned forest that everybody always asks about!

I know you get it now, CV, but I'm going to try to explain how my brain grasps it anyway. Sensitivity is a static quality of a speaker, just like its dimensions, number of drivers, ports, etc. Sensitivity is simply how loud a speaker plays a tone when supplied with 1 Watt of power, measured at 1 meter from its front baffle. A less sensitive speaker can be made to play as loud as a more sensitive speaker by supplying a little more power. This "little more power" is assigned to speakers by the channel trim levels. When X watts (at any given volume level) are sent to the more sensitive speaker, X + trim watts are sent to the less sensitive speaker.

Well, there goes my magic theory. Curse you, science!!

*shakes fist at sky*

Or at least "Curse you, sciency-sounding stuff pulled out of thin air!"

Not that I don't think Peter's right, but...

Mark wins that bout. And as always, I totally heard it in Stewie's voice which made it even better.