Center channel Impedance.

[Sir Terrence the Terrible]

I guess that I don't get it. It seems that we are talking about the same thing, yet somehow, it's different. I understand how the same signal reaching across the face causes a loss of imaging, but I don't quite understand what causes the notch.

We are talking the same thing, but at different points in the recording/mixing/playback chain. The cause of the notch is something called the interaural time delay, or the time it takes for the signal to wrap around the head, and get to the opposite ear. That delay causes the notch.

The only way I can understand this is to think that the left signal reaching the right ear is out of phase with the sound from the right speaker and when the two signals combine they create peaks and troughs, hence, distorting the sound. If I understand it correctly, what we hear is a softening (less distinct) of the sound .

Actually it is not out of phase, but in phase and delayed. You have a signal from the left speaker hitting the left ear, and a delayed signal(interaural time delay) left speaker signal hitting the right ear. That is what causes the notch, not a flip in phase.

Just thinking...

What's different that you do compared to Choueiri is that you ( I assume) equalize both channels at the same time whereas, Choueiri actually adds information to each opposite channel to cancel the cross talk.

So you understand what we are doing in the studio. When vocals are recorded, they are on a single mono track. When we mixed them into a stereo configuration, we split that mono track between two channels equally. This causes the vocals to lose about 3db of power in the mix, so we restore that power back by bringing up the vocals by that same amount. As a result of that split, vocals will take on a diffusive quality because of the HRT effect we discussed above. We equalize to get rid of that "fuzziness" on our side of the audio chain. When you play that same track on your speakers, the effect crops up again, but is somewhat lessened because of what we have done on the mixing side.

What Choueiri is doing is trying to eliminate the effect of HRT effects on the playback side, and his idea is really quite old. Bop Carver was doing this very same thing with his sonic holography circuit, and Polk was doing this with his SDA towers in the 80's.

It seems that Choueiri's method would better solve the problem except for one thing, his method is highly dependent on listening position since there is a timing/amplitude factor involved in when the signals reach the ears and moving away from that "sweet spot" ruins that timing/amplitude.

You method is a compromise that maintains a more even response regardless of the listening position.

Am I beginning to understand you?

Thanks

Actually we are doing the same thing, but on different sides of the recording and playback chain. We do it only to vocals and instruments mixed from mono sources into a stereo configuration. He is trying to correct the effect in a broadband way on the playback side.

[E-Stat]

Since my speakers are 91db efficient. This means it takes about 1/5 of a watt to drive my speakers to this volume (76db). Since my amp is about 6 watts, this makes my maximum volume about 97db at 4 watts.

Another difference with line sources like Maggies is their output drops in a linear fashion with a doubling of distance as opposed to a logarithmic drop of point sources (3 db vs. 6 db). Most efficiency specs are stated at a distance of 3 feet or 1 meter. At 6 feet, therefore, output would be 94 db, and at 12 feet, 91 db. A point source would have dropped to 85 db at that distance!

I leverage the 6 db gain from doubling Advents in the garage in a similar quality vs. quantity manner. I usually listen in the mid 80 db level where the amp is indicating -20 db output or 4 watts (yes, an '81 Stasis has peak reading output ladders). What that means is the class A voltage amp is providing all the output where the sound is marginally better with improved focus and a softer top than when running at higher levels where the AB current mirror is also contributing. The only time I've used the full output is at Halloween. For the past two years, my closest neighbors have thrown an outdoor party for their teenaged daughter. I've taken the system over to what used to be a tennis court. It will cleanly crank Rihanna in the mid 90s at some distance when the amp is wide open at 400 watts. Last year, doing so for three continuous hours destroyed the woofer surrounds on the older pair necessitating a re-foaming job.

Halloween

rw

[Feanor]

3 db from acoustic gain. 3 db from a doubling of power. If you had a transformer driven amplifier whose output was consistent, then you would get 3 db from merely doubling the speakers. Additionally, you are now running twice the power to double the speakers.
...

"If we have two sources of the same volume, the loudness acoustic gain is 3 db."

Before the Sound Labs, I used Acoustat 2+2s which had four panels per speaker. There was also a 1+1 which used the exact same transformer for driving half the number of panels. While there were different taps to the bass transformer to optimize the frequency response, the 2+2 was twice as efficient having double the panel area. 400 watts into a double pair of Advents has 3db more output than 400 watts into a single pair.
...
rw

I know I don't understand "acoustic gain", but I'd like to. It almost seems like you're getting something for nothing in terms of energy. Do you have a reference that I could read & understand, (bearing in mind the my science/engineering background is weak).

[StevenSurprenant]

Actually we are doing the same thing, but on different sides of the recording and playback chain. We do it only to vocals and instruments mixed from mono sources into a stereo configuration. He is trying to correct the effect in a broadband way on the playback side.

Thanks for taking the time to help me through that. It's a simple concept, but when you're as thick headed as I am, sometimes it takes time to sink in.

[E-Stat]

Do you have a reference that I could read & understand, (bearing in mind the my science/engineering background is weak).

There are tons of them around. Here's one that graphically depicts the table which Smokey posted back on page 2. What actually improves is the overall sensitivity from the added cone or diaphragm area. Professional sound reinforcement makes broad use of this concept.

rw

[Feanor]

There are tons of them around. Here's one that graphically depicts the table which Smokey posted back on page 2. What actually improves is the overall sensitivity from the added cone or diaphragm area. Professional sound reinforcement makes broad use of this concept.

rw

Thanks for that.

[StevenSurprenant]

I'm humbled!

I was scrolling through the posts and It just occurred to me that everyone that replied to my initial posting pretty much represented the backbone of this site.

Sir Terrence the Terrible - 5,841 posts
E-Stat - 4,672 posts
bobsticks - 6,647 posts
Smokey - 2,724 posts
Feanor - 5.965 posts

And then there's me at - 170 posts

I'm almost embarrassed...

All I can say is thank "you all" for helping me and opening my mind to new ideas. The saying that I am fond of repeating,' The more I learn, the less I know", truly applies here. At times I got a little defensive and I apologize for that, but understand that I listened to everything you folks said and I discounted nothing. You guys are truly great people. Thanks.

[Smokey]

However, and I'm not sure about this, at 6 watts which is way below what my present speakers can handle, I don't think I could blow a speaker even if I ran my amp into full clipping for a prolonged period of time. I'm not going to test this!

With lower power amps, you probably would not damage midrange or woofer, but you might put tweeter at risk if prolong clipping occurs. The problem is a clipped waveform (squarewave) will have much more higher frequency energy than its sinewave equivalent due to squarwave having more [high frequecny] harmonics.

So a tweeter will dissipate more power and heat from clipped wave than a non clipped sinewave.

Would like to know what is THD for full power output. I bet it does not look pretty.

And you would be wrong. Again.

FIRST WATT F2J

I am wrong if you consider 1% THD an acceptable number for distortion. As you can see from the chart, F2 amp reaches 1% THD at 1 watt, and F2J amp reaches 1% distortion at full power output (5 Watts). 1% THD in audio world is totaly unacceptable

Beside I don't know why you are advocating low power amps since your self recently bought a 300 watts amp, and said that having more headroom delivered more weight and far more clarity to the presentation(post #5):

Just joined the Emo tribe

At least you are consistently wrong. I assure you that the measurable levels I get at 400 watts is much higher than for 4 watts. You can teach me nothing in this regard. I know exactly the difference between what the Stasis output is at 0 db and at -20 db.

Ofcourse there will be level differenec between 4watts and 400watts, or 4 watts and 40 watts for that matter. But most typical HT setups uses between 1 to 20 watts (RMS) of their amp and rest of power are reserve for music peaks. That is why I said [extra] power is not for loudness.

[E-Stat]

I am wrong if you consider 1% THD an acceptable number for distortion...1% THD in audio world is totaly unacceptable

I refer you to the last time we (and Pix) had this discussion exactly a year ago. Neither he nor you backed your assertions with demonstrable proof via the online test which I had taken. My position was stated in posts 28,32,38,64, and 81 with Pix. You entered the discussion and my responses at 90,93 and 100 cover those.

Here's the thread. It was worth the review.

Beside I don't know why you are advocating low power amps since your self recently bought a 300 watts amp, and said that having more headroom delivered more weight and far more clarity to the presentation(post #5):

Sorry, but I guess I don't respond very well to those who make unqualified grandiose declarations which are not supported by fact. As for me, I've never owned a low power amp because my choice of speakers has always dictated differently. You would get consistent disagreement among horn speaker enthusiasts who are very similar with the sort. BTW, the XPA-5's output is 200 watts into 8 ohms.

rw

[StevenSurprenant]

Beside I don't know why you are advocating low power amps...

Of course there will be level differenec between 4watts and 400watts, or 4 watts and 40 watts for that matter. But most typical HT setups uses between 1 to 20 watts (RMS) of their amp and rest of power are reserve for music peaks. That is why I said [extra] power is not for loudness.

You made your point clear and I think everyone agrees, but I've had people tell me that 1,000 watts isn't enough power for those peaks. So how much is enough?

There is a reason why many people use low power amps. In many cases, they just sound better. Advocates of low power amps claim that there is more information reaching their ears and they usually attribute this to fewer gain stages. That was Nelson Pass's selling point of his Aleph 3 amp (30 watts). I listened to his amps from that era and from what I heard, he might be right.

I've listened to a butt load of amps and speakers and for the most part, the larger amps seemed to lack the micro detail of the smaller amps. This isn't a set rule. Besides, many claim that the larger amps have more control over their drivers, i.e. tighter bass.

The other thing is that many people advocate tube amps, not so much for their warm sound, but because of they claim that they create a more 3 dimensional sound stage as compared to solid state. I don't know if I can agree. The most defined sound stages I have heard were created with ss amps, notably Pass Labs and Spectral gear. Even a Linn system I heard did a pretty good job of fleshing out a decent sound stage. I have also heard tube amps that did a pretty good job too.

More power might mean that you will have ultimately louder dynamics, but dynamics is only part of the picture. As long as we keep our peaks below the distortion point, it should sound good. That's another issue (distortion). Back in the 70's my amp was rated at .3% distortion at rated output. A number of amps came out that were using a great deal of feedback to their drive transistors resulting with distortion figures of .0001% or some sort of low number. Spec wise, these things looked great on paper, but they sounded anemic.

The other thing is people with low power amps realize the limits of their system. They would never play Tchaikovsky's 1812 Overture at realistic volumes. I doubt that there are many systems that could do that.

Anyway, you are correct that peak power is important, but as long as you limit the speaker volume to the level of available peak power, there isn't an issue. If I took my speakers outside, like E-STAT did with his Advents, and drove my amp to full power, I doubt that I could disturb my neighbors. However, inside my smallish room and sitting 7 feet away from them, they are very loud.

This is all way over my head. All I know is that 6 watts is enough to play my system as loud as I want it to go without detecting any distortion.

My surround system, which has 130 watts per channel, gets played just as loud and I have yet to detect any distortion or compression from either system.

Besides, if I needed more volume I could bi-amp the system using a line level crossover (I've done this - same power - twice as many amps) or I could use a sub and limit the amount of bass going to the main speakers, or use multiple drivers as we spoke about earlier... or all the above!

You're making my head spin... Stop it!

[Feanor]

I enjoyed your reflections here, Steven.

People have different needs and different preferences -- two dimensions that we need to keep in mind whenever we give or receive advice.

My system, see under my signature, is the most satisfying that I've owned and, fortuitously, also the cheapest in a long, long time. However it conforms to my taste and needs, not necessarily to anyone else's.


My requirements and tastes:

• I enjoy the soundstage of the dipolar planar speakers
• I listen at only moderate volumes
• I don't require the "pounding" bass that some sorts of music demand
• I prefer maximum resolution and transparency (vs. warm, harmonic richness)
• I'm poor and can only affort a modest outlay.

[bobsticks]

I'm humbled!..stuff.

Steve, personally, I would turn that sentiment around on you. I suspect we're all just glad to have you as part of the community and I respect the fact that amidst this thread there have been a variety of conflictory opinions and theories expressed with a modicum of civil comportment.

A theory that I've always held is that "system synergy" and "the room" are just as, if not more important, than equipment. Yes, there are hard and fast rules to set ups and good kit is good kit and numbers are numbers, results being what they are...but take the best piece of gear and throw it in an unbalanced system and a rotten room and see what the results are.

Ultimately, I appreciate the way that you've raised some questions and expressed some opinions. That's how I started; that's how I learned. I suspect the others would say the same...

That's how we all learn.

[StevenSurprenant]

I know I don't understand "acoustic gain", but I'd like to.

It almost seems like you're getting - "something for nothing"- in terms of energy.

I think this is what Smokey has been alluding to all along... He has been telling us just that... You don't get something for nothing which makes perfect sense.

A perfect amp will double the current (and power) each time you halve the impedance. You can see this in the amps spec sheets. So, two 8 ohm speakers in parallel will consume twice as much power, but each speaker (separately) in this configuration will maintain the same acoustic output as if only one speaker were being used. Together, they will be 3db louder. That seems perfectly logical. All this happens without touching the volume knob.

Referring to the pages supplied by Smokey and E-Stat which show the acoustic power of multiple drivers... I think it is a little misleading and can be interpreted incorrectly (at least I did, at first).

This is important...The one point they don't seem to make is that everything is in reference to each (single) speaker outputting the same db in every case, irrespective to the amount of input power for the total system. In the real world, when we start adding more speakers, we start configuring them in series and parallel circuits to maintain an impedance that the amp can deal with.

For two speakers... If we wire them in parallel, the impedance drops, the current increase and the acoustic output doubles. Nice! If we wire them in series, the impedance doubles, the current to each speaker gets cut in half (is divided between both speakers), and the acoustic output goes down 3db for each speaker, but together, they have the acoustic output of a single speaker at the initial power, there is no net gain.

Someone correct me if I'm wrong.

For four speakers and in any situation where we double the number of speakers beyond that, we run into a different situation which seems to make the charts we are referencing above seem incorrect. If we have two speakers and we double them to four, assuming they are all 8 ohm, we need to wire then into two groups of series circuits (16 ohm) and then wire those groups into parallel with each other, which brings the total impedance that the amp sees to 8 ohms. I'm sure everyone understands.

This is where it gets sticky...

Without changing the input power, (combined) the 2 speakers that we wired in series only have the same acoustic power a single speaker driven at that power. Makes sense... Double the impedance -> half the current. Adding the second set of series wired speakers in parallel (without changing the input power) with the first pair, will bring the system impedance down to 8 ohms and the current/power to them will double. Combined they will have the same acoustic output as only 2 speakers in parallel..


In a nutshell, without changing the initial power from the amp, wiring two speakers in series will double the impedance and the total power will be divided between both speakers. The acoustic output of each speaker will decrease by 3db, but together they will gain that 3db back and play the same loudness as one speaker driven at the same power. Adding another set of these series wired speakers in parallel will halve the impedance and each set will play just as loudly as they did separately because the amps current will double. Together, they will play 3db louder than either group of series wired speakers. Now that should be clear as mud.

Let me repeat the most important part of this rumination. The charts that we were linked to are valid if every speaker is outputting the same acoustic power. This can only be accomplished with multiple amps, or if we wire the speakers in parallel and the amp can deliver the current needed at the lower impedances. Once we start wiring the speakers in series, the chart becomes invalid. I'm sure I don't need to explain it to you guys.

That pretty much balances the equation. If we get an acoustic increase from multiple speakers, it's because we've increased the total power. This means that we don't get something for nothing.

It seems so simple now. I know I repeated myself several times, but sometimes it helps in understanding to word it differently.

There is one after thought...

If your amp doesn't double the power into 4 ohms and you have a speaker with a tweeter at 8 ohms and a woofer (or series of woofers) at 4 ohms, the speaker will sound different than had you used a perfect amp which doubles the current at 4 ohms. In effect, the woofer/s would be at a lower volume (less current) with the lesser amp and cause the speakers to sound brighter.

Just thinking....

[Feanor]

The article linked by E-Stat asserts the fact of "accoustic gain" and links it to the area of the transducer. I'll admit I didn't understand how this wasn't "getting something for nothing", but I read the item only once so maybe it just when by a little too quickly.

The article does also discuss the issues of multiple drivers which I guess are mainly "lobing", i.e.effect of different phase of sound waves arriving at the listener from drivers being at different distances.

[E-Stat]

The article linked by E-Stat asserts the fact of "accoustic gain" and links it to the area of the transducer. I'll admit I didn't understand how this wasn't "getting something for nothing", but I read the item only once so maybe it just when by a little too quickly.

The answer is mutual coupling. You may be aware of the retired engineer over at AA who is a big Advent proponent. I've had the good fortune of visiting him during a business trip. Here is his explanation. It is true that such is not a consistent 6 db gain across all frequencies.

rw

[StevenSurprenant]

The answer is mutual coupling. You may be aware of the retired engineer over at AA who is a big Advent proponent. I've had the good fortune of visiting him during a business trip. Here is his explanation. It is true that such is not a consistent 6 db gain across all frequencies.

rw

Okay, I guess that I'll have to rethink this. Ohhh my head hurts!

I'll put my idea of electrical power aside and just think of it in terms of acoustic power.

Since this effect occurs mostly at certain frequencies and that the frequencies that this occurs at is dependent on how far apart the woofers are placed from each other, I might assume that this gain is a result of the waveforms at those frequencies being in phase with each other, hence additive in nature.

Is this closer to the idea?

[E-Stat]

I'll put my idea of electrical power aside and just think of it in terms of acoustic power...Is this closer to the idea?

I don't profess to fully understand the theory - just to apply it!

rw

[StevenSurprenant]

I don't profess to fully understand the theory - just to apply it!

rw

I understand....

Actually I was thinking of some applications for this. For instance, I thought of rebuilding my speaker boxes using the extra mid bass drivers I have in a 2.5 configuration. Since my speakers only go down to about 55Hz at -3db, I thought about adding a second woofer and crossing it over at such a point that it would bring that -3db point up to 0db and extend the bass down a little further. Anyway, this subject complicates things a bit, but it creates the possibility to increase the bass a total of 6db and could, in theory, allow me an even lower output. I'm just thinking, that's all.

Thanks again.

[Feanor]

The answer is mutual coupling. You may be aware of the retired engineer over at AA who is a big Advent proponent. I've had the good fortune of visiting him during a business trip. Here is his explanation. It is true that such is not a consistent 6 db gain across all frequencies.

rw

BTW I've heard of mutual coupling in another context. Thanks for that link re. stacked Advents.

I have the Big Box Pro and X-Over Pro programs for speaker crossover and cabinet designs. There they allude to mutual couple. X-Over Pro asserts that the mutual coupling will only occur where identical drivers are located within 1/8 wave length of each other. I.e. coupling is frequency dependant.

At one time I made a chart of Frequency > 1/8 wavelength:

Frequency Hz	1/8 wave, inches
20	84.8
40	42.4
80	21.2
160	10.6
200	8.5
300	5.7
400	4.2
500	3.4
600	2.8
700	2.4
800	2.1
1000	1.7
1200	1.4
1350	1.3
1500	1.1
1800	0.9
2000	0.8
2200	0.8
2400	0.7
2700	0.6
3000	0.6
4000	0.4
5000	0.3
6000	0.3
7000	0.2

So according to this it's quite practical to couple a pair of 7" mid/bass drivers below about 250 Hz but not much above. But it looks fairly impractical to couple 1.25" tweeters in typical 4" mounts at any relevant frequency.

[StevenSurprenant]

BTW I've heard of mutual coupling in another context. Thanks for that link re. stacked Advents.

I have the Big Box Pro and X-Over Pro programs for speaker crossover and cabinet designs. There they allude to mutual couple. X-Over Pro asserts that the mutual coupling will only occur where identical drivers are located within 1/8 wave length of each other. I.e. coupling is frequency dependant.

At one time I made a chart of Frequency > 1/8 wavelength:

Frequency Hz	1/8 wave, inches
20	84.8
40	42.4
80	21.2
160	10.6
200	8.5
300	5.7
400	4.2
500	3.4
600	2.8
700	2.4
800	2.1
1000	1.7
1200	1.4
1350	1.3
1500	1.1
1800	0.9
2000	0.8
2200	0.8
2400	0.7
2700	0.6
3000	0.6
4000	0.4
5000	0.3
6000	0.3
7000	0.2

So according to this it's quite practical to couple a pair of 7" mid/bass drivers below about 250 Hz but not much above. But it looks fairly impractical to couple 1.25" tweeters in typical 4" mounts at any relevant frequency.

Those numbers look very good for what I was thinking of. I was thinking about crossing over close to 200hz, that's where the curve begins to drop. The 8.5 inches distance is doable. but I'd have to make the boxes very deep.

BTW, how do you like the software you are using?

Thanks.

[Feanor]

Those numbers look very good for what I was thinking of. I was thinking about crossing over close to 200hz, that's where the curve begins to drop. The 8.5 inches distance is doable. but I'd have to make the boxes very deep.

BTW, how do you like the software you are using?

Thanks.

I have Bass Box Pro and X-Over Pro from Harris Tech, see HERE. They are easy to use and have great user interface; also they come with and extensive database of driver specifications.

A couple of shortcomings are that they cannot take into account driver relative offsets, nor will they optimized values for a given topology, (although they will calculate the results of manually changed component values).