Advent modification

[Pete B]

The Advent woofer has its own problems. There are better modern designs available to work with.

Name a modern woofer that will provide an Fc in the low 40s with a Qtc of about .8 in that size box, without serious response issues, and that will integrate as well with the tweeter with a simple single inductor for a crossover. The Advent woofer is available on ebay for about $20 so lets set an upper limit of $50, or I'll even give you $75.

It also should have comparable or better linear displacement, and particularly Xmax overload margin.

 

[Pete B]

Pete, you listed R2 as 4.3K. I cannot find a 4.3K resistor at Radio Shack. Can a 4.7K be substituted? Thanks.
http://www.classicspeakerpages.net/IP.Board/index.php?showtopic=2692&st=60&p=72883&#entry72883

4.7K should be fine, you can put a 47K in parallel with the 4.7 to get 4.3K if you are a perfectionist.

 

[Pete B]

Pete: Post the transfer function of your BSC circuit here, please.... :yes:

I simulated it in CALSOD, SPICE, and measured it with LAUD but the data is on my last system that I've not power up in 6 months or so. The circuit is very simple would take minutes to simulate. I'm not interested in a curve debate. I certainly look at speaker data, been doing it for over 25 years but I also learned that the room interface and psychoacoustic concerns are highly complex and I therefore put much weight on the in room listening experience.

Zilch, are you aware that some professionals discount boundary reinforcement due to wall flex? Geddes, Allison, and others come to mind. Obviously, it depends on the walls, concrete does not flex much. There is also often notching in the mid bass due to the floor and other boundary reflections; the in room response is complex.

Pete B.

 

[Pete B]

I'm off for lunch, and my own Advent collection is not completely finished, off to do some real work. I do not enjoy discussion based on how people think something sounds, so I'll bow out for a bit here. I hope some people try it and report back.

 

[Pete B]

Let me point out one more thing for now. T&S showed that a closed box speaker closely follows a 2nd order high pass transfer function. The peaking, or lack of it, is determined by the Q or Qtc of the system according to the well known classical theory for 2nd order systems. A maximally flat B2 (second order Butterworth) alignment requires a Qtc of .707, the Advent Qtc is .8-.9 which should be nearly flat or have less than about 1 dB of peaking. Your nearfield curve shows about 10 dB of peaking at 70 Hz as compared to 1KHz. Most closed box speakers follow this theory very closely, say within about 1 dB or so, small signal in the piston range.

How do you explain the large peak at 70 Hz?

Note that the Stereophile measurements also often show this peaking, whereas comparable measurements by Don Keele in older Audio reviews did not. Don Keele is the author of the paper that originated the nearfield measurement technique that is so widely used today. He received an AES award for that work.

Pete B.

 

[Pete B]

I've been running JBL 4430s as my mains for 20 years here, so it's two-ways with constant-directivity horns/waveguides and HF compensation.

I've not heard 4430's in probably more than 10 years, very nice for a high output system, overkill for the home environment - IMO. They look to be -3dB re woofer passband at about 40 Hz with the expected 4th order rolloff of a vented system. I'd want a sub woofer with them.

I seem to recall some ripple around the crossover region, +/-3dB per the published spec - good for an older system.

 

[Zilch]

Zilch, I suggest that you try BSC before trying to shoot it down with a long intellectual discourse. To be honest, I trust that old published curve much more than yours or Stereophile's measurements.

The CBS Labs anechoic curves show the baffle step quite clearly. I am merely questioning the actual magnitude of the 1.2 kHz peak. It seems to me that their hemispheric curve is more representative of actual. Some of it is present in my curves; I earlier suggested approaches to correcting it. Atkinson's curves show less of it than mine, even, I suspect because his are averaged over +/- 30° laterally.

Properly placed Advents do not have boomy bass, but will provide that kick in the chest feeling.

Well, that's the issue here, isn't it, after all? Take Advents away from boundaries, use them up on stands in free space, and you generate issues not present when used as intended, primariy among these being that the Advent bass is gone. How do we fix that? Stack them for gain from mutual coupling of the additional woofers and create another array of problems, or incorporate BSC and push their efficiency even lower?

What is "proper" placement for Advents? What you suggest is that they are better used with EQ in a different alignment than the original design intent. I have no problem with any of that; it's how I'm using them, in fact, for measurement purposes, without the EQ. What matters is that others understand and appreciate your "correction" as a system configuration PACKAGE, not merely $5 worth of Radio Shack components stuck in an insert loop.

Let's talk psychoacoustic effects, the threshold of audibility for response differences is about .1 dB if it is over a wide enough bandwidth and under the right A/B conditions. 1 dB is usually easily heard. 3 or 6 dB over the bandwidth shown in a region where our hearing is most sensitive, yes is screaming IMO.

We agree that some correction is in order. The 10 dB suggested by others clearly overstates the issue. Your own prescription acknowledges that the optimum correction varies on a case-by-case basis.

Do you listen at realistic levels? If not the equal-level loudness contours should be applied, and that peak is the opposite of what is required.

Not without hearing protection, no. In a larger sense, BSC IS loudness compensation. It's EQ, operatively similar to pushing a "Loudness" button for a different reason, and perhaps a different contour.

Are you familiar with the BBC dip, this is another factor concerning the difference between measured response and in room, phsychoacoustic considerations.

The secondary dip which remains AFTER BSC correction? There's two of them, actually.

BSC may be correcting for more than just BSC, what I can say is that it is a very significant improvement.

Which is why I have suggested parametric EQ to do the deed for whatever specific configuration and room conditions might exist. Again, what you're saying is not that BSC itself is an improvement, rather, that moving the speakers away from boundaries and application of the appropriate EQ is.

I simulated it in CALSOD, SPICE, and measured it with LAUD but the data is on my last system that I've not power up in 6 months or so. The circuit is very simple would take minutes to simulate. I'm not interested in a curve debate. I certainly look at speaker data, been doing it for over 25 years but I also learned that the room interface and psychoacoustic concerns are highly complex and I therefore put much weight on the in room listening experience.

I'm not interested in a curve debate, either. What I'd like to see is the curve, such that it might be used as a starting point by others for parametric EQ. This subjectivist "magic bullet" stuff is not real useful. There's no magic here. Post what you have found to be the optimum compensation curve, so that others may benefit from your work with this.

Zilch, are you aware that some professionals discount boundary reinforcement due to wall flex? Geddes, Allison, and others come to mind. Obviously, it depends on the walls, concrete does not flex much. There is also often notching in the mid bass due to the floor and other boundary reflections; the in room response is complex.

Yes. That's why I'm advocating in-room EQ in those ranges of the response where it can be effective. The whole "Audiophile" placement paradigm derives from experts declaring walls and floors to be the enemy of quality listening.

Let me point out one more thing for now. T&S showed that a closed box speaker closely follows a 2nd order high pass transfer function. The peaking, or lack of it, is determined by the Q or Qtc of the system according to the well known classical theory for 2nd order systems. A maximally flat B2 (second order Butterworth) alignment requires a Qtc of .707, the Advent Qtc is .8-.9 which should be nearly flat or have less than about 1 dB of peaking. Your nearfield curve shows about 10 dB of peaking at 70 Hz as compared to 1KHz. Most closed box speakers follow this theory very closely, say within about 1 dB or so, small signal in the piston range. How do you explain the large peak at 70 Hz?

Nearfield measurements overstate the LF response. They represent how the speaker behaves in 2-pi placement, which is how most owners use Advents, and how they were designed to be used.

Note that the Stereophile measurements also often show this peaking, whereas comparable measurements by Don Keele in older Audio reviews did not. Don Keele is the author of the paper that originated the nearfield measurement technique that is so widely used today. He received an AES award for that work.

Name-dropping is not helpful here. Let's see Keele's review of Advents, and his measured response.

I've not heard 4430's in probably more than 10 years, very nice for a high output system, overkill for the home environment - IMO. They look to be -3dB re woofer passband at about 40 Hz with the expected 4th order rolloff of a vented system. I'd want a sub woofer with them.

I seem to recall some ripple around the crossover region, +/-3dB per the published spec - good for an older system.

http://www.lansingheritage.org/html/jbl/specs/pro-speakers/1981-4430-35.htm

http://www.lansingheritage.org/html/jbl/specs/pro-speakers/1984-4430-35.htm

Keele designed the horn, and co-authored the AES paper:

http://audioheritage.org/vbulletin/showthread.php?t=7852

Behind the scenes of 4430 development as recalled by David Smith, including a sidebar discussing the "ripple:"

http://www.audioheritage.org/html/profiles/jbl/4430-35.htm

 

[archie2]

4.7K should be fine, you can put a 47K in parallel with the 4.7 to get 4.3K if you are a perfectionist.

Thank you. It need not be perfect, just an enjoyable perceived difference.

 

[Pete B]

Nearfield measurements overstate the LF response. They represent how the speaker behaves in 2-pi placement, which is how most owners use Advents, and how they were designed to be used.

Name-dropping is not helpful here. Let's see Keele's review of Advents, and his measured response.

You are correct that nearfield measurements represent the driver performance into a 2-pi position, however, T&S theory also represents 2-pi performance. So how do you explain the significant peaking, when theory suggests only 1 dB or so? Yes, the CBS measurements suggest no BSC compensation and agree fairly well with the theory, however your nearfield measurment suggests that there is BSC (actually overcompensated) .... something is not right with your measurements, IMO. The CBS measurements look reasonable.

Here is the AR-303 as measured by Stereophile:
http://stereophile.com/images/archivesart/303fig2.jpg

Ken Kantor's comment about this, suggesting that something is not right:
"To the extent that I am able to understand (calculate, model, measure) their behavior, the 303's have a Qtc of 0.83. I am including the source impedance of the xover in this figure. So we can expect a peaking of, what, about 0.3 or 0.4 dB dB? Certainly more than I would use for an NHT product, but likely to be swamped by other factors at 40 Hz.

I writing this from CES. If I run into JA, I will ask him if he remembers his measurement conditions. "

Other nearfield measurment issues:
http://www.classicspeakerpages.net/IP.Board/index.php?s=&showtopic=1487&view=findpost&p=58339

http://www.classicspeakerpages.net/IP.Board/index.php?s=&showtopic=1487&view=findpost&p=58338

Name dropping, oh come on, he authored the paper on a technique that you and Stereophile are using. Stereophile's nearfield measurements fairly consistently show a peak similar to yours that do not show up in Keele's or in agreement with loudspeaker theory. I consider this a sanity check, which good scientists do, and your nearfield measurements do not pass. I appreciate the fact that you share your work, but if you are going to use it to make a point I am certainly going to comment about the validity.

Pete B.

 

[Pete B]

The CBS Labs anechoic curves show the baffle step quite clearly. I am merely questioning the actual magnitude of the 1.2 kHz peak. It seems to me that their hemispheric curve is more representative of actual. Some of it is present in my curves; I earlier suggested approaches to correcting it. Atkinson's curves show less of it than mine, even, I suspect because his are averaged over +/- 30° laterally.

Well, that's the issue here, isn't it, after all? Take Advents away from boundaries, use them up on stands in free space, and you generate issues not present when used as intended, primariy among these being that the Advent bass is gone. How do we fix that? Stack them for gain from mutual coupling of the additional woofers and create another array of problems, or incorporate BSC and push their efficiency even lower?

What is "proper" placement for Advents? What you suggest is that they are better used with EQ in a different alignment than the original design intent. I have no problem with any of that; it's how I'm using them, in fact, for measurement purposes, without the EQ. What matters is that others understand and appreciate your "correction" as a system configuration PACKAGE, not merely $5 worth of Radio Shack components stuck in an insert loop.

We agree that some correction is in order. The 10 dB suggested by others clearly overstates the issue. Your own prescription acknowledges that the optimum correction varies on a case-by-case basis.

Not without hearing protection, no. In a larger sense, BSC IS loudness compensation. It's EQ, operatively similar to pushing a "Loudness" button for a different reason, and perhaps a different contour.

The secondary dip which remains AFTER BSC correction? There's two of them, actually.

Which is why I have suggested parametric EQ to do the deed for whatever specific configuration and room conditions might exist. Again, what you're saying is not that BSC itself is an improvement, rather, that moving the speakers away from boundaries and application of the appropriate EQ is.

I'm not interested in a curve debate, either. What I'd like to see is the curve, such that it might be used as a starting point by others for parametric EQ. This subjectivist "magic bullet" stuff is not real useful. There's no magic here. Post what you have found to be the optimum compensation curve, so that others may benefit from your work with this.

Yes. That's why I'm advocating in-room EQ in those ranges of the response where it can be effective. The whole "Audiophile" placement paradigm derives from experts declaring walls and floors to be the enemy of quality listening.

Nearfield measurements overstate the LF response. They represent how the speaker behaves in 2-pi placement, which is how most owners use Advents, and how they were designed to be used.

Name-dropping is not helpful here. Let's see Keele's review of Advents, and his measured response.



http://www.lansingheritage.org/html/jbl/specs/pro-speakers/1981-4430-35.htm

http://www.lansingheritage.org/html/jbl/specs/pro-speakers/1984-4430-35.htm

Keele designed the horn, and co-authored the AES paper:

http://audioheritage.org/vbulletin/showthread.php?t=7852

Behind the scenes of 4430 development as recalled by David Smith, including a sidebar discussing the "ripple:"

http://www.audioheritage.org/html/profiles/jbl/4430-35.htm

Look, I am not on trial here, I said this was a casual experiment, and I gave a table of values for 1 to 6 dB of compensation clearly recognizing that all rooms are not the same.

Why don't you just use your RTA and a cheap EQ to get a response correction that you like? Post it, then people will have more than one solution. I really have no need to win your debate, I see it as a waste of time because rooms are variable. You seem to be on a mission to prove something. I do notice that Advents need BSC even when against a wall, perhaps a bit less, but so what? You have framed your challenge for me to provide my ideal compensation curve when I have clearly stated that it will depend on the room and placement. I do not know the ideal curve for your or anyone elses room, obviously.

What I can say is that Advents most definately need some correction in every position I've ever used them in, even against a wall, typically 3 dB or more. They do not sound realistic or balanced without BSC, in any typical room position, this is my opinion take it or leave it. I'm not charging my consulting rate here after all, LOL!

Pete B.

 

[Pete B]

BACK TO the real subject. Can the BSC be put between the pre amp and power amp ?

Happy Forth of July All !!!!

What are your pre and power amp models? Yes it can go there, but the design values depend on the impedance levels of the equipment.

 

[Pete B]

Nearfield measurements overstate the LF response. They represent how the speaker behaves in 2-pi placement, which is how most owners use Advents, and how they were designed to be used.

So your saying that most users cut a hole in their wall so that they can flush mount and make sure that their wall is as rigid as say concrete? How do they minimize midbass issues due to reflections from other boundaries so as to provide an exact 2-pi environment?

Really, they were designed for 2-pi? That being the wall behind them, how do you think the designers were able to ignore all the other boundaries of the room per your claim that they assumed 2-pi?

I'm tired, I've got real work to do ....

Pete B.

 

[Zilch]

This is not rocket science - you put the mic 1/4" from the center of the dome, and measure using MLS, Sinusoidal, FFT, or RTA. Post nearfield measurements of the Advent woofer at variance with the results Atkinson and I show, and there'll be an issue to resolve. The subject is well discussed in Dickason -- nearfield overstates the VLF response. I get different results with fiberglass damping than the stock foam.

Just as you didn't commit more than "passing interest" to your BSC mod, neither am I about to make studying Advents my life's work here. I confirm Stereophile's findings, is what, and there are not many professional researchers who have measured more and varied loudspeakers than Atkinson.

YOU do it Pete, and then reconcile your results with theory and your own listening experience via whatever means you deem appropriate. Post them here along with the transfer function of your BSC circuit.... :yes:

 

[Pete B]

Funny, when I started the effort to restore a pair of OLA's I promised myself not to spend a lot of time on the effort. I did want to quickly figure out what was wrong with the midrange, and I was highly familiar with the design. Then I found out that there were several revisions of the OLA, and that the NLA was significantly different. I'm an all or nothing type of person, so I now have a pair of OLAs, 2 pairs of NLAs, a pair of 5002s, and a pair of Advent/1s. And I don't really like the speakers all that much, LOL!
Actually, it was probably the fact that I do like them with BSC that made me continue, that and curiosity.

My commercial reference is advertised as being +/- 1 dB from 36 Hz to 20 kHz as measured in the Canadian NRC lab. I'd expect this to include some smoothing and one should consider that it is their marketing literature. The Advents are not even close without BSC in A/B tests.

Here is a picture of my best personal design, obviously prototypes, this is a serious design with 2 - 12" woofers per side tuned to 20 Hz:
http://members.aol.com/basconsultants/spkrs.jpg

 

[Pete B]

This is not rocket science - you put the mic 1/4" from the center of the dome, and measure using MLS, Sinusoidal, FFT, or RTA. Post nearfield measurements of the Advent woofer at variance with the results Atkinson and I show, and there'll be an issue to resolve. The subject is well discussed in Dickason -- nearfield overstates the VLF response. I get different results with fiberglass damping than the stock foam.

Just as you didn't commit more than "passing interest" to your BSC mod, neither am I about to make studying Advents my life's work here. I confirm Stereophile's findings, is what, and there are not many professional researchers who have measured more and varied loudspeakers than Atkinson.

YOU do it Pete, and then reconcile your results with theory and your own listening experience via whatever means you deem appropriate. Post them here along with the transfer function of your BSC circuit.... :yes:

I don't need to do it, I trust what T&S predicts, Keele, CBS labs, but not you or Stereophile. You're trying to use a logical argument that because they do quantity they are getting it right? LOL! I showed several examples where they seem to be consistently wrong. Dickason is far from the final word on nearfield measurements, try the original article in the AES, or Joseph D'Appolito "Testing Loudspeakers". You have ignored my point that T&S represents 2-pi, and nearfield measurements properly done also represent 2-pi. Your measurements do not agree. I will drop out now, no point in continuing and I don't answer demands for work that I offer for free.

I have to wonder if your mic, or preamp is overloading with the increased mic volume displacement at LF. Consider the high SPL being so close to the cone.
What mic are you using?
Have you tried comparing ungated Sinusoidal to MLS, just as a sanity check?

Pete B.

 

[Zilch]

Really, they were designed for 2-pi? That being the wall behind them, how do you think the designers were able to ignore all the other boundaries of the room per your claim that they assumed 2-pi?

I am confident an AK Adventophile will soon link us to the official factory recommended placement conditions for Advents. Here's what mine say right on them:

The Advent Loudspeaker may be placed vertically or horizontally in any convenient locaton in a room. A convenient and favorable arrangement is simply to stand the speaker vertically at or near the base of a wall, with the low-frequency speaker nearer the floor. The system is small enough, however, to fit a number of shelf arrangements, and horizontal placement is perfectly acceptable.

If you have the inclination, you can experiment with various placements with the knowledge that room acoustics can often produce large apparent differences in the sound for small changes in placement. A simple rule of thumb is that the proportion of bass increases as you move a speaker toward intersecting room surfaces (wall and wall, wall and ceiling, wall and floor), and the greatest bass results from placement at the intersection of three surfaces (in a corner at the floor or ceiling). With that rule in mind, and with the knowledge that room acoustics often produce large apparent differences from small changes in placement, you should be able to obtain the balance you prefer with reasonable ease.

That says pi or 2-pi, as I read it, similar to other acoustic suspension systems of the era, and that's how most still use them. The suggested "convenient and favorable" placement would provide more enhanced bass response than even the nearfield measurements indicate, and I believe most users have long since determined that 1-pi alignment to be excessively "boomy" by contemporary taste and preference. You're telling us that 4-pi requires EQ or other measures to restore the design low-frequency performance.

 

[Zilch]

I'm an all or nothing type of person, so I now have a pair of OLAs, 2 pairs of NLAs, a pair of 5002s, and a pair of Advent/1s.

COOL! You have plenty of DUT measurement samples on which to report your findings here.... :thmbsp:

 

[Zilch]

I have to wonder if your mic, or preamp is overloading with the increased mic volume displacement at LF. Consider the high SPL being so close to the cone.

What mic are you using?

CLIO, as D'Appolito does op. cit. 4.4.2.

Have you tried comparing ungated Sinusoidal to MLS, just as a sanity check?

I will do that. I suppose there might just be something "special" about how Advents measure....

 

[Pete B]

COOL! You have plenty of DUT measurement samples on which to report your findings here.... :thmbsp:

I'm planning to do a more complete web page on the LAs and I might do some SPL measurements for it, however this will be as time permits.

 

[Pete B]

CLIO, as D'Appolito does op. cit. 4.4.2.

I will do that. I suppose there might just be something "special" about how Advents measure....

I'm not familiar with the CLIO mic, do you know if it is a Panasonic capsule?
I believe that JA uses Mightly Mic which is a Panasonic. It does have some distortion issues at high SPL.