Jump to content

Recommended Posts

On August 28, 2016 at 5:55 PM, Chief bonehead said:

to me, performance AND reproduction have to be one and the same......

Understood, but for me as a consumer there is going to be a price/performance and complexity/versus enjoyment threshold beyond which it makes no sense to live.  I have reached the level at which I enjoy immensely my music and don't spend too much time thinking about performance.  It's a luxury we consumers have over professionals.  I still wonder about getting some steep slopes crossovers in my khorns but I can live with and enjoy my current system until that itch someday becomes too much not to scratch. Warm regards, Tony

Link to comment
Share on other sites

On 8/25/2016 at 9:23 AM, Chris A said:

So this thread will illustrate the phenomenon of re-EQing to achieve certain loudspeaker sound signatures and give tips for those that are willing to move beyond typical "audiophile" organismic thinking related to passive crossovers and the exact materials used therein.

There are more topics to cover based on the subject identified above in the first post. The includes the sensitivity of the overall sound to:

 

1) flatness of frequency response required for increased overall realism

 

2) audibility of group delay and phase shifts due to bass bin ports and crossover filter issues

 

3) minimizing 2-inch HF compression driver "chatter" audibility

 

4) dealing with collapsing polar midrange horn spillover below its vertical polar control frequency in-room

 

Chris

Link to comment
Share on other sites

1) Flatness of frequency response for increased realism:

 

When I was dialing in my 5.2 array by hand using REW/calibration mike, I noticed that after taming those large humps in response of the Jubilee bass bin (anechoic response shown in the red trace below), that when I turned on good recordings, that there seemed to be a "magic" to the listening experience thereafter.  This was really unexpected since I'm aware that there are a lot of Khorn and Jubilee setups that don't use any form of EQ at all to tame these built-in frequency response issues, especially in the 60-200 Hz region.

 

Jub-KhornGraph.jpg

 

Once these humps in SPL response were tamed using parametric equalization filters (PEQs) in the digital crossover, the overall sound and ambience in the room seemed to calm down and reflect a much more realistic presentation. Below is a frequency response measurement using REW of a Jubilee bass bin in-room (in red trace) using the settings provided online:

 

Initial and final Jub bass bin performance.png

 

Once the bass response was smoothed out (green trace), the overall presence and sound in-room was markedly improved, with a natural or realistic sound that had never been present. 

Link to comment
Share on other sites

2) Audibility of group delay and phase shifts due to bass bin ports and crossover filter issues:

 

Some discussions centering on phase are based on notions of "absolute phase".  This section is not discussing "absolute phase", but rather "relative phase".

 

Note that impulse response factors into this discussion, since loudspeakers in their listening rooms will present a certain crispness and tightness of presentation if the relative phase and group delay across the audible spectrum is corrected to minimize them.  It's been my experience that any group delay growth in excess of ~5 milliseconds (ms) is audible, even below 500 Hz, and especially for bass reflex (ported) cabinets. 

 

In addition, phase mismatches in the crossover frequency bands, especially those phase shifts in excess of 360 degrees (i.e., time delay mismatches), show up as impulse response issues and indicate an issue with loudspeaker performance when percussion or other impulsive music event occurs, thus "smearing" the musical transient out.  This is especially audible on crash and ride cymbals on any recording, producing instead a very unnatural sound that significantly detracts from the realism of the performance. 

 

Here is the first of two phase plots for a 1979 Cornwall, taken at one metre:

 

LF Cornwall phase 1 metre.png

 

Note that I can't zoom out far enough to show you the total phase, so I have to show you in two plots.  Here is the phase plot for the midrange-tweeter frequencies:

 

HF Cornwall phase 1 metre.png

 

Note the scale used on the vertical axes.

 

In addition, increased group delay in the bass bin due to porting or other phase-inverting device (like a passive radiator) to convert the direct-radiating driver's backwave into an in-phase pulse--makes the bass "muddy" or "wooly" sounding.  Another way to say this is that any high pass filtering on the bass frequencies not only increases the phase/group delay, but also results in a degradation in the perceived bass tightness.

 

Here is a group delay plot for the above Cornwall.  Note that group delays outside of +/- 5 milliseconds are audible.  In the Cornwall, the woofer actually leads the midrange due to the length of the midrange horn.  Also notice the larger group delays of the bass bin below ~90 Hz due to the use of reflex ports:

 

Cornwall group delay 1 metre.png

 

Unfortunately, any electronic means to correct this via "FIR" (finite impulse response) digital filtering also requires an excessive number of "taps" (a very long digital filter length) to correct phase/group delay frequencies below ~500 Hz.  This means that the source of the issue (i.e., porting or use of direct radiators with passive radiators) increases the issues with bass performance in ways that can't easily be handled. 

 

In my experience, it's better to avoid the use of direct radiating woofers with backwave-inverting features (porting, passive radiators, etc.), and use an acoustic suspension or closed-box bass cabinet instead...or horn-loaded, and EQ in bass boost in the frequency band just below the resonance of the woofer in its enclosure. This approach minimizes phase and group delay shifts and actually can extend bass performance lower than the use of bass reflex design.  Care must be taken not to push this approach too far and use PEQs with very large boosting gains (generally, any overall gains above ~7). 

 

Correcting phase/time delay distortions in the loudspeaker's crossover regions is the main distinguishing feature of digital crossovers.  Once you hear these corrections in your living space using your loudspeakers, I find that no one that I've demonstrated this to wants to go back to using passive crossovers....the shift in the timbre of the loudspeaker toward a much more neutral sound is very apparent. 

 

However, it is noted that there were sound quality issues with first generation low-cost digital crossovers. such as the Behringer, dbx, and other crossovers employing less-than-spectacular analog circuitry, generally, many of the crossover units developed in the late 1990s and early 2000s.  Newer digital crossovers developed since 2010 that are quality units have apparently corrected thse issues.

Link to comment
Share on other sites

3) Minimizing 2-inch HF compression driver "chatter" audibility:

 

One of the design trade-offs using the newer "full range" 2" compression drivers is the cost of the moving diaphragm material itself.  In particular, use of materials like aluminum or titanium in these 2" compression drivers limits HF performance to below ~10 kHz, in m experience.  If beryllium  (Be) is used as the diaphragm material, and the compression driver design was optimized for the use of Be, this limit is easily pushed to 21 kHz or possibly higher.  The difference in the acoustical properties of Be over aluminum or titanium is about 7:1 in overall stiffness/weight--a very large advantage. 

 

Unfortunately, the cost of Be diaphragms has been high, although the costs associated with newer fabrication processes and the resulting material properties is apparently improving significantly in the past 5 years.  However, many "audiophiles" still refuse to invest in Be-diaphragm drivers because of perceived high costs (but they spend thousands on electronics, curiously, which have much less direct influence on sound quality) .  Until this situation changes more widely in commercially available compression drivers (high costs), the use of lower cost Ti or Al diaphragms will continue to be an issue.  The following suggestions for using these type drivers are the best alternatives identified and used by this writer:

 

a) Using a digital crossover, introduce a fairly sharp notch in the frequency response of TI and Al diaphragm driver around the typical 12-14 kHz "chatter frequency" to minimize the perception of chatter, using REW or some other measurement application to guide the placement and width of the notch filter, as well as correct any phase shifts of the higher-order notch filter(s).

 

Here is a color-intensity spectrogram plot of the K-69-A driver on a K-402 horn: note the chattering centered around 14 kHz:

New Center Wavelet Spectrogram K-69-A outside on-axis.png

 

b ) Use a tweeter crossed above ~8-10 kHz that has wide polar coverage that matches the polar coverage of the 2" compression driver/horn assembly at the crossover frequency, and carefully set the phase/time delays using a digital crossover as well as flatten the overall frequency response of the 2" driver/horn + tweeter to create a smooth overall response to 20 kHz or higher. 

 

Here is a color spectrogram plot of a K-69-A 2" (Ti diaphragm) compression driver on a K-510 horn crossed carefully to a Beyma CP25 "baby cheek" tweeter at 8 kHz, showing much better decay characteristics and absence of "ringing" than the above K-69-A driver alone:

 

Spectrogram K-510 + K-69-A crossed over to Beyma CP25 8 kHz.png

 

c) Save your pennies over time and buy higher quality Be-diaphragm 2" compression drivers. This is by far the preferred alternative in my experience.  Your ears will thank you.

 

Here is an equivalent color spectrogram plot of the TAD TD-4002 (Be diaphragm) driver on the same horn profile shape.  Note the same performance above 10 kHz as the K-69-A driver in the first plot, used alone:

 

Right Jubilee Wavelet TAD 4002 inside on-axis.png

Link to comment
Share on other sites

4) Dealing with collapsing polar midrange horn spillover below its vertical polar control frequency in-room:

 

This is a significant issue with the Heritage series of Klipsch loudspeakers, since all of them use the same type of collapsing polar midrange horns, whose vertical mouth dimension is much smaller than its horizontal dimension. This means that the midrange horn will begin to lose its polar control in the vertical axis at about 1700 Hz for this particular horn family.  This puts a lot of excess energy from 1700 Hz down to the bass bin crossover frequency on your ceiling and floor.  Obviously, the solution to this unbalancing acoustic energy reflected from ceiling and floor is to provide extra damping around the top and bottom of the horn mouth and on top of the loudspeaker cabinet, place extra absorption on the floor and ceiling (or diffusion panels that are good in the frequency range of 400-2000 Hz on the ceiling), and moving any near-field reflective objects away from the midrange mouths--within about 6 feet of the midrange horn mouth. 

 

Using PEQs to decrease the output of the midrange horn in this frequency band, while listening to favorite music selections, is another way to minimize the audible effects of the loss of midrange horn vertical pattern control.  The application of certain attenuating PEQs and shelf filters in the midrange should be appropriate for the room and its decor, and the personal tastes of the owner/listener.  Since this is what Roy calls "salt and pepper EQ", it should be done using a variety of music selections at mid-to-higher loudness levels while adjusting the individual crossover filters to taste.  Understand that this is a trade-off between direct and reflected SPL, so some time and patience should be employed in this process.  I've witnessed Roy do this using a K-510 horn versus a K-402 horn in the listening room adjacent to the Hope anechoic chamber

 

The ultimate solution is to use a midrange horn that doesn't lose its polar control down to its high pass crossover band (~400-800 Hz for the various Heritage midrange horns).  The K-402 comes to mind, but there are other midrange horn solutions that are available from third parties that have sufficient mouth dimensions and internal horn profiles that result in improved midrange performance and controlled directivity.  If using a midrange horn with an internal slot of 2" width, note that you should cross over to a tweeter at 4 kHz or lower to avoid a "frying bacon" sound of higher order harmonics (HOMs) present in these "constant directivity" horns.  If using a horn with a 1" slot width, you should cross over below 8 kHz.

 

Here is a plot of the horizontal off-axis frequency response of the K-402 horn, from on-axis (zero degrees) out to 90 degrees.  Note how it holds its polars quite well out to 50 degrees off-axis:

 

New Center Polars - Hor (10 Deg Incrs).png

 

Here is the equivalent plot as above, in the vertical axis of the horn.  Note that this horn holds its polar coverage in both axes down to its Fc of about 175 Hz:

 

New Center Polars Vert (10 Deg Incrs).png

Link to comment
Share on other sites

Note that the above uses of a digital active crossover, combined with measurement plots using REW and a calibration microphone, gives a visual representation and a much greater control of the performance of your loudspeakers to significantly improve the listening experience. 

 

Measurement brings much greater understanding and troubleshooting capabilities, and the control provided by a good quality digital crossover results in outstanding in-room performance that heretofore was not available in home two-channel or multichannel setups.

 

Chris

Link to comment
Share on other sites

Also note that the above REW plots were all generated from single upsweeps, and all views generated from those single saved upsweeps are generated from the same easily saved data file, i.e., one saved data file = many different performance views that can be generated. 

 

The data that I've presented in the above thread has been saved from that captured over the past 2 years, although I didn't spend even 40 hours over that length of time capturing that data--much less time than building a couple of loudspeaker cabinets, applying or repairing veneer, and trying to tweak passive crossover networks for purpose using only listening as a guide.

 

Each individual change in crossover settings has made a significant improvement in the resulting sound--much more than any change in in passive crossover parts, IME.  The payoff in investing that small amount of time in this activity has been much more than the effort involved.  In sum total, the difference in sound quality is difficult to quantify.  It's been my experience that you have to hear it to believe it.

 

Chris

Link to comment
Share on other sites

I've heard many comments about "it's too hard", or "it's too complicated" on this forum as justification for spending disproportionate amounts of time and money on "the old way of doing things".  I'd say that you shouldn't sell yourself short...you're more than capable to gaining much more performance and satisfaction for less time and money, even if that performance and satisfaction is a great deal less than what I show above. 

 

I often remember a quote from Heinlein:

 

Quote

 

A human being should be able to change a diaper, plan an invasion, butcher a hog, conn a ship, design a building, write a sonnet, balance accounts, build a wall, set a bone, comfort the dying, take orders, give orders, cooperate, act alone, solve equations, analyze a new problem, pitch manure, program a computer, cook a tasty meal, fight efficiently, die gallantly.

 

Specialization is for insects.

 

-Robert A. Heinlein

 

A little over the top...but you catch the drift.

 

Chris

  • Like 1
Link to comment
Share on other sites

Besides using impulse plots and the phase plots to find if your loudspeaker is time-aligned between drivers, you can also look at the "step response" plot within the impulse response plot within REW.  Sometimes it's much easier to see the delay mismatch using this view:

 

Right Jub step response.png

 

In the above step response plot, you can see a 0.22 ms (equating to 0.23 inches) in excess in delay between the bass bin and the K-402 horn.  While this isn't nearly as much as a 1/4 wavelength, it's visible within the step response view and can used be used to achieve better time alignment.

 

Chris

Link to comment
Share on other sites

Obviously, there is a steep learning curve with this approach, and most have no orientation or experience with it. Chastising those without the technical background to execute the approach is counterproductive. You assume that people understand the terms and concepts - a mistake. 

 

There has been no discussion of needed equipment or associated cost - something that would helpful (and necessary).

 

Granted, a lot of good information - but I wouldn't expect most Heritage users to pursue this anytime soon. As they say, "the devil that you know ..."

 

I appreciate the work you put into this thread - but I know there are a lot of questions surrounding setup and cost.

Link to comment
Share on other sites

20 minutes ago, Deang said:

Obviously, there is a steep learning curve with this approach, and most have no orientation or experience with it. Chastising those without the technical background to execute the approach is counterproductive. You assume that people understand the terms and concepts - a mistake. 

 

There has been no discussion of needed equipment or associated cost - something that would helpful (and necessary).

 

Granted, a lot of good information - but I wouldn't expect most Heritage users to pursue this anytime soon. As they say, "the devil that you know ..."

 

I appreciate the work you put into this thread - but I know there are a lot of questions surrounding setup and cost.

 

If you actually read the thread, you'd see that there is no "steep learning curve", and the equipment needed is a USB microphone and a computer.  That's it.  The USB microphone is ~$100US.  If you've got a good bus from your computer to your DAC or preamp (most here already do), then that's it.  Why you would take that position (i.e., "It's too hard") again seems as if you've got a axe to grind.  I'm fresh out of grindstones, I'm afraid.

 

The reason why I posted the above is simply that this approach's time has come.   When you look at the HT marketplace, many features of a digital crossover are already built-in to the AVPs and AVRs.  Some folks here use 2-channel HT setups already.  It's not difficult to go truly hi-fi now and it's now less expensive than "the old way of doing things" - by about 1/3 to 1/2, judging from the prices talked about.

 

Chris

Link to comment
Share on other sites

I didn't say it was "too hard", I said there was a steep learning curve. If you think everyone is going to catch on to this because of its "simplicity", you couldn't be more wrong.

 

"Axe to grind" - where did that come from? I thanked you for the work you put into the thread, and was only offering up some constructive criticism to help you make the thread better. 

Link to comment
Share on other sites

35 minutes ago, Chris A said:

When you look at the HT marketplace, many features of a digital crossover are already built-in to the AVPs and AVRs.  Some folks here use 2-channel HT setups already.  It's not difficult to go truly hi-fi now and it's now less expensive than "the old way of doing things" - by about 1/3 to 1/2, judging from the prices talked about.

 

Sorry, but 2-channel listeners don't use HT receivers, and those that do, aren't seeking the pinnacle of performance.

 

So, "some" are using HT receivers - so please show them what they need to do to make this work.

Link to comment
Share on other sites

Why would you--a supplier of third-party passive crossovers-- come on this thread to complain about "steep learning curve"?  The only learning curve I had was making sure that the microphone used in REW was the one that I set up 1 metre in front of the loudspeaker and not the one in the laptop.  That's not much of a learning curve.  Once I did that, each sweep that was run is still on my hard drive, and I can see all the information plotted above even years after I ran the sweep.  That's a pretty shallow learning curve. 

 

Additionally, each person using REW has the option of posting their entire sweep on line for others to help them with.  This has worked well in the past.

 

7 minutes ago, Deang said:

 

Sorry, but 2-channel listeners don't use HT receivers, and those that do, aren't seeking the pinnacle of performance.

 

So, "some" are using HT receivers - so please show them what they need to do to make this work.

 

This is getting a bit abusive.  Why don't you let "them" decide for themselves and ask their own questions?  I didn't see any questions from you.  Are you someone's agent here?

 

Your proclamation of HT receivers not being the "pinnacle of performance" could certainly be turned on the products you sell here.  I simply don't agree with your statement -- I believe that it's not reflective of reality at all.

  • Like 1
Link to comment
Share on other sites

Now I'm "abusive"? I think it's obvious that being considerate to you doesn't work - so I give up. 

 

You asked where the party went. I can tell you that no one is posting because they either don't understand, or have grown tired of your condescending attitude. 

 

Seems that everyone here knows what's going on with me and my business except you - and while I've promoted my work in the past, I've never solicited.

  • Like 1
Link to comment
Share on other sites

I'm good for specific questions on the above information.  If someone is lost reading the above post, I welcome any questions--even "I don't understand this" questions--but pointing to a specific post where they got lost.  I'm also good for step-by-step instructions, if someone is interested in doing any of the above for themselves.

 

I find that most people initially want to adjust their frequency response and perhaps double check their crossover performance.  The above posts are meant to be viewed as informational-only on what you can do--so that those that might be interested have a "glossy brochure" on those capabilities with some examples.  The posts above haven't been arranged in "step by step" fashion because of that introductory purpose. 

 

Step-by-step expansion can be done on any part of the above if someone is trying it out and needs that kind of support. I believe that I can do that part, too...no worries.  I've helped many people that way, and a lot of those were via PMs.

 

Chris

  • Like 1
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...