Loudspeaker Design - from the perspective of a loudspeaker design engineer...

[Chris A]

I've noticed that this forum is a bit unique among the various audio forums on the Web, and I feel that this viewpoint actually needs to be encouraged--especially in this sub-field of expertise (i.e., loudspeakers in real listening rooms).  Currently, I think this sub-forum is suffering a bit from a lack of understanding what its role actually is in the schema of audio subforums, and I for one would like it to be a little less threatening and fear-inducing for those unsure of themselves in this new realm.  But nevertheless, the intent is to stick to the charter that its creator has set, namely...

 

Quote

•  
• 
•  
• 26194 posts

• Popular Post

•  

Posted February 17

Welcome to the new space solely for objective audio discussions. Subjective impression threads, anecdotal or experience only based comments are not allowed here. Have some wonderful news to share about a USB cable? Great, do it in another sub-forum here on Audiophile Style. 

 

Other than the logical uses of this sub-forum, it will also be used to house objective-based comments that seek to refute someone's personal experience  posted elsewhere. For example, if someone says that two bit identical files sound different, the new Objective-Fi sub-forum is the place to begin / continue the discussion unabated and away from appeals to authority or unprovable psychoacoustic comments. 

 

As always, feedback and questions are encouraged. I'm sure we have some kinks to work out, but we'll get there. This is audio, not brain surgery. Nobody is saving babies or killing puppies. 

 

Assuming that the subject of this thread is in-bounds for this sub-forum, my intent is to introduce first-hand comments/observations/measurements from a notable master loudspeaker design engineer working for a large loudspeaker/audio engineering company--in a specific manner--as it has been seen and related by one of his disciples stretching back over the past 13 years or so.  His recent loudspeaker designs have won many rave reviews from others on far reaching forums and audio blogs (most of which are not in the "objective hi-fi realm--which this thread can begin to be remedied). 

 

What this thread will attempt to do is to bring a more engineering perspective to the discussions in this forum on the subject of loudspeaker design and capaiblities.  This forum in general seems to have a preponderance of computer/electrical/electronic engineers and perhaps practicing physicists. But I notice that precious few acoustics/audio transducer engineers are posting that can call themselves truly "loudspeaker engineers" here. 

 

Why loudspeakers...and not some variety of electronics or other digital signal processing systems in this discussion?  I recently rediscovered a comment buried in a closed web forum located elsewhere on the subject of loudspeaker design, and thought of this thread (which I have heretofore not posted into).  Here is a comment from that loudspeaker engineer, who is my mentor:

 

Quote

I think from what I have seen and experienced that changes/differences are, or least ought to be categorized by how much impact they bring. To me and maybe this is why I am intrigued by loudspeakers; they can sure make a whole lot of difference--what I call large percentagers [a.k.a, “big hitters”]. I mean, they are at the end of the audio chain before the energy is converted to acoustic energy to vibrate air and then reach our ears. I have been in situations before where I make a call on an amp, preamp, a CD, etc., then I change the speaker and I find myself having to do it all over again.

I think that once you get speakers as close to what we call perfect, such as it is in this imperfect world, then we reach a state of how talent is distributed within the system of components making up the setup, i.e., can I hear that cap? can I hear that wire? can I hear that different compression driver? tube vs. solid state? The better resolution speaker can act as a transformer/magnifying glass to help heighten the accuracy/precision of whatever talent we were blessed with…You can make the biggest change to your system you can make; and all other changes will pale in comparison to this large percentager. Pretty dramatic.

The loudspeaker engineer can't control your room by designing it for you... I speak only of the audio chain and those elements used to spit out acoustic energy--from CD/tape/LP/mental telepathy to moving air to your ears.

This is the comment that had me thinking that, finally, a non-subjective (read: "non-religious" and "informed by science, not belief systems") discussion based on repeatable measurements and physics on loudspeakers could be attempted, and that the resulting discussion would significantly educate those on the edge of the field to make at least better buying decisions, if not encourage them to expand their engineering skills into real loudspeaker design, test and evaluation.

 

So, if this thread isn't out of bounds (and I'm hoping that the admin agrees that this sort of thread is sorely needed on this forum), I will attempt to introduce some of the observations, measurements, and points of view/wisdom that have been shared by this prolific (and who will remain unnamed) loudspeaker development engineer. I've learned a great deal, which I've been able to turn into my own loudspeaker designs that might pass muster in any first-rate audio engineering company.  These are specific statements that are often clouded by individuals not versed in the actual engineering that are of upmost importance--the root of the technical knowledge that matters, and that can direct our attention to those specific loudspeaker designs having a "big hitter" or "high percentager" status.

 

First, and foremost, the preamble to what's being attempted with loudspeaker design:

 

Quote

What should the goal of a loudspeaker be? My answer: to recreate the event; to be a reproducer (not a producer) of the acoustic event.

 

This means that "pleasantness" and "euphonic sound" that's not accurate to the input program material isn't the design goal, but rather an accurate reproducer of the input signal (just like it is for all the electronics and other digital systems in the signal stream).  If it still sounds really good subjectively, then that design goal has been validated as a correct one.  So if that otherwise means to you that the loudspeaker and its corresponding electronic/digital signal chain up to and including the recording microphone or signal generator can reproduce a square wave, then so be it.  If that capability is instead traded against other performance capabilities (including cost as a design constraint), then that's the aggregate design goal.  This is objective performance, not subjective.  The interplay of various loudspeaker performance capabilities in each category against each other can be complex, but not unknown in their desirability to the finished design.

 

There is a great deal more that can be said, but the emphasis here is on something that's more specific and a little less generalized.  For instance, a loudspeaker measuring ±90 degrees of phase in-room from 100-20,000 Hz is well within the realm of attainability (without the use of FIR filtering), as well as other more familiar measurements of merit including flat SPL response, controlled directivity response (-6 dB) horizontally and vertically, and other distortion thresholds, such as modulation, harmonic, compression, impulse, and other forms of measurable distortion measured at 10% of input power handling capability (an AES standard).

 

The next portion of our preamble (or starting) discussion is aimed at those people just itching to put their mark on this thread (even before reading this)--these of the issues of the "audiophile needs regarding loudspeaker performance in-room":

 

Quote

[Another non-engineering observation from an audio forum]: “I think a lot of the audio forum mindset filters down to seeking the approval or advice of others." 

 

Not just the forum mindset, I'd call that the audiophile mindset. In my experience, many who consider themselves “audiophiles” are extremely insecure about their own opinions. They need outside approval to find whether something is “good” or not. Witness those that obsess over the latest S'Phile-recommended component lists, or suddenly get disenchanted with something because someone else tells them it is no good.  The best thing one can do in this hobby is to learn their own likes/dislikes, trust that and gain some confidence in their own choices. This results in much less gear swapping and maybe a little more focused path when they can do that.]

 

This is the crux of loudspeaker discussions that seem to pertain to the subjective inclinations of the "audiophiles".  The old phrase...

 

Quote

"measure with a micrometer, mark with chalk, cut with an axe"

 

...seems to be tailor made for audiophiles that argue whether they can actually hear DAC SINAD at below -100 dB, but then are totally oblivious or otherwise insensitive to what their loudspeakers/rooms are doing to the signal after it leaves their setups in-room and travels to their eardrums.  This is endemic to the discussion, unfortunately, and one that pulls otherwise objective minded individuals into the subjective abyss.  I would encourage those itching to comment here to instead wait for a little while and digest the discussions as they unfold before pushing the reply button.

 

I'll break here as a good place to end the first post on this critically important subject.  More to come.

 

Chris

[semente]

Great subject.

If you cannot name the engineer nor speaker brand, will you provide a comprehensive set of measurements of a couple of his designs?

[Chris A]

So assuming that tacit approval has been granted by the admin (whose charter for this subforum is repeated above in order to get his attention on the forum...and intervene if not), continuing on with the introduction...

 

So what divides the wheat from the chaff in this business (loudspeaker design) seems to be that the consumer or casual DIY builder of loudspeakers uses mainly subjective or perhaps a mixture of subjective and objective performance measures (the set of which requirements specifications is often not validated as relevant to actual loudspeaker listening performance).  This is never the case for those loudspeakers that are designed in large enterprises committed to the production of loudspeakers for a living.  Commercially available loudspeakers were built to a set of stated or agreed-to performance requirements, a set of form/fit/function constraints, or some mixture of acoustic performance plus size plus shape plus appearance requirements. and unit production price.

 

What are the requirements that they use to develop loudspeakers?  It turns out that these are typically held as trade secrets by many/most large companies. The hierarchy of these requirements--i.e., their relative importance in order to trade their exact performance levels in the finished product.  Certainly the detail requirements in terms of individual driver performance, enclosure performance, and crossover performance (either passive or DSP/active) are usually held to be only consumable inside the company, and only by authorized employee functions, such as engineering, marketing, and executive organization (including product planning). 

 

So how can I relate the requirements and their relative precedence hierarchy/trade-off structure? I can't--not to the minute level of detail required by a particular design (except for my own personal experiences--which is at a lower level of importance here). 

 

...But I can related the priority requirement types and why they are deemed important--because that's what I was gifted with, and the usability of those results have been noted by others.  So the details of how each particular driver or crossover or enclosure for a particular project may be company private (but can be inferred by the parts used in the finished product), or the processes used in taking bought components and modifying them to company needs are considered to be trade secrets, the "things to look for" and their appropriate importance/threshold level of performance has been shared over those 13 years of apprenticeship mentioned above.  Not that I've uniquely qualified to be the source of this information, but it was the fact I was picked by this noted teaching engineer to be included in a small group of non-company individuals to share these insights with--over an extended length of time--that's most useful to this discussion. 

 

So I'm not generally speaking for myself, but rather drawing the experience of that expert.  It is these insights which do make the difference, I've found.  So let's carry on and see where this goes...

 

First up, the most important single design goal in this expert's experience: low distortion numbers.  What kind of distortion?  Well, harmonic distortion (HD) at 1/10 the rated input power of the loudspeaker is an overriding requirement.  But HD is a stand-in for modulation distortion, which the HD turns into when using real music inputs to the loudspeaker:

 

Quote

[Harmonic] distortion is normally measured at 1/10 the input max input power for a single speaker/driver (AES standard). When I compare two speakers, I pick a target SPL that would be about average (I think in [the case of the present loudspeaker] it was 100 db at 1m.) [Harmonic and modulation] distortion are usually nonlinear. So at low volumes it’s not there, but then as you increase the volume, it goes up exponentially. So an increase of .1% to 1% is okay but an increase of 1% to 10% is very audible and annoying.

 

How do you measure modulation distortion?  Dual tone measurements...

 

 

The notional sideband levels seen here are from a production product from that company.  This example is not at the 100 dB rated level in this case, but are provided only as a visual example of what you'd be looking for.  The Klippel nearfield measurement scanner (NFS) automates these measurements so that the stacked set of dual-tone measurements can be compiled into one chart for many dual-tone measurements.

 

Next up: directivity response

 

Chris

[Chris A]

The next set of performance parameters is usually alien to most people scanning through loudspeaker performance specifications: directivity indices vs. frequency (otherwise reported as "smoothness of sound power").  The most important criteria include smoothly changing directivities vs. frequency, especially at the crossover frequencies.  The directivities of the lower frequency drivers/waveguides must match the upper frequency driver/waveguide (waveguides are mentioned if they are present) at the crossover frequencies.  If they don't match at the crossovers, Toole has shown that these loudspeakers will suffer lower subjective performance evaluations (see his book Sound Reproduction) and Sean Olive's patent on subjective loudspeaker parameters as they correlated to overall loudspeaker subjective performance.)  Here is a pie chart of those performance parameters as measured by Harman/JBL for reference:

 

 

(Note that on-axis SPL response is the easiest to correct, but directivity/sound power defects cannot be corrected), so the discussions concerning Olive's performance parameters will necessarily be significantly different than that author reports.)  You may notice the notable absence of distortion in Olive's parameters--something that differs from the set of requirements precedence hierarchies noted above and by JBL/Harman.

 

Chris

[barrows]

@Chris A, It is a requirement of these forums that if you are associated with a manufacture, that you reveal your association in your signature.  I cannot tell for sure if you are actually associated with a manufacturer or not from your previous posts?  Please do no mis-understand me, I welcome interesting and informed discussion of objective loudspeaker design elements, but being associated with a manufacturer requires transparency of such.

[Chris A]

No formal affiliation with any manufacturer, no compensation, and no money at all involved.  That is, in fact, the wonderful part of this role: it's entirely voluntary and self-directed. 

 

You can infer from this the motives of the individual involved, but note that he did all of his mentoring in his spare time--sometimes with the blessing of the company he works for and the off-hours use of its test and evaluation facilities, but always in a no-special-arrangement relationship other than the interests of the class of students and the willingness of the mentor to share what he could share with us. 

 

As I mentioned, there are things that he couldn't share (as in specific performance parameters, suppliers, and trade-sensitive internal manufacturing processes) but the lion's share of the learning was directed at testing (anechoic) and modifying loudspeakers they had on hand and first-hand listening sessions using at least single-blind listening techniques (often double blind when the differences were very small).  There was a lot of homework that ensued over the years, too.  That's something that a lot of casual DIY loudspeaker engineers are usually not willing undertake.  Those sessions were like gold, however.

 

Chris

[Chris A]

There are two more main requirements categories to be discussed: inherently linear/flat transfer function (small signal SPL and phase response), and impulse/decay response like that which is found in ribbons, planar loudspeakers and certain waveguide-loaded drivers.  These are higher importance requirements categories used by certain companies that have chosen to specialize in those areas.  But for the moment, I'll let the above thoughts germinate into deeper insights as to that which occurs in today's hi-fi loudspeaker marketplace.  There are some points to ponder there that you will not find in any textbook or audiophile journal.

 

The real wisdom of the discussions with the particular mentoring loudspeaker engineer reside in observations of the particular technologies used by his company.  These are the real gems of insight that can be used to great utility in one's own designs, if using those technologies, and on the more generalized insights in the entire loudspeaker marketplace. 

 

Chris

[Chris A]

3 hours ago, semente said:

Great subject.

If you cannot name the engineer nor speaker brand, will you provide a comprehensive set of measurements of a couple of his designs?

Sorry--that would identify the man to identify his designs--because he is too well known.  However, I've tried to share some of the measurements, above.  There will be more discussions on measurements that will follow, as well as discussions on how my views have evolved from his views.  A lot of independent thinking and experimentation can occur over 13 years...

 

One of the things that I've learned over time: the more you know, the more you understand why other manufacturers do what they do, and the rationale for their approaches become more apparent.  This is also true of not only loudspeakers, but also the differing amplifier topologies (including old technologies like tube) and crossovers (active and passive), and room treatments, etc.

 

Chris

[feelingears]

In the spirit of wanting to learn something here about speaker/room interactions and how measurements inform the future PhDs of audiophile progress, it's unhelpful to use biased statements like, "...the subjective abyss" even if it's accurate in the inherently subjective nature of human auditory experience. Do wine tasters go at each other this way? Do coffee drinkers? Chefs? 

 

Instead, perhaps we may learn something about why measurements don't seem to capture all that people report hearing. Statements made in the first post suggest that all of an input signal's characteristics are known, and that those interactions result in all observable changes in data. It seems safe to observe that this is not in fact the case, and listeners are listening for different things that for them result in a sense of sonic satisfaction.

 

Having auditioned and heard the effect (or not) of many tweaks and traditional components in my listening room the last five years, I know that what I hear/listen for differs from most others. I've also observed the dearth of language available to communicate this phenomena to others, let alone measurements. Sure, some of that is the fact I am not an audio pro in any way and I don't play one on the Internets.

 

So the point is, I hope @Chris A will enlighten us with useful language concepts as much as measurements, rather than furthering yet another divide in this world.

 

Cheers.

 

[Chris A]

8 minutes ago, feelingears said:

In the spirit of wanting to learn something here about speaker/room interactions and how measurements inform the future PhDs of audiophile progress, it's unhelpful to use biased statements like, "...the subjective abyss" even if it's accurate in the inherently subjective nature of human auditory experience. Do wine tasters go at each other this way? Do coffee drinkers? Chefs? 

 

Instead, perhaps we may learn something about why measurements don't seem to capture all that people report hearing. Statements made in the first post suggest that all of an input signal's characteristics are known, and that those interactions result in all observable changes in data. It seems safe to observe that this is not in fact the case, and listeners are listening for different things that for them result in a sense of sonic satisfaction.

 

Having auditioned and heard the effect (or not) of many tweaks and traditional components in my listening room the last five years, I know that what I hear/listen for differs from most others. I've also observed the dearth of language available to communicate this phenomena to others, let alone measurements. Sure, some of that is the fact I am not an audio pro in any way and I don't play one on the Internets.

 

So the point is, I hope @Chris A will enlighten us with useful language concepts as much as measurements, rather than furthering yet another divide in this world.

 

Cheers.

 

 

Sorry, but your comments are specifically out of bounds for this subforum (i.e., subjective/non-measurable psychoacoustic factors).  I plan to stick to the guidelines laid down by the forum admin (and you probably will cross those lines to your own peril, I might add).  There are a lot of psychoacoustics discussions that are out there that you are probably not aware of already,  but this forum and in particular, this thread, are not really focused on those topics.

 

If you want to talk about those items, choose another subforum and start you own thread.  Perhaps I will be able to contribute there.

 

Chris

[feelingears]

Ah well my apologies for inadvertently crossing the lines! I nonetheless look forward to learning something from your  posts.

 

Best,

Laurence

[Chris A]

No worries.  I hope that you will find something here that you can use. However, it might take a little while to fully develop the topic, so patience might be a good characteristic to have extra helpings of here...  😉  Some of this is probably going to be a bit subtle for those not into loudspeaker engineering. 

 

Chris

[cjf]

6 hours ago, Chris A said:

First up, the most important single design goal in this expert's experience: low distortion numbers.

Where does this mystery individual rank speaker cabinet design in the list of must haves when attempting to achieve low distortion numbers? The cabinet appears to be much like the frame of your car or the foundation of your house. Compromises made here would seem to only pollute anything else you attach to it thus resulting in higher distortion numbers.

 

At least that's the pitch of the big money brands who tend to spend alot more time on designing exotic cabinets made from even more exotic materials in an effort to achieve the least resonant box possible. The more flavorful brands tend to try and let the box resonate more in an attempt to behave more like a musical instrument.

 

Which of these mind sets does the mystery individual belong to? Or perhaps neither? (which would be interesting)

[Chris A]

Rigid box--no flexing is the goal (same for the waveguides).  This of course must be traded against weight and material cost, but in general, the trend over time has been toward stiffer boxes.  Remember the goal of a loudspeaker discussed above: it's not a "maker of music" but rather a "reproducer". 

 

Chris

[semente]

You've mentioned "box" and "waveguide".

Since there are a few different speaker topologies (panel and open-baffle, waveguided box, horns) and because neither is perfect will you be looking into which characteristics make the best speaker for each one of them?

Or have you already settled for one particular topology which is a way of defining your preference?

 

Speakers are not listened to in anechoic rooms and different topologies interact differently with the room.

Besides audiophiles do not show a distinct (majority) preference for either wide or narrow dispersion nor for either treated nor untreated rooms.

[Audiophile Neuroscience]

2 hours ago, semente said:

audiophiles do not show a distinct (majority) preference for either wide or narrow dispersion nor for either treated nor untreated rooms.

 

All audiophiles prefer a treated room. I know this from my mentor, the unnamed guru who told me so 🙄

[PeterSt]

2 hours ago, semente said:

Or have you already settled for one particular topology which is a way of defining your preference?

 

That.

 

27 minutes ago, Audiophile Neuroscience said:

All audiophiles prefer a treated room. I know this from my mentor, the unnamed guru who told me so 🙄

 

And if you are not careful, my friend, PWK himself will come after you.

 

haha

[Chris A]

4 hours ago, semente said:

You've mentioned "box" and "waveguide".  Since there are a few different speaker topologies (panel and open-baffle, waveguided box, horns) and because neither is perfect will you be looking into which characteristics make the best speaker for each one of them? Or have you already settled for one particular topology which is a way of defining your preference?

 

Speakers are not listened to in anechoic rooms and different topologies interact differently with the room.

Besides audiophiles do not show a distinct (majority) preference for either wide or narrow dispersion nor for either treated nor untreated rooms.

Patience, grasshopper.  All will be revealed in its time... 🙏

 

Continuing on, now taking up the more well-known/advertised design capabilities of transfer function response and time-based response (impulse, decays, etc.), these capabilities are more easily shown in a hierarchical format...

 

 

I created this notional loudspeaker capabilities/defects hierarchy a couple of years ago for talking purposes in another thread, and some of the naming and placement of the quantities would likely change a bit, but for the purposes of this thread it's good enough for now, I believe (i.e., not necessarily complete).

 

The items in green background come from Sean Olive's paper on subjective loudspeaker preferences (i.e., a statistical correlation to a relatively large number of blind-blind listener trials performed at JBL-Canoga Park).  The two items in lighter blue background are correctable to some degree using digital signal processing.

 

So for the discussion of transfer function response, about half or perhaps 3/4s of the loudspeaker performance nowadays can be "corrected" via use of IIR or FIR filtering.  Crossover response in this context is dominated by the all-pass filter (phase growth) effects of the chosen crossover filters, assuming IIR filters are used (analog or digital).  FIR filters permit control of both amplitude and phase response vs. frequency, thus allowing the crossover designer the freedom to correct both amplitude and phase independently--at the cost of added time delays for the loudspeaker amplification channel overall.  Additionally, time-based response of impulse and even decays can be controlled via signal processing techniques. 

 

The point of the transfer function and time-based response portions of loudspeaker design is that these capabilities are largely correctable (to a degree) if the consumer wishes to correct these defects.  This is what I'd term the "new age" of home hi-fi loudspeaker development, which the studio monitor marketplace has largely already accepted.  There are also some PA installations with limited FIR filtering corrections, but these are limited to the inserted time delays which detract from their usefulness in live sound applications.

 

The distortion hierarchy breakdown (which I identified as the most important in the above discussion) shows some distortion types.  Notably missing is compression distortion (including thermal compression effects on crossovers and heated voice coils).  That type of distortion characteristic was omitted from the hierarchy originally because I typically don't deal with lower efficiency loudspeaker designs, but if you were, you'd have to add that as a significant source of distortion.  This would disproportionately affect planar membrane loudspeakers (direct radiating) and other loudspeaker designs employing only direct radiating drivers on flat or open baffles.

 

This is a good place to break for now...

 

Chris

[Superdad]

On 9/20/2020 at 10:16 AM, Chris A said:

I've learned a great deal, which I've been able to turn into my own loudspeaker designs that might pass muster in any first-rate audio engineering company. 

 

Hi Chris:

Do you have photos of your aforementioned loudspeakers that you can share?

You posted this photo of your system in your profile:

And you described the loudspeaker system as follows:

"2007 Klipsch Jubilees (two-way w/TAD TD-4002 HF drivers in front corners), full-range multiple-entry K-402/BMS 4592ND with two Crites 15" cast frame woofers (center), '82 Belle bass bins with bi-amped ESS AMT-1s (surrounds), 14 Hz DIY tapped horns behind each Jubilee bass bin (subs)...a fully horn-loaded 5.1 surround sound setup."

Seems like some fine DIY (though you room looks rather tiny for those large horns), but not really a completely original effort.  Or is there some other system you have built?

 

Thanks!

 

 

[Chris A]

The one in the middle in the picture above is my prototype: it's a full-range multiple entry horn (MEH) with 90x60 degree coverage from 100-20000 Hz, and SPL response is 18-20000 Hz, and has phase and group delay response that beats most non-FIR-filtered studio monitors (i.e., the one in the above picture doesn't use FIR filtering--just IIR filters).  It was created from a midbass module from a behind-the-screen cinema setup, reusing only the horn and box.  There have been a few others of this configuration built by others based on this design. 

 

I'll talk more about my designs (if you wish) after I get through these discussions in this thread.  It seems there is a standing militia in the wings waiting to pounce on anything I say to pigeon-hole me with some particular loudspeaker technology.  That's not the purpose of this thread--which is to help others with understanding how commercially available loudspeakers are designed and developed, particularly from a requirements standpoint.

 

26 minutes ago, Superdad said:

...Seems like some fine DIY (though you room looks rather tiny for those large horns)...

That's a matter of opinion.  You should make a trip here to listen...I think you'd change your mind on that point.  I live about 25 minutes from D/FW International A/P.

 

Chris

 

 

[Chris A]

How's that for a mixed metaphor...?  (the bold text, above).

[Superdad]

30 minutes ago, Chris A said:

How's that for a mixed metaphor...?  (the bold text, above).

Well I for sure avoid militias--whether they are standing or sitting. 

[Tintinabulum]

This reads like an advert. Mini, cheap marketing campaign dressed as something else. To me.

[Chris A]

6 minutes ago, Superdad said:

Well I for sure avoid militias--whether they are standing or sitting. 

See what I mean...

[Chris A]

I like this one better (attributed to Clay Krames, RIP):

 

Quote

Like a pack of hyenas jumping on the bandwagon to get their pound of flesh.

😄