Trade-offs in loudspeaker design

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi all,

I have recently became aware of trade-offs in loudspeaker design and couldn't find a thread about so here goes.

I've been reading here actively for past few years and somehow totally ignored/missed the importance of thinking through the trade-offs while designing a loudspeaker and I would have liked to realize it a lot sooner :) I'd like this thread to help others to get the awareness early on.

A thread without content isn't very valuable so hopefully many members would share their thoughts. This is more philosophical subject than anything specific so I'm not sure if this thread leads anywhere. Anyway, here is my big insight.

There are some "main categories" of compromises in a DIY loudspeaker design project which in my view are cost, size, aesthetics, audio quality and having fun. Some decisions one have to make along a loudspeaker design process have trade-offs in more than one of these main categories. For example limiting the loudspeaker enclosure size might affect the bass extension, cost might limit options for the surface finishing and so on. I think prioritizing these main categories for a given project provides good basis for the decision making and gives confidence to successfully meet the project targets, or fail early if the project was too unrealistic. Ability to follow the set priority would define how many of the decisions, or compromises one was able to take, were actually good for the project.

Trade-offs seem to be everywhere and I find that being aware of their existence and what they are is very important for success of any project. IF one want's to make the best audio quality loudspeaker in the world, like many do, all trade-offs possible should be chosen from the cost, size and aesthetics categories and in this order. Trying to pursuit such project one should better be wealthy, have dedicated listening space and be blind which isn't fun at all so better relax from the goal a bit :D Now go and make the project you are involved successfully finished while having confidence on the decisions and lots of fun!

edit: I found out term "Designing from first principles" that would match what I'm trying to grasp here.
 
Last edited:
diyAudio Moderator
Joined 2008
Paid Member
Some kinds of speaker bring these compromises forward, some are more forgiving.

An example could be a horn system. Space is an issue (for performance, not WAF). Parts need to go in certain places to get the right room interaction. They need to be a certain size to do it. Each way might be trimmed back as much as the nature of sound will allow. It defines/dictates the design.
 
Hoffman's Iron Law

Having a wife is one such trade-off :) Another is by this man

Josef Anton Hofmann - Wikipedia

His law (tradeoff) basically says for a given woofer, you can choose two of these three (you can't have all):

Size of box
Low frequency extension
Sensitivity (power handling)

Another good rule (more general) is similar: Good, fast, cheap, pick any two. Applies to cars, computers and sundry other things :)
 
Last edited:
trade offs are always there, it's a matter of choices. Some want to build the perfect speaker, others want something that does sound good but also looks good. Some prefer this kind of sound, others an other.

And that is a problem on this forum and almost all others, people can't understand that some other guy (or woman) wants something else than they want and start cussing for that... That should not be the case, you should be aware of that, and choose your most important factors and take the disadvantages of your choices with it without being attacked for that. Not everybody wants the perfect monitor, just like not everybody wants class d amp and dsp or a passive crossover and tube amps or...
 
Nice topic for a discussion, thanks.

Designing and building anything involves compromises, whether it's a car - ride v handling; an airliner - performance v fuel economy; or a speaker - my alleged attempts at cabinet making v sound quality and quest for value.

Aesthetics seem to me to be the largest area of compromise. I've lost count of the number of people I know who like a certain famous brand because it 'fits in with the decor' or 'doesn't take up much room'. My Hendrix-themed Classix II DIY speakers sound far better and are far cheaper than any of those four letter brand speakers but look really amateurish.

On a different theme, I recall a 1970s Sonab omni-directional speaker and amp system which would only perform at its best if installed in a particular sized room, now that's a constraint!

With regard to budget, things may not be so clear-cut: I've heard some expensive retail speakers which sounded quiet average. A decent DIY project like the Tritrix would knock them out in the sound quality department and cost but a fraction as much.

As for the fun factor, I enjoyed the challenge of making my own cabinets, soldering the crossovers, doing the Hendrix decoupage on my Classix and Tritrix and then listening to favourite music. Very satisfying!

My first pair of DIY speakers - Curt Campbell's Slapshots MTMs, really opened my ears as to how good a DIY project could sound. They even look good as a family friend made the lovely cabinets for us. The budget was set by their being a Christmas present from my wife; the cabinet size was set by the room size and layout and the need not to dominate the room too much.

The Slapshots sound better and more natural than anything I've heard in a retail shop for at least three times the price including the drivers, XO parts and cabinets.

Could I have spent more on a different project with 'higher end' drivers, etc? Certainly, but there's the law of diminishing returns to consider and as a 'Baby Boomer', my hearing isn't what it used to be, so I don't consider the budget to be a compromise in this case.


In the end it comes down to taste: I'm not going to slag off someone just because they own and like Brand X if I don't.


Geoff
 
Last edited:
Another good rule (more general) is similar: Good, fast, cheap, pick any two. Applies to cars, computers and sundry other things

In reality, it is "Good, Fast, Cheap, pick two if your design team is smart and competent, otherwise pick one if you are lucky... often you get none".

== == == == == == == == == ==

Trade-offs: A big trade-off is SPL capability, distortion, and dispersion. An example; imagine a tweeter which has low distortion down to 2.5 kHz, but below that it starts to rise, and by 1.5 kHz it is becoming unacceptable. Now imagine a choice between a 5 inch midwoofer, a 6.5 inch midwoofer, and an 8 inch woofer.

The 5 inch driver will have good dispersion up to 2.5k, so the resulting system (5 inch crossed at 2.5k) seems quite good, but it will have limited SPL capability and limited bass extension.

The 6.5 inch driver will have good dispersion up to just 2k, so now the tweeter must by crossed lower down in a region where it has more distortion. SPL capability is improved, and it will dig deeper in the bass.

The 8 inch driver will have good dispersion up to just 1.6k, but here the tweeter has nearly unacceptable distorion... The woofer has a lot of SPL capability, and it can go deep, but the tweeter may have too much distortion.

If the 8 inch is crossed at 2.5k, the tweeter is happy, but now the off-axis response is uneven compared to the on-axis response.

Large drivers have (all other things being equal) high SPL capability and usually low distortion... Small drivers have better off-axis response... This is a tradeoff... I would love to have an 8 inch driver with the SPL and deep bass capability of a 15 inch driver, and the off-axis response of a 4 inch driver... but this does not exist (yet)...
 
Some kinds of speaker bring these compromises forward, some are more forgiving.

An example could be a horn system. Space is an issue (for performance, not WAF). Parts need to go in certain places to get the right room interaction. They need to be a certain size to do it. Each way might be trimmed back as much as the nature of sound will allow. It defines/dictates the design.

Yeah totally, the who are well into speaker design and are planning to build for the audio quality the depth of details that go into trade-offs is immense. On the contrary if the cost is main priority almost nothing else matters, just hook a dumpster driver to an amp, maybe a carton as a box and enjoy. Both systems should be fun, if successful!

Having a wife is one such trade-off :) Another is by this man


Josef Anton Hofmann - Wikipedia

Hah, the classics :D well, better be friends with both and call it a successful project when there still is some bass and wifey around.

..
And that is a problem on this forum and almost all others, people can't understand that some other guy (or woman) wants something else than they want and start cussing for that... That should not be the case, you should be aware of that, and choose your most important factors and take the disadvantages of your choices with it without being attacked for that. Not everybody wants the perfect monitor, just like not everybody wants class d amp and dsp or a passive crossover and tube amps or...

This is exactly why I thought to raise the topic, it is all about the objectives and choosing the tradeoffs to get there. If the objective is not understood it is impossible to help someone, or get helped. Awareness of the topic hopefully helps people to get better help and less confusion.

...

Aesthetics seem to me to be the largest area of compromise. I've lost count of the number of people I know who like a certain famous brand because it 'fits in with the decor' or 'doesn't take up much room'. My Hendrix-themed Classix II DIY speakers sound far better and are far cheaper than any of those four letter brand speakers but look really amateurish.
...

With regard to budget, things may not be so clear-cut: I've heard some expensive retail speakers which sounded quiet average. A decent DIY project like the Tritrix would knock them out in the sound quality department and cost but a fraction as much.

As for the fun factor, I enjoyed the challenge of making my own cabinets, soldering the crossovers, doing the Hendrix decoupage on my Classix and Tritrix and then listening to favourite music. Very satisfying!

My first pair of DIY speakers - Curt Campbell's Slapshots MTMs, really opened my ears as to how good a DIY project could sound. They even look good as a family friend made the lovely cabinets for us. The budget was set by their being a Christmas present from my wife; the cabinet size was set by the room size and layout and the need not to dominate the room too much.

The Slapshots sound better and more natural than anything I've heard in a retail shop for at least three times the price including the drivers, XO parts and cabinets.

Could I have spent more on a different project with 'higher end' drivers, etc? Certainly, but there's the law of diminishing returns to consider and as a 'Baby Boomer', my hearing isn't what it used to be, so I don't consider the budget to be a compromise in this case.


In the end it comes down to taste: I'm not going to slag off someone just because they own and like Brand X if I don't.


Geoff

True! commercial loudspeakers, for the most part, must have the cost very high priority. Some don't mind about the cost but they come with niche aesthetics. This is why I think DIY speaker beats the commercial ones. Aesthetics for me, yes! I'd like speakers be invisible and sound incredible, can't find that on stores :D Takes for ever to get decent finishing.. well

In reality, it is "Good, Fast, Cheap, pick two if your design team is smart and competent, otherwise pick one if you are lucky... often you get none".

== == == == == == == == == ==

Trade-offs: A big trade-off is SPL capability, distortion, and dispersion. An example; imagine a tweeter which has low distortion down to 2.5 kHz, but below that it starts to rise, and by 1.5 kHz it is becoming unacceptable. Now imagine a choice between a 5 inch midwoofer, a 6.5 inch midwoofer, and an 8 inch woofer.

...

Large drivers have (all other things being equal) high SPL capability and usually low distortion... Small drivers have better off-axis response... This is a tradeoff... I would love to have an 8 inch driver with the SPL and deep bass capability of a 15 inch driver, and the off-axis response of a 4 inch driver... but this does not exist (yet)...

Yeah there is lots of this kind of seesaw kind of things. Sometimes there just seems no good choice exists, until step back and look the project again. If it could be a bit more costly and sizable 3-way speaker a 4" mid + 15" bass would make this problem go away but another set of trade-offs is introduced. A lot of preferences and reality based guidelines outside from the audio quality realm often dictate the trade-offs.

I find it very interesting thinking process to go through the options, avoiding dead ends. This keeps the project ever evolving and never finishing, a failure :D Lots of fun though
 
Last edited:
Another interesting thing is that when looking for inspiration from the commercial speakers one should think about the priorities the commercial one has behind it and if they reflect the DIY project priorities. For example seeking advice / inspiration for audio quality related compromise better look the expensive end of pro audio where the audio quality is above all, be it PA or studio speakers. PA speakers of course have compromises chosen for audio in big venues but this info alone helps to appreciate their perspective to the compromises. HiFi market speakers might have the aesthetics and cost before the sound quality. Not hard rules here, mere awareness is enough to pull more useful info from the examples for the own project I think.
 
Imho the most important tradeoff is the room. Apart from those lucky enough to possess a home theater room anyway. Most living rooms are constraining to the least. Positioning of speakers is dictated by other considerations. Acoustic treatment not in compliance with aesthetic requirements. I won't mention the SWMBO here, because I myself like a well designed living room too. It has lead me to the conclusion that, when designing a system, the (im)possibilities of the room have to be into the equation. Often a great effort in speaker quality is severely crippled by the limited options of the environment. And in such situations often a less perfect system performs more than well enough.
 
commercial loudspeakers, for the most part, must have the cost very high priority. Some don't mind about the cost but they come with niche aesthetics. This is why I think DIY speaker beats the commercial ones.

Better speaker beats the worse no matter who has designed and manufactured and what is the price. Everything else is just prejudice.

Designer/manufacturer probably think that own is better, more suitable, cost effective or whatever no matter actual quality (~common or experts' opinion). Difference is that diy is almost worthless for the others because it's designed by "nobody", guarantee and service uncertain.
 
Yep, and the aesthetics is for the DIYer. Even if the quality was there they might not please anyone else. Aesthetics is the main driver for me to pursue DIY speakers. I feel most commercial loudspeakers are too flashy, I would like the loudspeakers to disappear or at least blend in.

Anyway it is a different motivation and different set of targets for every project. One project is a hobby build for a DIYer and her space and another project might be for the HiFi shops across the globe targeting some customer segment or anything between. Very different set of goals and compromises for each so comparing them might not be useful, each could still be a successfully executed project on their own. Reaching the goal of a project requires choosing the right trade-offs and appreciating them as such.
 
I tend to view cost as a constraint, rather than a trade-off. A constraint is another name for design requirement, and it is typically a go-nogo parameter. My design must be within the constraints or else it is unworkable. An example of a constraint is something like overall height: my room has 8 ft cielings, so my speaker must be no more than 7ft 11 inches. I can make it any height I want, so long as I don't exceed 7' 11". A speaker design that is 8' 1" may be heaven on earth, but I can not use it... Same with cost... we all have a budget, and while our budget may be flexible, there is an upper limit.

A trade-off is the interplay between two or more competing parameters or attributes. It would be desirable to increase both attributes, but unfortunately, reality has other plans... increasing one attribute leads to a decrease in other attributes. Managing the trade-offs means balancing the parameters to achieve some optimum result.

As long as cost is well below the budget constraint, it can be handled as a trade-off... but for me, I tend to be willing to spend right up to my budget maximum, so for me it is a constraint. :)

j.
 
frugal-phile™
Joined 2001
Paid Member
Every loudspeaker designer has to choose a large number of compromises, So many, that even if one takes a bunch of the very best equally valid designs, they will often sound very different.

Room is a HUGE variable and likely the largest. The speaker has to fit in the room both aesthetically, practicality, and its synergy sonically.

Box size. As mentioned. Size, extention, efficiency.

WAF. Often a big factor if you have an SO. Cosmetics can be very important.

Budget. Jim considers this a constraint. Another word for compromise, but his wording does point out that a clear set of goals is a good place to start.

Let’s consider one set of compromises. 1-way, 2-way, 3-way, more? In theory more drivers means one can better cover the large variation in the lengths of the wavelengths we are trying to reproduce.

But, if more than 1-way we need to add an XO. In general, XOs are an evil thing speaker designers have to deal with. Given the disparate responses (FR, dispersion, colourations…) it is hard to develop something that smoothly passes one set of frequencies to one driver to the next. And if the XO frequency is such that the quarter wavelength of that frequency is greater than the centre-to-centre of the drivers being crossed one starts to have issues (lobing & more) with the drivers sound coming from a different point in space. But getting that quarter wavelength C-C means one has to make other compromises. A 1-way* gets rid of the XO, but it takes a large cone to move a lot of air, and a small one to be able to do highs decently. So a diiferent compromise.

*(note that some full-ranges, for example those with whizzer cones, are actually 2-ways with a mechanical XO. Another compromise)

So when one is starting to design a speaker, one has to at least broadly consider: Room. Room. Size. WAF. Budget. Listening habits — if you don’t listen loud, like particular kinds of music in specific kinds of ways, do you value dynamics, soundstage/imaging, the ability to reproduce very small pieces of information, more will all play a major role in your design.

And do not forget that what feeds the speakers also has to be considered. If you do not feed it the information to the loudspeaker it cannot reproduce it. The amplifier’s output impedance, Rout, also has to be considered. Most amplifiers today have low Rout and so not interact with the speaker’s impedance. But we are seeing more & more amplifiers with high Rout (SETs, ACA, and a growing number of true current amplifiers), where any deviations n the impedance curve will directly affect the FR. And in some cases, the speaker wants to see a high Rout amplifier (Fostex FExx6 series drivers for instance).

dave
 
  • Like
Reactions: 1 user
The factor that amuses me most as a tradeoff is the polar dispersion one. The two camps couldnt be anymore different. One wants the same frequency response at any angle, the other, nothing except at a beamed-to listening position.

The classic tradeoff I see is speaker complexity to achieve ideal "polars", versus a simpler design that "beams". Each "camp" has their reasons. I wonder which one is "correct" for various listening situations?

One could probably say loudspeaker design tradeoffs depend on the listening situation, which could range from sedentary and isolated to space-active and social. Or "you are there" vs "they are here". Now even that seems contradictory - no wonder speaker design compromises are so difficult to suss out!
 
Related to compromises, I think goal dependent decision tree hierarchies would be helpful.

IF {want center channel}
THEN

People come into the forum and say they want to build speakers but they don't know where to start because they don't understand the decision tree structure.

I think I'd put room size and speaker location at the top of the hierarchy. In turn, that lets you know how to evaluate your first set of compromise decisions. For example, hifijim mentioned driver diameters vs SPL/polar response in Post #6.

IF {Small room with 250Hz transition}
THEN: your main-channel drivers don't necessarily need to push below 250Hz.

You'd pretty much need to dedicate a website with hyperlinks and an appropriate navigation structure to make something like that usable.

Also, there's psychoacoustic knowledge black holes that result in shrugging when it comes to getting an definitive answer on some compromises. Some DIYers will claim certain compromises are critical when there's not sufficient psychoacoustic research to support their opinion. Yet, they may have great physical science reasons to support their opinion. So there's a hierarchy within the compromises; some compromises are rock solid while others are questionable.

Or course, the ultimate confounder of compromises is the Circle of Confusion. Some people claim that making good speakers is actually detrimental in the face of poor recordings because you can hear the flaws. Or that narrow polar responses degrade poor recordings while wide polar responses improve them. In turn, a goal based hierarchy would help alleviate that confounder a little bit because a speaker for home theater might generally benefit from the higher recording quality used by Hollywood studios.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.