DIY versus Brand Name Speakers

[Atjan]

A few years ago my Thor's were alot easier to build than I expected

You trained well! 

[dstrauss]

OK, I have a question...
So what makes a better speaker?
The enclosure can only cost so much, so it can only be the driver(s) and tweeter, right?
What is the difference between say a Kef driver and a Sonus Faber Driver?
Is it only the quality of the driver/cone/magnet/wiring
What makes a speaker tick? What makes it tick better than other speakers?
Please take into consideration that I am a newbie and trying to understand the difference between good and awesome.

[Ampdog]

Firstly I will save time and go with Byrd's post #5 and others in similar vein.

Then in part answer to Dstrauss's questions: As Byrd and others said, there remains not many unknown factors in loudspeaker design, and more than one quite powerful CAD program exists making it comparatively simple to optimise such a design on your PC. But there are a number of somewhat conflicting factors in loudspeaker design. Not lazy, but matters like group delay, damping, optimal Q for this or that design, size etc. can be Googled, with better explanations than I can give here in several long posts. Others can also reply with more authority.

As is very often the case generally, it could be difficult to design and make an inexpensive loudspeaker for much lower cost than a commercial model of similar performance. But go up into the so-called 'high-end' category and one begins to pay for name. Closer to my field, one can build an amplifier for say twice as much, equalling something about ten times the cost of a commercial entry level unit. Diminishing returns and all that.

[alternativeroute]

What makes a speaker tick? What makes it tick better than other speakers?
Please take into consideration that I am a newbie and trying to understand the difference between good and awesome.

I think one of the biggest factors must be the 'recipe'. You can take the best of everything, chuck it together and it might still not sound good...

All the speakers I have built so far have been 'proven' recipes... One day when I am bigger I will try my creating own recipe...

Also see this thread: http://www.avforums.co.za/index.php/topic,7440.msg88096.html#msg88096 kinda related to the 'recipe' thing...

[d0dja]

Just throwing this out there as a theory... but maybe another factor in DIY speaker building (all things being equal in terms of components, materials, construction technique) is that the DIYer will construct, damp, design XO, etc for the drivers that will actually be in the box, measuring and compensating for actual real-life response, while the mass manufacturer (unless it's super high end) will have developed a design, and then build to that recipe, even though there are manufacturing differences between supposedly identical drivers.
 
But mostly it's a hobby.

[ghostinthemachine]

Let me state at the outset that it is not my intention to stir a hornets' nest. I simply wish to gain a better understanding of the logic that informs a decision to build your own speakers as opposed to buying them from your local dealer.

My personal opinion on this subject is the following:

PROS:
For a far lower price you can build a pair of truly impressive loudspeakers
You can finish the set in ANY color you like

CONS:
Not everybody are master carpenters
Quality drivers can be a bit difficult to source since we are on the gat-kant of the globe
X-over networks are usually a bit of drama

[Shonver]

Private builders (i.e., for personal use) are not bound by patent laws, so can clone an Earl Geddes horn/waveguide with no worries. Also, sometimes the same drivers used in boutique loudspeakers are available on the open market, enabling DIYers to clone proprietary models.

[Stereophreak]

IMHO:

I think every audiophile owes it to himself to build a speaker.

DIY guys never seem to talk in mystical and magical phrases about speakers. They might say there is a peak at about 4 kHz that needs taming, instead of "forward sounding and slightly obnoxious" etc.

Trying my hand at building for the first time has given me a whole new understanding of what goes into a speaker, and I'm learning daily. How else can you listen to a speaker without crossovers (argh my ears!), or to an enclosure designed for bass reflex, without a port (bass, what bass?), and to each driver on it's own etc (wow this tweeter does a lot more than tsk tsk)? Well, I suppose you could, but with diy it's almost mandatory.

My advice is, if you are on a budget and you are looking for great speakers, diy is almost certainly not the answer, second hand or even entry level will get you there easier and with less drama. Like Ampdog I think the real benefit of diy speakers comes at the higher end. (Should I mention JimGore's HE build again? Oops, I just did)

[JimGore]

From my own point of view, and speaking for myself only:

I have been building speakers for a really long time now.  I own all of the equipment I need, I have the cabinetry and finishing skills, I have the crossover design software, calibrated microphones, etc.

I had quite a large cash-outlay for all of the above equipment, but they have paid for themselves over the years.  When I purchase tools, I try to buy the best I can afford because I know it will last me long, meaning it will most probably outlast the time it takes for the tool to pay for itself.  Even if it does not, I can usually get parts for these tools so a fixing costs less than a new tool.

My current situation then is such that each set of speakers I build these days are comprised of an investment in drivers, crossover components, wood, finishing, consumables, and time only.  This enables me to build "cheaper" than some others who need to invest in tools.

Hypothetically speaking, if I had to build a stereo set of speakers with a R 25k budget (for myself, thus not counting time), I am confident I will be able to blow a R 100k set out of the water.  They will look the part, sound the part, and weigh the part.

Now, if I didn't have the tools and equipment the input cost may be double, which still gives a 2:1 ROI.

Another thing to consider is that there is some "black magic" in designing a good speaker.  You can using the best of everything, do everything right down to the last spec of polish on the plinth, and the speaker could still sound bad.  For me then, this final bit of the puzzle takes quite a bit of time.  It is the "voicing" of the speaker.  I have been known to try up to 6 different crossover implementations, countless resistor padding combinations, countless combinations of cabinet daming to get this right.  In my opinion, this is the line that seperates a passable speaker from a really good one.

When you do get everything to work together, it is as if the sound "snaps into place".  This usually happens quite unexpectedly: I will be sitting on the carpet with the crossover in front of me, playing around with my crocodile clip leads and resistors and caps all around me while listening to my reference tunes.  All of a sudden you clip that lead onto the network and everything comes alive: You have coherency between the drivers, the imaging is spot-on, the midrange is well defined and clean, everyting us just right.

The other side of the coin is that you can do amazing things without using the best of the best.  I have come to realize that the voicing and crossover is probably the most imoprtant aspect of a well designed speaker.  My Litil kit speaker is a point in case - the drivers are small, don't cost much, and are not all that impressive on paper.  The implementation is however quite startling.

Regards,
Ian.

[Blues]

I have come to realize that the voicing and crossover is probably the most important aspect of a well designed speaker.

couldn't agree more, and so does george short (north creek), who designed my speaker kit (rhythm). which is also the reason why I rather build a kit from someone who actually knows what he's doing, rather than mess it up myself.
this is where I believe you get the best value for money.

[bbe22]

It is the "voicing" of the speaker.  I have been known to try up to 6 different crossover
Ahhhhh so true, on spkr I modeled 16 variations, built 4 and actually listened to them over 10 weeks , then
settled on my favourite and measured again - how do u price that !?

[chipwelder]

It works very well when you have lots of time and little money, when that inverse law flips, it is hard to justify, but the learning experience of building a speaker is very very worthwhile...
Another positive is that it is playing in your room, you can taper bass response for your room, hell you can make the bass and mid response work back against the wall - if you wanted too, not ideal, but better than any mass market design will work...
It never flipping ends though. You will probably struggle to leave well enough alone.

Oh ja and you will probably never be satisfied with a commercial speaker anywhere the cost of your diy jobby, as the compromises will be too big and you have now trained your ears to listen for them.

P.S. you can choose your compromises according to your preferences

[JimGore]

on spkr I modeled 16 variations, built 4 and actually listened to them over 10 weeks , then
settled on my favourite and measured again - how do u price that !?

What I usually do is quote double or triple the time it would normally take to do the crossover - that gives me a reasonable time in which to do the voicing.

Oh ja and you will probably never be satisfied with a commercial speaker anywhere the cost of your diy jobby, as the compromises will be too big and you have now trained your ears to listen for them.

That is very true!  I doubt I will ever have a commercial set as my reference set again.  Sad but true 

Regards,
Ian.