End Game DIY Loudspeakers

[mhardy6647]

I don't have any specific suggestions, but I will posit that the best possible choices will have an "X", "Z", or "L" in their model designation, and also very possibly an "i" or "e" - and an outside chance of odd caPitalization and/or Punc.tuation

All of those together -- you're talking statement product.

 

[dwkdnvr]

This is a tiny bit dramatic. Copying someone's design should be fairly straight forward.

Most home improvement shops can deliver straight to your home cut to measure planks of the wood of your choosing. Doing baffle cutouts with a router needs a bit of practice, and clamps are surprisingly expensive for what they are. But it's not as difficult as you describe it to be.

With that said, getting a finish on wood that looks similar to a Revel, now that requires SKILL. Not to mention incredibly expensive equipment if you want to do something like powder coating.

Uh, no - 'home improvement' shops don't deliver ready-to-glue-up material. You're lucky if you're within 1/4" of desired dimensions, and obviously they aren't going to do bevels, tapers, rebates, cutouts etc. maybe something you can leverage for a simple subwoofer box build, but not for an 'end game' style project.

IF you have a fully-realized 3d CAD model (and they're getting more common but still rather rare) you can work with a CNC shop to cut the material and that will work well as they can likely handle the detail cuts. WAY different proposition than just picking up the phone and calling Home Depot, though. Or, you can contract with a local cabinet/furniture maker to build the cabinets, but then you're paying custom labor rates - maybe still cheaper than commercial speakers, but you're seriously undercutting the value proposition.

And, 'finishing' isn't difficult just at the final step of applying-the-finish. Construction quality and accuracy is critical before you even get to that point as the finish - whether paint/lacquer or veneer based - telegraph imperfections rather noticeably.

So, yeah - my post would be overly dramatic if OP was talking about a simple subwoofer box that he was going to cover in truck bed liner. For something that is expected to be a legit alternative to a $20-30k commercial product, even just assembling and finishing requires non-trivial skill development.

 

[mcdn]

There is an element of drama in some of these comments. Certainly before the era of near-perfect DSP and amplification there was a lot of artistry in getting driver and cabinet integration right - and there still is for passive designs and less conventional designs, but for some scenarios it's less scary than it used to be.

With rare exceptions the expensive bits of a conventional commercial speaker are the design effort and the cabinet construction, not the drivers. An active DSP "clone" of the Revel F328be would be perfectly straightforward to execute technically for your local woodworking shop, because the DSP removes the need to futz around tweaking crossover and cabinet dimensions to the last Hz and mm - though of course the cosmetic result would be very hard indeed to match.

Now to be fair, even with DSP there's often a lot of time to be spent optimising drive unit placement for diffraction etc. For example see the Directive design thread, but that was made more complex by being a 2-way, and it's just computer time, not inherently expensive like making and testing prototypes is. Also 3-ways like the f328be dramatically simplify things by removing some of the tradeoffs that need to be made. If you start with an existing design as a basis then a lot of other design choices have already been made for you as well, such as cabinet dimensions, directivity targets etc. Even drive unit selection is getting to be fairly simple (e.g. Seas DXT is a bit of a no-brainer for a tweeter, and there are many excellent options for mid and bass)

Not that I'm recommending any of this, but I wanted to make the case that for a 3-way active speaker with DSP based on an existing commercial design you like, it's not as scary as it seems. And the result would be super-satisfying!

 

[Mr. Widget]

So, for all the DIY experts here, what would be your recommendation for an end game DIY loudspeaker that would equal or exceed what we on ASR would consider to be the best available on the retail market?

Unless you have the same resources as Harman or another top flight speaker manufacturer and you have the same level of practical design experience as their best tenured loudspeaker designers how are you going to better or match their best efforts at a lower cost? Assuming you write off all your time as recreation and provide the labor for free... you still probably won't have the ability to add a couple of turns to the voice coil and take out a few grams of cone mass as needed to achieve your goal.

Cloning a seriously good existing design is a good way to go... if you can get all of the exact drive units and know the other details of the design.

40 years ago when most speakers were built using basic TS box design and selecting from an assortment of drivers from Addax, Phillips, Scanspeak, SEAS, Dynaudio etc., you could save on the high cost of US or European labor. Those days are long gone. The best bang for the buck speakers today are seriously great deals and at the high end, the designs are generally highly specialized and require resources beyond the realm of the DIYer.

 

[hex168]

I've never bumped a previous comment of mine before (post #11), that would generally be borderline rude, but I'm surprised no one reacted to the JBL M2 clone suggestion. It's well-documented and the OEM drivers and waveguide are available.

 

[don'ttrustauthority]

DIY speakers are great because you can install high grade drivers into existing designs, just find a woodworker to make a cabinet for you.

Compression drivers and horns are fantastic sounding and inexpensive compared to home audio speakers. Find a nice woofer to suit the sound you want (efficient, deep, w/e).

Compare what you get inside a Focal, for example, for five grand, you can do MUCH better for MUCH less.

 

[McFly]

I've been at DIY speakers for almost 4 years now. I'm 33. And I still haven't even begun to design/build my ideal end game. I'm only just at the point where I have a 2-way damn near perfect (thats passive XO - DSP is a bit easier). After 4 years I'm only just starting to build my 1st prototype 3-ways, and its essentially just my 2 way with woofers added. I'm using DSP for the 3 way, as the passive 3- (and up) way XO is just too much cost and even more time IMO. And my woodworking and finishing skills and tooling capability are still nowhere near how I would expect a commercial speaker to look.

So even if I got to my DIY end-game, I know I'd still be unhappy with the aesthetic, and ultimately have to pay a CNC shop some serious dough to do the cabinets and an automotive painter to finish the cabinets.

Sometimes I think, f*** it and just give up, spend large on some nice commercial speakers and be done with it, because they look nice. But then I know ill never be satisfied.

Good luck to all who DIY. You need time, a lot of time.

 

[Raxumit]

I've made a bunch of DIY speakers, and I listen to them daily. You should only build your own speakers if you think you will genuinely enjoy the process, and want the challenge/satisfaction. Money saving is an illusion, as I've found out. You'd do far better buying a used commercial speaker if saving money is the objective.

If you care about aesthetics to the point of wanting a beautiful piece of audio furniture, DIY is mostly going to disappoint, unless you have a super talented workworker or are one.

All that said, one thing I do like about DIY speakers is it is easier to evaluate the potential performance of a given DIY design by looking at the drivers, which are often comparable across designs because they'll be the same drivers with the same part numbers. Like, if it's a Scan Speak Relevator three-way, it'll be pricey (for a DIY design, chump change for a $30k budget of course) and has the potential to be excellent.

From there you can look at the measured response and crossover design and see if you think it'll suit you. For example I really want to like Troels Gravesen designs, but his cabinets are often faceted or complex, making the woodworking out of reach for me personally. Most fatally, his response curves tend to be sloped, deliberately so I think.

I do wish there were a few more designers like him who reached for the best performing drivers, regardless of cost. The Parts Express community is focused mostly on value and very good performance, and I honestly can't fault them for it.

This makes it hard to make a solid recommendation for a maximum performance, price-not-an-object DIY speaker.

If you do go the DIY route, despite everyone warning you to be careful, I suggest building a starter proven DIY design first of more modest cost, say $500-$2,000. Find something you can get flat packs for, and hopefully a parts kit from Meniscus or the like.

 

[mcdn]

Cloning a seriously good existing design is a good way to go... if you can get all of the exact drive units and know the other details of the design.

Cloning a passive design into an active one of the same physical layout though is much easier, because you don't need the exact same drive units, internal dimensions and so on! The F328be would be cheaper and higher performing if it was active and used DSP, but that's not what the domestic speaker market wants. That's the real appeal of DIY I think, to create things that are strictly better but which the commercial market serves poorly.

Now granted the professional market does have more options, see the love for Genelec and Neumann here, but the aesthetic is very different.

 

[Raxumit]

That's the real appeal of DIY I think, to create things that are strictly better but which the commercial market serves poorly.

Totally.

My first DIY speaker was the infinite baffle subwoofer I built into my attic to use with my home theater setup. You simply can't buy bass performance like an IB sub offers commercially, at any price. I'd wax lyrical but this forum prefers objective so here's a response graph:

This is pre-DSP, from the listening position, 80 hz sub crossover. The sub comfortably hits infrasonic. The whole room is actually pressurized at high SPLs, as it would be in a bass head's car.

Naturally since nobody can easily sell you an IB sub, most haven't heard of it, and the principal forum is called, appropriately, Cult of the Infinitely Baffled.

 

[dshreter]

If the goal is to save money, buying used speakers seems like a much better path. Overall fit, finish, and condition I think you get a nicer speaker in your hands with a used commercial pair than new DIY.

 

[Pablo27]

With that kind of budget I would be tempted by these little (huge) babys:

Ellipticor-A50

You can pay someone to cnc and assemble the cabinets, this would possibly another £3000 / £4000? give or take. Obviously the bigger and more complex the cabinets the more cost is involved. You can make changes to the cabinets providing you stick to the fundamentals.

There is a builder on the DIY audio forum who built a version of the troels TLS 1 (the large black ones in the attached pic, he did add drivers at a later date). I personally love them, but they may divide opinion.

I have built the SB Acoustics 10 by Troels. They (in my subjective opinion of course), produce a "soundstage" unlike anything I have heard elsewhere, "holographic" at times. The SB Acoustic drivers are very nice.

I would not advise building a "high end" speaker yourself unless you have good woodworking skills, a very good (accurate) table saw and lots of spare time.

I built DIY because I did not have the funds to purchase a commercial equivalent that uses the same standard drivers etc (something like a wilson audio Sophia).

Personally I would take that 20 / 30k and buy some big genelecs.

 

[alex-z]

Doing "end game" speakers DIY is difficult, because certain elements like FEA optimized cabinets and CFD optimized ports are not available at the consumer level. You can follow best practices regarding bracing and constrained layer damping, but won't achieve the quality offered by something like a KEF Blade One Meta.

For that reason, my "end game" would be something like a pair of KEF Reference 5 Meta converted to active crossover, using a miniDSP DDRC-88D, trio of Topping D90SE, and 6 channel Hypex NC502MP amp. That would be $22000 for the speakers and $5200 in electronics. Or just buy a pair of Genelec 8361A and save myself a lot of time.

 

[TonyJZX]

I think this comes down to where the SOTA is now.... the 8361 and other speakers on that level are bringing 'end game' to about.... $10,000... as as the proprieter says... with the amps and dacs and dsp/room correction built in, what else do you need but your source material and to sit your ass down and enjoy until your final days on this earth.

Can you beat all that for $20k? $30k???

As people have said here, unless you're bringing something truly world beating to the table on just about every aspect of design and construction you cant really compete.

Is it all possible that Amir stamps stuff with the "SOTA as it stands now" as a symbol of the best there is at this moment...

if someone like a genelec or harman wants to go all out then can even a resourceful small team beat them at their own game? It is questionable. Can one man? very doubtful.

 

[Burning Sounds]

Not according to the reviews coming from DIY Iowa this past October. It did not fare well.

Do you have a link to these reviews?

 

[abdo123]

Uh, no - 'home improvement' shops don't deliver ready-to-glue-up material. You're lucky if you're within 1/4" of desired dimensions

I didn't know people have so many bad experiences with this, I usually order my wood online and it seems to be cut in a central warehouse somewhere by people who do it on a daily basis.

Showing up to a random branch nearby and counting on the random employee present to cut it for you might provide less satisfactory result for obvious reasons.

For something that is expected to be a legit alternative to a $20-30k commercial product, even just assembling and finishing requires non-trivial skill development.

Oh yeah in that context you're absolutely right. No one who has never held a tool before is going to make something remotely similar to the KEF Blade or something.

 

[FrantzM]

Hi

DIY from a distance, seems like a good solution, even an ideal one... until you try .. for anything. The results are usually very, very far from what you saw on the web, magazines or YouTube.
To be blunt END Game DIY requires a level of skill and knowledge that few people, audiophile or otherwise possess.
... Let me go on...
Cost and Time:
The cost of material, the so-called BOM (Bill of Material): The cost of parts required to make something... Well, after a few mishaps one quickly realizes, that the BOM is, perhaps, more the mere listing (and cost) of parts and accessories.. Most people, unless you have tried, do not realize it also involves parts, tools and sometimes, truly tangential and/or peripheral items.. such as solvent, glue, cleaners, accessories-specific tools... those tend to come when least expected... Sometimes you need a new fan in your shop because the solvent you use requires it... or ... Different drill bits for the wood type or grinder blades or ...

Finish
The level of finish, one finds routinely in, even inexpensive products, say a $200.oo loudspeaker, is beyond most mere mortal, unless they are already well-versed hobbyists in woodworking or metalworking or other craft, 3-D printing for example; else? Truly poor results ... Good finishing takes a long time, requires practice, patience and knowledge

And the cost. Even for something as basic as a sealed subwoofer: Relating my own experience. I am not a woodworker, my experience in woodworking is below zero; yet, these seemed like a good solution for my 3rd subwoofer:
There are some flat-packed subwoofer kits on Parts Express. The one I was fixated on:

Dayton Audio 15" Ultimax Subwoofer and Cabinet Bundle​

Good subwoofer with a 15 inches driver, the kit comes in CNC cuts. So all you need to do is assemble it, @ $360.00 complete and shipped? Super!
... You need these too:
Wood Glue.. About $20, total
Polyfill acoustic stuffing ... About $100.00
Feet for the cabinet.... About $30.00
Binding posts.... about $20.oo
Wires .. frankly anything 12 ga. ... $5.oo ?? Less?
Wood clamps for the required size ... The big surprise for me you need 8.. They better be good
Yep! before you know, a good set of 40 inches clamps is about $60/ pair fora total of $250.00, yep!
You decide to go "natural" so won't finish it... it stays with the wood it came in.. finger marks and glue marks on it? No problem? $0.00 for finish.
You then need an amp for this thing, 500 watts of more amp... No worry, Parts Express has those, some at as low as $250.00. Excellent.
Then you have to mount this amp in the box, It doesn't have the proper cut-out.. You have a shop make it for you, $50.oo? More?
...
At the end, your subwoofer cost you, a bit over $1300.oo ... Plus your time ( which you gave freely, who are we to account for it? ). You are happy... but on the commercial side, you could have bought an SVS SB-3000 for $1200.oo in piano black, no less, with a Bluetooth App. , or a Rythmik F15 HP for $1600.oo or ... These are likely to perform at least as well, perhaps better, much better, measurably better than your DIY ? Perhaps?? I'll grant that...
...
My own conclusion and advice:
Don't even think about DIY . It is a fool errant. Unless you are/were already deeply versed in speaker DIY.

Happy holiday!
Peace.

P.S. Edited for clarity

 

[Jukka]

A loudspeaker is a collection of compromises. You need to set a target, a goal before you can start choosing components or designs.

I DIY to save money. With 5-6k and employing a pro carpenter you can create an end game set. But you need to define it first.

 

[kemmler3D]

I think the M2 clone is worth looking into if you like the M2.

My advice would be to demo everything you can.

But I also think DIY with a $30K budget is silly unless you already have pro-level skills in (at least) driver selection, DSP, and woodworking or other means of construction.


Honestly DIY @ $30K is pretty dumb, unless you happen to be making your fortune as an expert cabinetmaker, but custom work (done by a pro) at $30K is a pretty good idea IMO.

What does it take to beat a Revel Salon? Well, better parts and at least equally good construction.

Better parts are hard to do at any price. They simply may not exist. SOTA is SOTA. Construction can be beaten if cost is no object. But only if you know what you're doing... better to hire someone for that with the $10K you don't need to spend on parts.

 

[dshreter]

Better parts are hard to do at any price. They simply may not exist. SOTA is SOTA. Construction can be beaten if cost is no object. But only if you know what you're doing... better to hire someone for that with the $10K you don't need to spend on parts.

A woodworker can make something very fine, but will it sound better? And if it doesn’t, then what? And if you’re hiring someone who is an experienced loudspeaker builder, that’s a bespoke speaker not DIY.