Were there any guides or tutorials you followed? 20 years ago or so, I knew someone who built their own speaker cabinets and I couldn’t believe DIY speakers could match companies that spent hundreds millions on audio engineering. I’ve heard enough anecdotes that I would be interested in learning more.
Jeff Bagby is another big name in the DIY speaker world, although he died of COVID in March 2020 and I'm not sure there's a centralized site of his designs :(
I'd like thank you and the GP for posting these links. My daughter has a synthesizer that she would like some monitor speakers for and making them might be a fun project. These links are great.
Very important: price and sound quality are at best very loosely related in the consumer market! I mean off-the-shelf products here. Never assume expensive stuff sounds great. This is counter-intuitive but true.
Thanks for these links. Do they make kits for soundbars? I'd like to get something short to put in front of the TV and it feels like I should be able to make something. I guess the problem would be integrating ARC into it.
I have a semi-crappy Onkyo CR-N575 bookshelf receiver[1] on my desk because I wanted a small system that could play CDs, radio, Bluetooth, Spotify, etc... (there's not much out there for this anymore).
I've always wondered if buying (or now I'm thinking about making) better speakers would make a noticeable difference in the mediocre sound? I only ever play at pretty low volumes and the receiver itself is no gem.
The manual says RMS output power is 20W + 20W, 1kHz, 10% THD 6 ohms, but then a couple of lines later says THD is 0.05% line in, digital in 1kHz 1W. That's pretty bad, right? If so, then speakers may not be worth upgrading. If they are, how could I tell if these[2] are compatible?
There's a variety of books out there. Sound Reproduction by Dr Toole is a very comprehensive reference to the big picture, but doesn't really get into speaker and crossover design specifically. There's other books you can find for that, but I'm unaware of the most recent good ones as I read most of this stuff decades ago. In that era the Vance Dickenson book was a big one.
The main reasons DIY can rival hifi are:
1. Good components are available to DIY builders. Yeah there's some proprietary stuff that's cool you can't get. But there's a ton of very top quality stuff you can get.
2. Retail hifi has insanely restrictive costs. They'll save every penny possible. So simply by buying better capacitors, etc, or using a slightly more complex crossover, or building a bigger box than is fashionable you can gain an edge.
3. The digital room correction systems offered on most home AV receivers today do a lot to fix simple flaws in DIY designs, giving more slack.
I honestly believe crossover design is overrated these days (for DIY). It was a necessity in the past, but today we have cheap amps and great DSPs. "The age of the active speaker" I'd call it. The smallest ADAU145x can do the crossover (and corrections) for a 2-way speaker. If necessary even as a FIR. Products I have are the freeDSP-aurora (8 channel DSP, 300€) and a Beocreate (4 channel DSP+amp, 170€). Both use the ADAU145x family, but the aurora is more powerful.
Of course this only applies to DIY. Though I wonder what's the cheapest commercial speaker one can get with an ADAU145x in a crossover-less design.
Yeah, I didn't wanna just blather at people but I should have mentioned that with minidsp and all its descendants you can treat the crossover design as a software problem, which definitely simplifies.
But also what I was getting at with mentioning the common room correction systems is these days, you can do something like take a textbook Linkwitz Riley 4th order crossover, do the little bit that's necessary to match levels, and then just call it good and let the receiver's DSP handle the large scale FR response. That advice might annoy some more knowledgable people, but the blunt truth is it'll work and done with reasonable care you'll have a truly impressive system for the price.
I don't know pro gear that much but my understanding is there's a bunch of vendors now that package DSP + ClassD amps in a way the speaker companies can white label as their own. Kinda like half the AV receivers on the market are ultimately mildly tweaked Audyssey designs.
For those who want the ultimate TL;DR on Dr. Toole's book, it's something like this:
Listeners prefer objectively accurate (ie, the input matches the output) audio reproduction, with some added emphasis on the bass frequencies. Because most of the sound we hear indoors is actually reflected, it is crucial that frequency response of the speakers remains constant as we move off axis.
This might seem blindingly obvious, especially the first part, but it's groundbreaking work as far as demystifying audio is concerned.
For so many decades, the snake oil salesmen dominated the hifi industry with bogus claims. They wanted us to think that building speakers was like crafting the perfect wine. But it's not. It's engineering. =)
The alternative (or an engineering trade off!) can be doing the opposite: Only care for on-axis sound, and while at it put most of the energy on the axis to reduce energy of reflections. Drawback: Works best with one or at most two listeners, plus speakers and listener are constrained in their relative positions.
Bad for the living room, not an issue in a study.
The bundling is a matter of membrane diameter and frequency. Bigger diameter result in stronger beam forming, however lower frequencies have less beam forming.
You can see the effect very well in the directivity pattern of this 8" wide band speaker: https://www.visaton.de/sites/default/files/dd_product/B%2020...
I've a FAST/WAW with a 4" wideband (250Hz XO). As personal monitor speakers these work great with an amazing stage and "resolution" (the illusion of being able to pinpoint the exact spatial location of a single instrument).
//woah: Thanks you two (you&GP) for triggering me on this (not /s)! While grabbing a coffee I spent a few minutes thinking about beam forming in audio applications; and came up with the line array. Not so novel, I know, but I'm happy to have gotten there on my own.
Late reply so I dunno if you'll notice this, but I think you'd enjoy reading about Keele's Constant Beamwidth Transducer line array design. He passed away some years ago, but archives of his website are around, and his AES papers are still there. The TL;DR is he adapted ideas from sonar beam forming to audio reproduction.
What I love about Toole's book is it's basically the whole big picture of audio, completely sourced and grounded in real deal empirical research, but research that also worked to correlate objective measurements to subjective preferences.
All those internet arguments over the years about what matters and what doesn't, tubes vs transistors, class A vs class AB amps, you name it... Toole's book is the cliff notes of what's actually real. It's such a huge advantage vs when I first started reading this stuff in the 90s and it was a lot harder to find information that wasn't just someone's "golden ear" opinion imagined out of thin air.
No one spends hundreds of millions on audio engineering. Or at least, not at achieving the best audio experience. The company that spends the most on audio R&D are probably Bose, and their whole goal is to produce speakers that look good and sound acceptable.
The reason DIY audio gear matches really expensive factory built gear, is that the secret to good sounding audio is good sounding speakers, and good sounding speakers are heavy and big, and have intricate cones made out of expensive materials. Imagine how expensive it is to manufacture such a thing and have it marketed and shipped, and you'll realise they need large margins on something that's already expensive. Combine that with that the actual construction of speakers is very simple anyone could it do with a basic set of woodworking tools, not to mention that many people enjoy doing it as a hobby.
Of course good amplifiers are also part of the story, but their development is a sort of by product of the computer industry, as better transistors has generally meant better amplifiers, with class D amplifiers basically being the state of the art right now.
I knew someone who built their own speaker cabinets and
I couldn’t believe DIY speakers could match companies
that spent hundreds millions on audio engineering
Yeah! The good news is that in many ways, audio is a solved problem - the industry is over a century old, so all the fun tech has had lots of time to trickle down to DIY folks. =)
That said, designing your own high quality (I'm defining "high quality" as "measures objectively well") speaker "from scratch" is quite a bit of work.
Most folks, even designers, start by building other peoples' designs. If you really want to design your own speakers, then absolutely go for it. It's not rocket science, but it's somewhat involved if you want to do it right.
If you just want a pair of high quality speakers for cheap (I am one of those people!) then use an existing design w/ a flat-pack kit from Parts Express or elsewhere.
