My experience with DIY and it's many frustrations that are not talked about.

[Astrozombie]

I sidelined my idea to try the CSS Critons, If I have room and feel like taking on a beastly project like the 1099s or 1299s I guess I could see the appeal, though those speakers usually need EQ with response like a rollercoaster..........

 

[badspeakerdesigner]

I sidelined my idea to try the CSS Critons, If I have room and feel like taking on a beastly project like the 1099s or 1299s I guess I could see the appeal, though those speakers usually need EQ with response like a rollercoaster..........

There are good ones that don't, like the VBS 10.2 which I'm working on two at the moment.

 

[fpitas]

Sometimes the "celebrities" are not to blame for poor performance of kits, as for example when driver designs change. The driver manufacturer might notify a large speaker manufacturer, but hardly cares about the DIY crowd.

 

[egellings]

I'm kind of wondering what the OP's point is. Design is by definition an iterative process, and it isn't for everyone. And speaker design and speaker performance is highly subjective.

The reason why someone is attempting DIY is another question. I am highly skeptical that an individual can save money while matching or exceeding the quality of a mass produced product by DIY. Especially for a budget speaker.

But I think there is a financial advantage to pursuing a high end build, because high end speakers are low volume products, meaning your costs are more similar to the manfacturer's costs. But it still takes a lot of time to design the speaker, and making the critical decisions is always harder when you are paying for the materials...

I'm building a design that doesn't exist, or atleast the closest things to it that do exist are miles out of my price range. So I potentially get something unique and high quality for a relatively low price, but it costs me considerable time and energy.

Of course starting out with a challenging project is it's own issue, so to do it right you should do multiple small projects to build experience...and at that point you might as well forget about the "saving money" incentive for DIY.

But on the other hand, any speaker you build will be the best speaker in the world, atleast to you, atleast until you build something else that you think is better!

A DYI speaker could turn out well if it was done using a tried-and-true design. To try to build a speaker that works well by guess work is sheer folly.

 

[fpitas]

A DYI speaker could turn out well if it was done using a tried-and-true design. To try to build a speaker that works well by guess work is sheer folly.

At that point you'd have to put it in a weird swoopy cabinet and sell it for a fortune

 

[ElJaimito]

Warning: longish. When people speak of DIY speakers they mostly mean building boxes - not drivers! I tried that, but too hard. I've built quite a lot of speaker boxes though, mostly squarish, starting in 1965 or so as a kid. Boxes go with drivers. I started out with Wharfedales, then weird Jordan metal cone drivers I could never get to sound generally good. Back then the speaker boxes you could buy were generally poorly damped and braced, and DIY versions could be better than manufacturer's built-to-a-price boxes. Later I moved into professional audio studio design and still designed boxes for my Tannoy monitors, foldback monitors and so on, but as time went by it became poor return for effort. Better to buy commercial monitors and improve the boxes (often easy). I must have owned around 100 different high quality speakers and monitors over the years, including many famous ones, and all of them had flaws, some of them acceptable, some not, but I sold almost all of them all eventually. Only kept a few unusual boxes: JBL, Lowther, Visonik, plus my newish Neumanns and Kali concentrics and well perhaps there's a few more around!

At the moment I do plan to build some new 2x15" 1x5" 1x super tweeter sealed boxes for my studio, and also a set of three sub-resonant sealed 15" boxes for the subs. All of these are (relatively!) easy to design and characterise as they are sealed, with high quality known drivers with specs, they will be flush-mount so I know the room matching, they are 12dB not 24dB in roll-off, and will be active with fully electronic cross-overs. Then, I am now able to measure and qualify and quantify their performance and tune the electronic crossovers accordingly, before room EQ with SoundID or Trinnov to final polish. Moreover I have other good monitors (Neumann) to compare them with, as well as some Kalis with subs, big old JBL studio monitors, and sets of high end domestic speakers to compare them with as well. It will still be a lot of work to do, but many (10+) thousands cheaper each main monitor than buying equivalent professional monitors, and I only have myself to satisfy. In these circumstances it is still probably worth building my own for production use. But note that I purchased Neumanns and Kalis and subs and am happy I did. Big boxes are easier to get right, especially sealed ones. Is speaker DIY economically worthwhile? Except in special circumstances as described, probably not. But what I know about audio was learned in part through building lots of bad boxes and making them good. That may be invaluable.

I also built studio consoles (mixers), rebuilt studio tape machines (1" 8 track), master recorders (30 ips 0.25), designed and built noise gates, electronic crossovers, etc. etc. but not power amps - I never found those cost effective to get right. I was employed as a consultant to design stuff for companies, and they paid me, so they must have been happy. I also designed the sound-proofing and acoustics for a few studios and listening and performing rooms. None of this is intended as a boast, just a reflection on a career (defined as "sliding downhill out of control"?). Doing most of it now is not cost-effective, for audio custom stuff you'd do better learning DSP programming I reckon (not that hard - there are kits). But as a way to find out what works and why and why not, doing all of the above is highly recommended...! Best of luck to all.

 

[Gahf]

Howdy, I don't feel this website....

I just chanced upon this post and have now gone through all the responses. Since the responses cover pretty much everything with all levels of advices we can receive, I had thought of giving it a skip but then felt compelled to add my notes for whatever they are worth.

1. Maybe many of us here wish to achieve celebrity status, too. But most celebrities are so in their own right. PC is widely copied by newbie speaker DIYs. The first satisfactory speaker I made was a PC design and listeners do like them. If you use decent quality drivers and not make too many mistakes which nature (physics?) doesn’t like, you too will arrive at decent sounding speakers. Troels takes it to the next level by using still expensive drivers. I have no doubt in time you would too be able to inspire many after you come back in your next attempt.
The MTG design you mention is a good sounding speaker covering the range above the deep bass. Now the deep bass is something which is the most difficult to achieve, and has, probably, a very high contribution towards making an accomplished speaker. I can only guess but with your range of drivers in your drawers you too can make a similar sounding speaker which covers maybe 80hz to 15khz if you use the right box volume.

2. One way I have found out which the good drivers are by making many mistakes in my journey. I am now instinctively able to zero down on drivers by their specifications and price point which my build requires. You will be surprised the best bass array I have built consists of an inexpensive driver built by a local Delhi manufacturer and covers the 30-140hz area easily when in an array of 4. I use pro drivers extensively for the range beyond 80/100hz and the overall speakers still sound amazing. The key is to arrive at a level of experience where you can easily separate the noise from genuine advice, and trust me, this capability too you would acquire in time.

3. I disagree. It is easy to build decent sounding speakers if you follow basic principles of physics. A starting point can be the Loudspeaker Cookbook by Vance Dickason which so beautifully teaches these basic principles.

4. Agree. I have a load of cabinets and drivers which I have discarded. But after I remodelled my approach to make separate units for separate frequencies first (so separate boxes for subwoofer, mid-woofer, mid +high) I have learnt not to make expensive mistakes. I often recycle my cabinets as I would not paint them before I hear the set, which helps in reducing heartburn. Maybe a similar approach can work for you too?

5. Once you go through the grind, it is hard not to make a better sounding speaker for less than what the market is selling for. But yes, the first 20/30k or so is the price nature demands before she lets you in her secrets. The 2-3k or below costing simple speakers will be always better straight off the shelf as every company worth its salt has this widely selling range covered well. Only when you go beyond those where the requirement is that music from your speakers moves you emotionally, you strike a real bargain building one for yourself. The path is not easy, though, and requires effort at learning and unlearning and learning again.

6. Science, or at least physics will always win. In time I have learnt not to ignore these principles. You would too, I am sure.

7. If you post to this community the photos and whatever details of the drivers you have, and in sets you can best categorise them into, you will be pleasantly surprised how many good people on this forum selflessly take you through the effort to convert these into awesome sounding speakers. I once found a mentor too and I am so very grateful that I took advice.
Building monitors is not a good starting point. I would suggest you try a 3 way with one 10”/two 8” woofers, a capable mid woofer, and a simple tweeter, using an active crossover and inexpensive amps, you might just find that you have managed to make a fairly good speaker. At this stage I would care to apply the principles in the VD cookbook but not worry about the on/off axis difference. You would get rid of the booming resonances too with tweaking the equalisation if the enclosure is made following those rules.

30 years ago, I built my first amplifier and my first speaker using the cheapest components. The entire setup costed my about $50, but I could listen to music the whole day long on those. Then I went into a job and with all the traveling and workload I just forgot what good music was like. Just before the start of the pandemic I went shopping for good speakers, and after buying a set of Bose, a Marantz, a Yamaha and a Polk audio, I was told if I wanted the quality I am looking for, it would cost me $30k or more. I then decided to build my own, and after 2 years of failures and heartburn and filling drawers with drivers and discarded cabinets, I can proudly say I have built a complete range, from single driver horns to bookshelf to towers, all measuring well on REW and also liked by listeners. Music out of my best system moves me emotionally, a feeling reported by several other listeners.

I have attached photos of some of the speakers I have built. I am willing to share specifications if it helps you.

 

[montyliam]

I thought I'd chime in here with my experience with DIY. I recently had a lot of time off this summer and decided to properly give DIY a go. Im into HT and listen to a lot of music, ranging from classical, rock, pop and electronic. I wanted to create a new LCR set of speakers that would be just as fine with movies as with music. I had previous experience using the Genelec 8030c's as LCR and loved the sound from these just wanted more SPL. The next option was to just buy the 8350's and be done with it, but I thought I would give DIY a go. I purchased a Hypex FA122, B&C 8NDL51 and used an 8" waveguide from Somasonus with the SB Acoustics SB26ADC. I spent weeks designing the boxes in sketchup and finally settled on a sealed cabinet for a THX 80hz roll off. I built a prototype box, built a DIY custom turntable for measurements and spent weeks measuring and tweaking until I got a result I was happy with. I then listened extensively in mono and really liked the sound. I then decided to build the actual versions. I spent another few weeks researching different types of cabinet constructions and settled on CLD braces. I got all the wood cut to my specifications. I then spent another 2 weeks, full days, building the boxes, which involved routing complicated shapes, sanding, filling, gluing, clamping etc. I finally got the boxes finished and painted and quite frankly was not pleased at all with the end result (aesthetically), the boxes looked good but they looked very DIY imo. I am a big freak when it comes to aesthetics, I want everything to be perfect. At this point my enthusiasm was low to even bother finishing the boxes properly, as this would involve even more sanding and filling etc. I saw online a good deal for a pair of Genelec 8350's and bought them knowing I could return them if they weren't for me. I plugged one monitor in and began listening side by side with my DIY design. The Genelec was better in every way. Not that the DIY was bad, in fact they sounded pretty great imo, the Genelec just sounded (and looked) better. I felt I had just wasted a summer building this DIY product when I could've spent a little more and just gone with Genelec right away. I did learn a lot about speaker construction and measurements along the way mind. One thing that is very important to consider is that DIY is NOT just the cost of the drivers and amps etc. I ended up spending around £1000 to get the first monitor complete. The Genelecs cost me £1350 each. Lesson learnt.

I won't be DIY'ing for a while now, speakers anyway. Subs I think I will continue to DIY. They are fairly easy to get right, I just need to perfect the stages pre-painting, so that my efforts in designing and building won't be wasted and I ultimately won't be happy with the painted product.

Rant over!

 

[neRok]

Ultimately though I find that building speakers has really cut into my enjoyment of music.

I agree with other peoples suggestions, in that it is important to keep a main/reference system functional so that you can use it, and then you can DIY on the side. I made the same mistake when I was younger with cars, in that my cars ended up off the road for so long whilst I was fixing/upgrading them. Whilst I somewhat enjoyed the DIY experience and learning all the ins and outs of carbies/camshaft-specs/etc, if I had of just put the car in the workshop or bought one already finished, then I could have had much greater experiences from actually using the car a lot more.

Fine, I see what you're saying ... here's my heightened FR version with scales included ...

View attachment 303365

The FR variation 500 to 10k is no worse than about +/- 1.5 dB (yellow line) -- hardly terrible. Also, the on/off axis consistency is pretty decent. Overall not bad if not up to the top Genelec / Neumann benchmarks. EQ will render these small deviations irrelevant.

The off-axis response of good speakers trends generally down, like my black line. And they definitely don't have off-axis responses that peak higher than the on-axis, like the areas all my orange lines point at, and especially not at worst area my red line points at (2kHz-3kHz).

 

[Gorgonzola]

I agree with other peoples suggestions, in that it is important to keep a main/reference system functional so that you can use it, and then you can DIY on the side. I made the same mistake when I was younger with cars, in that my cars ended up off the road for so long whilst I was fixing/upgrading them. Whilst I somewhat enjoyed the DIY experience and learning all the ins and outs of carbies/camshaft-specs/etc, if I had of just put the car in the workshop or bought one already finished, then I could have had much greater experiences from actually using the car a lot more.

The off-axis response of good speakers trends generally down, like my black line. And they definitely don't have off-axis responses that peak higher than the on-axis, like the areas all my orange lines point at, and especially not at worst area my red line points at (2kHz-3kHz).
View attachment 312123

Thank you for your sermonizing on DIY. I don't necessarily disagree with what your saying but, meanwhile, I have been very content with the Zaph ZRT's as my main speakers.

Again, I take your point about off-axis FR. Your 'black line' is an idealized version of that response that is perfectly achieved by few speakers, even the some of the best-measuring. Vis-à-vis the Zaph ZRT's will argue that the off-axis isn't bad: the off-axis tracks the on-axis pretty well although the fluxuations from flat are a more extreme.

However at this point I'll terminate my Zaph ZRT apologetics. Sufficeth to say the are at once highly resolving and pleasurable. Admittedly I listen on-axis and us equalization that removes residual FR irregularities.

 

[tktran303]

I think that OP brings up some good points.

The comment about

“Celebrities - Boy this hobby has a lot of celebrities in it, and while at first I listened and absorbed their information, the further along I get the more I realize many
designers are simply propped up by users with little reference as to what good sound is and not so much that their designs perform well. I see A LOT of popular designers utilizing what
I'd consider very dated approaches to design delivering sub par performance. I've had to take a hard stance and only consider designs that have extensive and trustworthy measurement suites.

I'd consider this design one of dated design principles and a poorly assessed speaker.”

I see what you see but I interpret it a little differently.

You don’t have to name the ‘celebrities’ but I know who you are referring to. Some are amateur, some are professional, some are junior, some are very senior.

I think the community just trying to be kind and considerate to each other.

20 years ago, when I was still wearing training wheels, I learnt a lot from many people.
They were my supervisors. My mentors.
And I was very grateful.

Now that I’m in my middle age/mid career, I don’t want to Trash Talk all those from whom I’ve learnt. Some people have retired. Some are sick, some are no longer interested, some are no longer active. Some are still doing the same thing they were doing 20 years ago. Why mess with success?

But I myself have now become the supervisor, the mentor for a younger generation.
All that I have learnt that is useful, I impart to the younger generation. All that was not quite correct/completely wrong/superseded by new evidence, I won’t.

I hope that my processes and methodology remains up to date. But eventually, I too,may become bored, tired, or sick. I hope to be succeeded by the younger generation, doing what I do, better than myself.

I hope that one day I can rest and become useless driftwood, as other people before me.

For now, I don’t agree with everyone’s process or preferences. But I do ask questions and answer some for others.

For instance, I don’t agree when designers use only on-axis measurements. I don’t agree with boutique passive crossover components. I don’t agree when designers say distortion doesn’t matter. I don’t agree with the idea that waveguides are a necessity. I don’t agree that dipole, cardioid or monopole are inherently better than the other(s). I don’t think that “directivity mismatch” sounds as bad as it reads.

I disagree with the lack of fuller copper sleeves in motor systems, or the dome/cone mismatch in coaxials.

But these speakers, by other designers, can still sound pleasant and enjoyable. My wife can enjoy classical music played from her phone sitting face down on the kitchen bench, but her “minimum standard” for a piano costs US$10,000. So I get that.

I believe that home / studio audio monitors are a solved problem. With the advent of DSP/software I believe that many small room acoustics issues that were discussed in that seminal book, first published in 2008, are currently being advanced in automotive audio.

So you know, in 20 years I hope my processes and methods and are also superseded…