And there seems to be more and more on social media spreading misinformation as time goes on.
jeff
Agree 100%. That is absolutely an issue that all forums deal with in varying hobbies. Misinformation, poor information, different objectives, different applications for similar information, and the ever present subjective preferences are all sources of confusion...especially in speaker building. There are so many differing views of what's best or how something should be, and many are still correct even when they don't match our own views. Pro sound for live performances, pro sound for DJ, dance clubs, and movie theaters all have very different objectives and different variables than hi fi speaker applications, in addition to the subjective preferences. I suspect tidbits from different disciplines get passed on as universal fact, when they're often really very specific. That's one of the reasons it can take a DIYer some time to figure out what direction to head.
Given the very large number of compromises any loudspeaker design has to make, that a bunch of equally valid designs can sound very different.
dave
For those who don't rely on measurements, there's one thing measurements have taught me.. it all changes when you put a speaker in a room. Even moving it in the same room. Reviewers must be challenged because I know it takes time to voice speakers, and it takes even longer to adapt acoustic issues.
My thoughts are that "better" means "sounds like what I was looking for." In the same sense that sometimes cooking at home can give you the exact set of flavors and nutritional balance you can't buy in the streets. So in that sense, I've built speakers that are far better for me than I could have afforded otherwise, if I could even find what I was looking for.
Also in the category of "better" is what do you want to own? Do you want to brag to your neighbors how much you spent at the store, or do you want to show them what you built?
I'm the kind who thinks the latter is a lot cooler way to live. As a side benefit I get to interact with some very smart and interesting people here who otherwise would never speak to me, so for that alone I'm very happy.
🙂
Also in the category of "better" is what do you want to own? Do you want to brag to your neighbors how much you spent at the store, or do you want to show them what you built?
I'm the kind who thinks the latter is a lot cooler way to live. As a side benefit I get to interact with some very smart and interesting people here who otherwise would never speak to me, so for that alone I'm very happy.
🙂
Totally agree that we tend to be less sensitive to subtle nuance in low frequencies than we are at say 500hz-5Khz-ish. I like good bass as much as the next person, and have even been exposed to my fair share of top notch bass, but it can come at a huge expense. I know the low bass is where a lot of people focus, because it's fun, but from a musical standpoint, its less critical (unless there's an obvious issue) than octaves above the bass because fewer instruments have fundamentals in that range. Honestly, I wish I could have it all, but with my budget and the cost of getting really top shelf bass, if something's gotta give I'd gladly make a compromise below ~ 80-100hz long before I'd compromise from 125ish to ~ 4000hz, where the vocal range, and the core range of so many common instruments resides.
My system is currently fairly mid-fi below 80hz, and much higher fidelity above that (officially declared by yours truly! 😊) It took me years to chase the tiny gremlins that irritate me in that critical lower midrange to the upper-mid/lower-treble range ....the sound simply mattered a lot more to me in those frequency ranges, and was a constant motivator to keep plugging away at improving it. Store bought speakers would have surely deterred me from making many of the changes that I made to the crossovers and other aspects of the speakers. Perhaps at some point I'll focus on improving the bass to the next level too, but I just haven't been overly motivated to spend the time and money for a few percentage points of improvement because somewhat lesser performance in the lower bass is simply easier to tolerate and even still enjoy.
You are so right Allen!
I starting to think it´s maby better to just begain measurements in the room direct, and save you self a lot of time and frustrations later.
If you just want to se a drivers capacity....Okey
But building a diy speaker?
And i always wonder how many hour´s John Atkinson spend on "a review measurement in the home of someone"
Like this
https://www.stereophile.com/content/magico-a5-loudspeaker-measurements
Hello
Yeah except for if you move them same issues as with a speaker designed traditionally and if you put them in another room they won't work. With a traditional design you have a better chance of them working going from room to room. IMHO
Not only that but we know what a the predicted in room response will be looking at Kippel measurements. Of course minus the schrouder range where any speaker will have issues in any room typically sized room we have in our homes.
Rob 🙂
At higher and middle frequencies a controlled directivity design should measure similarly in room as anechoically. If it doesn't then the preferred way to deal with it will often be in the acoustic design, the baffle and the positioning. If that's not an option, tweaking it flat at the listening position can be an overcorrection because a reflection isn't just a wiggle of the response. There is an element of delay, and it may be heard as distinct due to the angle it comes from. There's not necessarily a complete fix for it.
At some lower frequency the baffle becomes small and all the room walls get illuminated. We're less sensitive to the delay but keeping the response flat in these conditions below the baffle step can be challenging.
At some lower frequency the baffle becomes small and all the room walls get illuminated. We're less sensitive to the delay but keeping the response flat in these conditions below the baffle step can be challenging.
Yeah good acoustical design and then DSP output filters and EQ to make good anechoic response, then DSP input EQ to tailor to the room. And, it's mainly room modes and overall balance (to final listening angle) one can fix. Toe-in is now utilized to optimize early reflections, and then the listening axis tuned to nice balance, if it already isn't because it 's good acoustical design 🙂
But I see situations like this elsewhere, too. If some company quotes 'X' amount for a product, sure, they may use a template and have everything ready very quickly, but you can't because to get to where they are right now, you have to go through a huge learning process for which there is no short-cut.
I find that the expertise gained in doing DIY loudspeaker designs and some amount of ear training (like JBL/Harman proposes and has provided an app to help doing blind testing) really helps in the realm of getting the loudspeakers dialed in well, using something like REW to help measure what you've got, so both your eyes and ears can understand what is happening.
In terms of some companies doing incoming inspection of supplier drivers, this could be an issue if suppliers have a lot of variability in their products (variability as defined in the lean/six sigma sense).
I know one company (Klipsch) had to do 100% inspection of supplier drivers when PWK was still working there. Apparently, the only real reason why they changed drivers over the years was due to supplier issues in quality control (woofers, midrange drivers, and even tweeters).
You would think that supplier quality control systems would nowadays be able to control their manufacturing systems better than 60-70 years ago, but apparently that's not really the case. Good ISO 9001 quality control systems apparently aren't nearly as widespread as one might think.
However, using DSP to correct for amplitude and phase issues (i.e., the transfer function of the drivers) makes that dialing task much more productive nowadays than in the time of "passive crossovers only".
Chris
Smaller companies are going to be reluctant to spend the money to implement and maintain ISO 9001 certs.
jeff
Amen. I rely on measurements, but not specs. Meaning, whatever a speaker manufacturer quotes as the -3 dB point, it's a pipe dream because once you put the speaker in a room it can go in any possible direction. Perhaps the worst example is buyers who decide on a sub and configuring a sub based purely on their main speaker measurements. "Oh foolish mortals" says the demon that lies between the lowest room modes.
A lot of money?
I am sorry, but a professional wood worker can make you a quick prototype for about 200-800 bucks.
And unless you work for free, but those simulation tools don't do anything by themselves.
At an average hour rate between 75 and 200 bucks, it's often cheaper to just quickly get a prototype made, instead of spending days on simulation software hoping the results will end up the same.
(Hour rate is not the same as what you see on your paycheck, there are overhead costs etc)
This is particularly true for BEM/FEM based software.
Besides that if you design a loudspeaker, there are always many other (mechanical) things that can only be judged when seeing things in person and in real life.
A design process is never linear, so seeing things in the flesh will result in choices you would have never made by looking at just simulation software.
In my professional experience this happens extremely often.
So you can spend 4 days simulating stuff. That will you also cost roughly 2500-3500 bucks in time. Or a lot more for complex projects.
You can iterate at least 2-4 physical prototypes for that amount of money.
(Depending on the complexity, size etc)
So in the end there's a balance to be made.
Besides the fact that software programs are not a substitute for knowledge and experience.
Including experience with how things will turn out or be ready for production and assembly. Finding the right and trustworthy suppliers etc etc etc
Something software definitely won't be able to tell you but often is one of the most critical steps right from the start.
Please remember that we are speaking about diy here.
Most makers do not do not work by the hour.
Building the simulation model or prototype you decide.
It is all about the journey of doing it.
Thanks DT
You might need a set of design documents to construct the prototype.
Your previous post I responded to was CLEARLY focused towards OEMs
The rest is all a bit obvious, so no idea why to explain that even?
It's a total obvious given that diy don't charge hours. 🤷♂️
This is how the modern little speakers (JBL Flip 6) are able to produce seemingly physics bending amounts of sound out of something the size of a 600ml pop bottle that still sounds good. I was curious just how it was able to sound good (not fantastic, but good) with rock and dance music, outdoors, even far away at surprisingly high volumes, so I used a frequency analyzer app and saw how: It just completely eliminates anything below about 75HZ when it's turned up unless the particular sound is relatively quiet.
The speaker wasn't mine, and I couldn't pick the music - I'd love to hear what something like Massive Attack sounds like through it, just to see how it handles a song that is mostly below 80HZ thrown at it.
Of course, once you start to listen for it, your ears learn to hear what's missing but it still sounds far better than hearing distortion from something being overdriven or clipping.
I am in the middle of building a (much larger, but still pretty small) boombox that I can't wait to hear, and hope it's able to produce the level and quality of sound that I'm after.
Possibly some kind of adaptive filtering. A while ago I came up with an idea for an analogue circuit that could do something similar. Never implemented it, but I could still do it if I ever feel the need. Open sourcing it...
Concept:
Start with a full-range active speaker.
Create an active XO filter to split the signal high/low. E.g. 200Hz.
Take the low output and pass it through a soft clipper (there are circuits for this on the forum), specially tuned so the bass output never exceeds some limit that you give it, e.g. +/-2mm.
Recombine the high & low signals. Here the trick is to use simple first-order RC type filters so the transients sum together nicely without lag or phasing effects.
Then send the result to the amplifier.
Normally, when the bass starts hitting the speaker's power limits, it just saturates and destroys all the other signals riding on top of it.
The aim of the circuit, is that no matter how crazy and distorted the bass gets, it still leaves a spare +/-1mm or so of overhead for clean high frequencies.
Concept:
Start with a full-range active speaker.
Create an active XO filter to split the signal high/low. E.g. 200Hz.
Take the low output and pass it through a soft clipper (there are circuits for this on the forum), specially tuned so the bass output never exceeds some limit that you give it, e.g. +/-2mm.
Recombine the high & low signals. Here the trick is to use simple first-order RC type filters so the transients sum together nicely without lag or phasing effects.
Then send the result to the amplifier.
Normally, when the bass starts hitting the speaker's power limits, it just saturates and destroys all the other signals riding on top of it.
The aim of the circuit, is that no matter how crazy and distorted the bass gets, it still leaves a spare +/-1mm or so of overhead for clean high frequencies.
Hello All,
I make for for the experience of it.
First off, we they all use the same OEM parts to start with; resistors, capacitors, transformers and speaker/drivers.
I started with vacuum tubes because that is what I had. The first projects were constructed on scrap pieces of plywood. Cages were added to protect cats and children. I still use some tubes on occasion, I like the glow.
Plywood breadboards are not front and center in my living room.
I have built more than a few circuit boards with fiberglass sheet stock with press in mil-grade turrets. They are not often in my living room either.
For the bench built projects that I do usemake and use in Sparky's lab, I most often use 16Ga steel electrical junction boxes as enclosures. There is some Spice simulation and some prototyping with some amounts of iteration in between to locate the sweet spot. Sparky is my cat.
A Benchmark Headphone / Pre-Amp has permanent residence on my bench, I like the remote. It is difficult to build something that measures better.
Thanks DT