Hei everyone! 😊
Too long; didn't read:I’m building a website to make DIY speaker design easier with features like build plans, calculators, 3D design help, and driver selection. Launching in 2 months—what top 5 features do you need most when building speakers? Let me know!
I hope you're all doing well! 😊 I wanted to share something I'm incredibly excited about, and I would love to hear your thoughts and feedback.
As an engineer and web developer who's always had a passion for DIY projects, I've spent countless hours building speakers over the years. One thing I've always thought about while working on my own builds is how amazing it would be to have an all-in-one tool that helps guide the design and building process from start to finish. A place where I could get help with calculations, design suggestions, and even visualize my speaker in 3D before ever cutting a piece of wood.
That's why I've decided to stop dreaming and start building a website that does exactly that! 🎉
My goal is to create a platform where anyone—from complete beginners to seasoned audiophiles—can easily design and build their own speakers, all with the help of integrated tools like:
• Build plans: Step-by-step guides to take you from concept to completion.
• Speaker calculators: Automatically handle the math behind speaker enclosures and ports.
• 3D design assistance: Visualize your speaker in 3D and adjust it as you go.
• Driver selection help: Get recommendations for drivers based on your design goals.
• And more...
The ultimate goal would be to have an integration with ChatGPT or similar, so you can get guidance throughout the whole process, going from an idea to build plans automagically through a simple conversation.
I’m planning to launch the first iteration of this website in a few months! But before I dive too deep into development, I want to make sure it’s as helpful and useful as possible.
So here’s where I need your help: What are the top 5 features you wish you had (or currently need) when building your own speakers? Whether it’s a tool you’ve always wanted or a problem you regularly face, I’d love to hear it!
Your feedback will be super valuable in shaping the platform, and I’d love to build something that truly meets the needs of the DIY speaker community.
Thanks so much for your time and input—I'm genuinely thrilled to see what we can create together. Let's make speaker building easier, more fun, and more accessible for everyone! 💡🔊
Cheers, Eirik
Too long; didn't read:I’m building a website to make DIY speaker design easier with features like build plans, calculators, 3D design help, and driver selection. Launching in 2 months—what top 5 features do you need most when building speakers? Let me know!
I hope you're all doing well! 😊 I wanted to share something I'm incredibly excited about, and I would love to hear your thoughts and feedback.
As an engineer and web developer who's always had a passion for DIY projects, I've spent countless hours building speakers over the years. One thing I've always thought about while working on my own builds is how amazing it would be to have an all-in-one tool that helps guide the design and building process from start to finish. A place where I could get help with calculations, design suggestions, and even visualize my speaker in 3D before ever cutting a piece of wood.
That's why I've decided to stop dreaming and start building a website that does exactly that! 🎉
My goal is to create a platform where anyone—from complete beginners to seasoned audiophiles—can easily design and build their own speakers, all with the help of integrated tools like:
• Build plans: Step-by-step guides to take you from concept to completion.
• Speaker calculators: Automatically handle the math behind speaker enclosures and ports.
• 3D design assistance: Visualize your speaker in 3D and adjust it as you go.
• Driver selection help: Get recommendations for drivers based on your design goals.
• And more...
The ultimate goal would be to have an integration with ChatGPT or similar, so you can get guidance throughout the whole process, going from an idea to build plans automagically through a simple conversation.
I’m planning to launch the first iteration of this website in a few months! But before I dive too deep into development, I want to make sure it’s as helpful and useful as possible.
So here’s where I need your help: What are the top 5 features you wish you had (or currently need) when building your own speakers? Whether it’s a tool you’ve always wanted or a problem you regularly face, I’d love to hear it!
Your feedback will be super valuable in shaping the platform, and I’d love to build something that truly meets the needs of the DIY speaker community.
Thanks so much for your time and input—I'm genuinely thrilled to see what we can create together. Let's make speaker building easier, more fun, and more accessible for everyone! 💡🔊
Cheers, Eirik
Only reflex and sealed enclosures? How wide a browser support?
Note that ChatGPT has been shown to spew about 50% misinformation with things diyAudio.
dave
Note that ChatGPT has been shown to spew about 50% misinformation with things diyAudio.
dave
Last edited:
Yes, please don't rely on AI for information at all.
While this sounds like a good idea it will inevitably in some way compete with this forum. People who are involved in speaker building are probably not going to spend the time to post on both. It's hard to predict what the actual effect will be here, but there is sure to be some change.
Anything that adds (correct) information on the hobby is to be welcomed but as others have mentioned above AI drivel scraped from the web is more likely to degrade rather than enhance a site that places any value in being factually correct. If you want to use AI as more of an expert system then that can avoid problems with being factually wrong but you run the risk of irritating people with irrelevant answers. Of course this can be avoided by having enough information to draw on but that is going to be a lot of work. Would be good to see such an expert system but I don't think they are common. If you want to use AI as more of a neural network to optimise the parameters in a loudspeaker system that is perhaps doable but how useful are the answers likely to be given the cost of specifying the system one wants to optimise. It's not that AI can't be useful it's just that working out how (in the real world) is still a bit of a challenge.
Your content list is broad and fairly well covered by other sites already some of which specialise in covering just one of your points in depth. This doesn't mean people won't find your site useful but it is likely to require a lot of work to gain traction. I am also putting together a site but have opted to focus on 2D and 3D simulations for detailed design and quantitatively understanding the physics which is not widely covered at present. If it starts to gather interest then I would hope to broaden and include other aspects. If it doesn't then I have spent time on an aspect I enjoy and not worked away for little gain on aspects I enjoy less (I have made a mistake like this in the past). Is this approach an option?
Your content list is broad and fairly well covered by other sites already some of which specialise in covering just one of your points in depth. This doesn't mean people won't find your site useful but it is likely to require a lot of work to gain traction. I am also putting together a site but have opted to focus on 2D and 3D simulations for detailed design and quantitatively understanding the physics which is not widely covered at present. If it starts to gather interest then I would hope to broaden and include other aspects. If it doesn't then I have spent time on an aspect I enjoy and not worked away for little gain on aspects I enjoy less (I have made a mistake like this in the past). Is this approach an option?
A guide in website that will ask the speaker builder...
1) Budget
2) Choosing currently available driver.
3) As linear response as possible.
4) Wide frequency response.
5) compression / power response ?
problem is choosing currently available drivers (in different countries/location) for speaker builder's needs. And each individual design need will require participation from experts.
How about a tool/app that if fed with TS parameters and requirements will spew out 'basic' speaker plans. Which can be refined later by experts ?
regards
Except they didn't laugh at the Wright brothers for long. Engineers were laughing at the AI chaps in my college in the 80s (they probably were in the 70s as well but I wasn't around then). An engineering company I used to work for introduced a couple of expert systems in the 90s but soon withdrew them. Around the same time they were involved with a neural network/monte-carlo simulation project that I wouldn't classify as a failure though I don't think it was put into production. AI has been around as a tool for 50 or so years although without the extraordinary levels of promotional BS we have seen in the last few. Doesn't mean AI isn't useful for some things. Programming support is one I expect to be pretty useful when developed. It is being rather oversold though.
To compare the Wright brothers with the billion dollar investments into ai technology does no justice to both. Different things put into one pot.
A website which gives a good overview what exists already as help pages on the web for this topic "loudspeaker building" would be great.
And please leave ai topic out. Its not charming at all.
A website which gives a good overview what exists already as help pages on the web for this topic "loudspeaker building" would be great.
And please leave ai topic out. Its not charming at all.
Eirik,
First of all, good luck with your project. It's very ambitious, but has the potential to be excellent.
A feature that would be excellent would be a cutsheet generator. Ideally, it'd optimise the layout to minimise cuts. Bonus if it updates in real-time from the cabinet design that's currently being worked on.
Integrating an Edge-style diffraction calculation would be good for optimising baffle placements. Taking that a step further, it'd be good to be able to pick a driver and simulate it on that baffle. NB - quite a lot of driver data required there.
Finally, if we could then apply passive/active crossover and/or EQ to the simulated frequency response, that'd be great. I'd want a plot of 0-degrees, 30-degrees and 60-degrees showing simultaneously, so the user can see how the on/off-axis response varies as the crossover is altered.
Bonus: an auto-EQ that makes FIR files to flatten the frequency and phase response curves with selectable sample rate, taps, etc.
Chris
It is already availlable but not in english ( translator should work though):
www.petoindominique.fr
It's not perfect but offer a great fast way to gauge a design/driver/ideas.
This site exist for at least 15 years and driver database grown extensively thanks to users which can add the brands/drivers by themself. At least it was such until someone decided to hack the database some years ago, so for now it's much more secure ( read less open source).
I really like the fact all choices made and math used are given in dedicated chapters hence you know limitations and limits of proposed design.
I think it's a good starting point for such kind of site. And i agree with Chris661, a cutsheet generator would be very nice! A sbir one too but it's probably a lot of work to implement.
Last edited:
🤣🤣🤣🤣🤣🤣🤣
Please let this loser impostor and improvised tribune where he is, for what it's worth, and his cohort of French Idiophiles followers ! When he's not bullshiting, he steals ideas from others (me included) for his website, and when he doesn't claims for it, he mentions "found on the web"... Bin voyons !
Fortunately yes it's not in English : already this guy barely writes correctly in French ! Skip it, by any means : there's much more clear and reliable information to find elsewhere on the subject, definetly.
T
@tubelectron,
I've no interest in personal issue with the guy or others things related like theft of idea, intelectual property of concepts or suchs ( you can find anything on the web, including claims of being 'inventor' of any concept/idea, dubious behavior, anythings really...). If i need this kind of info i know where to ask or who can be read to have access to historical perspective on things ( and many of those people are members here 😉 ).
He may be what you describe, i don't know the guy, neither do i follow french audiofools or anything* audiophile related in french (or other language either) as i'm not interested in this and it's not the way/reason i came to Diyaudio (rather the inverse in fact as our community is fullfiled with engineers or people with such minds/approach to subjects).
My point is his site is not a such bad example of what could be done in the spirit of EirikDyrset's project description.
As i said it's not perfect by any means but it's by far (imho) the fastest/easyest way to try things driver related.
If you have personal issues with him, ok point taken from my side.
We could talk about it in pm but i'm not interested in doing public bashing cause i've had a number of issues with some french audio 'celebrity' and i largely prefer not to talk about them in forum rather than doing 'free advertising': ...been there, done that and people being people it always end up as being benificial to them rather than the inverse ime: one of those 'celebrity' used things i wrote in a french forum for his own profit for years and then scammed me afterwhile as he couldn't made enough money from me in a transaction we had... and i largely regret to have helped his business and promoted his 'guru' attitude.
So i think i understand your point if you think you've been flawed, however from the experience i had i decided to react differently to such things. I hope my point is clear and we can move on from there. 😉
* at least i try not to get involved in such things too much... but when you want info related to loudspeaker's designer it's often the only source, sadly.
I've no interest in personal issue with the guy or others things related like theft of idea, intelectual property of concepts or suchs ( you can find anything on the web, including claims of being 'inventor' of any concept/idea, dubious behavior, anythings really...). If i need this kind of info i know where to ask or who can be read to have access to historical perspective on things ( and many of those people are members here 😉 ).
He may be what you describe, i don't know the guy, neither do i follow french audiofools or anything* audiophile related in french (or other language either) as i'm not interested in this and it's not the way/reason i came to Diyaudio (rather the inverse in fact as our community is fullfiled with engineers or people with such minds/approach to subjects).
My point is his site is not a such bad example of what could be done in the spirit of EirikDyrset's project description.
As i said it's not perfect by any means but it's by far (imho) the fastest/easyest way to try things driver related.
If you have personal issues with him, ok point taken from my side.
We could talk about it in pm but i'm not interested in doing public bashing cause i've had a number of issues with some french audio 'celebrity' and i largely prefer not to talk about them in forum rather than doing 'free advertising': ...been there, done that and people being people it always end up as being benificial to them rather than the inverse ime: one of those 'celebrity' used things i wrote in a french forum for his own profit for years and then scammed me afterwhile as he couldn't made enough money from me in a transaction we had... and i largely regret to have helped his business and promoted his 'guru' attitude.
So i think i understand your point if you think you've been flawed, however from the experience i had i decided to react differently to such things. I hope my point is clear and we can move on from there. 😉
* at least i try not to get involved in such things too much... but when you want info related to loudspeaker's designer it's often the only source, sadly.
Last edited:
In no way, absolutely no way. Most of his datas are unclear, not checked or erroneous. No really, I would not recommend to anyone such a falsifier's so-called expertise... Just think : until recently, he did not knew the existence of Vance Dickason's, Joe d'Appolito's and Martin Colloms' reference books (thank you tubelectron) !
Any good online calculator will do the job way better and quicker by far - this one, for example, is well known for years :
https://www.micka.de/en/index.php#ideal
T
Last edited:
OK.wasnt aware of that. Will browse the website.
----------------
What I was suggesting is; Old speaker drivers comes to end of life, new drivers are marketed. So a basic input tool for current tweeter, mids and woofer data* to be fed in the website and the website shows ideal/workable crossover and enclosure (Already good website mentioned by tubelectron) design.
*Absolutely critical data for speaker design would be published in text in website and than the DIYer would look for drivers in the market than use the proposed website.
regds.
Yes. Simple, well designed, immediate and efficient. In one word : pragmatic. No Arrias discourse, guru-confusing idiophile blah-blah
à la pétoin here...I wanted to thank "Micka" for his very useful, reliable online (and free) calculator tool contribution - like so many other people, I guess ! - but unfortunately, the provided contact link is not available anymore... Possibly overcrowded ? 😕
T
Alright.. as someone looking at the various options (WinISD, etc).. to build speakers.. I keep coming back to a software product that I used 30 years ago called Term Pro (by Wayne Harris). If you haven't seen that gem yet.. it does pretty much everything you're talking about doing.. only its a $250 software product and their site seems to be early 1990s with the last update 2 years ago.
The product is loaded with capabilities.. so if you are serious.. I'd buy a copy of that product and see what all it does. Anything you develop should do at least what that product does. It has a number of features others do, like driver database, box design, etc. But it adds a 3D rendering engine so you can specify box (square/wedge styles) and place drivers, ports, etc.. and visually see from different angles if the driver (Based on driver data) is too big/small for box, same with ports, etc.
Perhaps the most important feature it has is the ability to export CAD files for use on CNC to cut the parts. That's a big deal today given the easy access to somewhat affordable CNC machines (or ability to pay to have a CNC cut a file you send it and ship the results back).
Not only will you need to handle all aspects to T/S parameters, a driver database, and common boxes (sealed, ported, 4th/6th order) and designs (square, wedge, round) but account for displacement of driver, ports and bracing in the box, and model all of those accurately, but you'll definitely want ways to automate cutting things out. For example, a modern day version of Term Pro would integrate in with CNC company's to offer customers to cut/ship their design for those that dont have CNC machines. Perhaps it would also allow for customizations like acrylic/plexiglass panels (and account for the differences in how that works with sound waves to model it accurately).
You also absolutely want a way to build custom ports.. like folded ports in a box, flanges, round/square ports, and model angled vs rounded port turns (e.g. if you build a folded port to provide more length IN the box and it takes 5 turns.. use wedged cuts in the corners or rounded cuts in the corners to offer better air flow/dynamics than just no edges at all (e.g. 90 degree turns only).
Given today's tech.. can you PLEASE look at ways to allow others to support/add to it. If I were doing this today (and I have thought about it as I am a software developer as well).. I would absolutely use WASM technology and "extension points" as ways to allow others to work on WASM modules in whatever language they want to contribute to the application. I have actually written a plugin engine (in Go) for WASM extensions, and am porting it to Java, Typescript, python, rust, zig and C# (eventually). If you're interested in collaborating a bit.. be happy to discuss. But the general idea is that anyone can use the engine to support WASM plugins that can load dynamically and extend an application. It's oddly quite easy to build WASM binary modules (at least in go, rust and zig) and they are quite small and near native performance as well.
The product is loaded with capabilities.. so if you are serious.. I'd buy a copy of that product and see what all it does. Anything you develop should do at least what that product does. It has a number of features others do, like driver database, box design, etc. But it adds a 3D rendering engine so you can specify box (square/wedge styles) and place drivers, ports, etc.. and visually see from different angles if the driver (Based on driver data) is too big/small for box, same with ports, etc.
Perhaps the most important feature it has is the ability to export CAD files for use on CNC to cut the parts. That's a big deal today given the easy access to somewhat affordable CNC machines (or ability to pay to have a CNC cut a file you send it and ship the results back).
Not only will you need to handle all aspects to T/S parameters, a driver database, and common boxes (sealed, ported, 4th/6th order) and designs (square, wedge, round) but account for displacement of driver, ports and bracing in the box, and model all of those accurately, but you'll definitely want ways to automate cutting things out. For example, a modern day version of Term Pro would integrate in with CNC company's to offer customers to cut/ship their design for those that dont have CNC machines. Perhaps it would also allow for customizations like acrylic/plexiglass panels (and account for the differences in how that works with sound waves to model it accurately).
You also absolutely want a way to build custom ports.. like folded ports in a box, flanges, round/square ports, and model angled vs rounded port turns (e.g. if you build a folded port to provide more length IN the box and it takes 5 turns.. use wedged cuts in the corners or rounded cuts in the corners to offer better air flow/dynamics than just no edges at all (e.g. 90 degree turns only).
Given today's tech.. can you PLEASE look at ways to allow others to support/add to it. If I were doing this today (and I have thought about it as I am a software developer as well).. I would absolutely use WASM technology and "extension points" as ways to allow others to work on WASM modules in whatever language they want to contribute to the application. I have actually written a plugin engine (in Go) for WASM extensions, and am porting it to Java, Typescript, python, rust, zig and C# (eventually). If you're interested in collaborating a bit.. be happy to discuss. But the general idea is that anyone can use the engine to support WASM plugins that can load dynamically and extend an application. It's oddly quite easy to build WASM binary modules (at least in go, rust and zig) and they are quite small and near native performance as well.