I've been looking for great sounding 3 way speakers but am coming to the realization nothing unless I spend LOTS and up nothing is going to satisfy me. I used to use Madisound for LEAP designs and made a few nice pairs 20 years ago but I see things have changed. I'm again looking to build a nice 3 way for myself with top notch drivers and xover parts, I am very technical and I use REW now for my studio and PA but have to be unpresumptuous and realistic, it would take me LOTS of time to sift through all the software options and learn exactly how to get something close.
Any suggestions would be great.
Thanks,
John
Any suggestions would be great.
Thanks,
John
I think an indicated budget would be useful to understand if this is even realistic? Designing a crossover - especially for a discerning audiophile such as yourself (...) - will take several iterations, in-room/in-cabinet measurements etc. If you want to pay someone by the hour to do that it's unlikely to be cheap.
I'm not sure how the Madisound x-over service used to work, but if you bought a simulated x-over made from raw driver measurements but nothing in-room/in-cabinet and were happy with that, I would bet you could get similar results by learning the tools such as REW, Passive Crossover Designer etc. yourself over the course of a few weekends. If you are used to REW and "very technical" I doubt it will even take you that long 😉 - and if it doesn't sound the way you wanted it to, you know who to blame 😀
What has changed in the last 20 years is that active filtering and DSP is much more accessible, whether for a final system or just for prototyping it can cut down on what you have to commit to spend on passive components by a fair margin. Again, with a budget in mind you can easily price out if you can get close to what you think will be satisfactory to you discerning tastes and see if the project is realistic.
Last, but not least: If you want the most audio quality for your money, the best recommendation is always to buy used gear. Your only up front "investment" is going to be time for desk research and test listening to find out if what you really want within your budget. Used speakers are also much more likely to retain their value than DIY-stuff.
I'm not sure how the Madisound x-over service used to work, but if you bought a simulated x-over made from raw driver measurements but nothing in-room/in-cabinet and were happy with that, I would bet you could get similar results by learning the tools such as REW, Passive Crossover Designer etc. yourself over the course of a few weekends. If you are used to REW and "very technical" I doubt it will even take you that long 😉 - and if it doesn't sound the way you wanted it to, you know who to blame 😀
What has changed in the last 20 years is that active filtering and DSP is much more accessible, whether for a final system or just for prototyping it can cut down on what you have to commit to spend on passive components by a fair margin. Again, with a budget in mind you can easily price out if you can get close to what you think will be satisfactory to you discerning tastes and see if the project is realistic.
Last, but not least: If you want the most audio quality for your money, the best recommendation is always to buy used gear. Your only up front "investment" is going to be time for desk research and test listening to find out if what you really want within your budget. Used speakers are also much more likely to retain their value than DIY-stuff.
20 years ago, these services would take an axial measurement and design a crossover to make that flat or similar.
Now you can do something similar by simulating an enclosure. You don't have to touch a real speaker. The result is going to be usable especially if you then use a global equaliser to help voice the speaker.
There is a limit to a crossover of this nature, and part of designing a crossover ought to involve ensuring the different bands and their baffling work together to produce a system with a shared goal. It's usually a safe bet to combine a 5-6" woofer and a dome tweeter, and consider baffle step compensation simply because these are known quantities, but there are still limitations.
Now you can do something similar by simulating an enclosure. You don't have to touch a real speaker. The result is going to be usable especially if you then use a global equaliser to help voice the speaker.
There is a limit to a crossover of this nature, and part of designing a crossover ought to involve ensuring the different bands and their baffling work together to produce a system with a shared goal. It's usually a safe bet to combine a 5-6" woofer and a dome tweeter, and consider baffle step compensation simply because these are known quantities, but there are still limitations.
Thanks for all the advice... The speakers I made years ago I made to replicate the Avalon's with the removed angle baffle. I had an ETON mid and tried various tweeters(no BE back then or just coming out) so I get the baffle interaction. I am not sure I could start from the ground up with a crossover design without guidance even if I got perfect measurements with REW, I'm sure there is an art to getting the phase best crossover points, and using the natural rolloff to get the most out of the drivers. I was initially wanting a 4 way with MTM but I have been reading there are problems with the off axis due to the two mids. Now I'm leaning on a 3 way with BE tweeter, TeXtreme and 2 10" Scan aluminum Revelators. It would be nice to be able to confirm the design with measurements after a design, what are the pro's using now that LEAP is basically gone ?
I'm sure if you could provide good measurements, someone could put a little something together for you.
An accomplished crossover designer can use any simulator and come out on top. Many these days use Xsim or Vituixcad but there are several others. Xsim focusses on circuit design in the way of conventional simulators. Vituixcad has more tools and methods and a steeper learning curve.
An accomplished crossover designer can use any simulator and come out on top. Many these days use Xsim or Vituixcad but there are several others. Xsim focusses on circuit design in the way of conventional simulators. Vituixcad has more tools and methods and a steeper learning curve.
I would absolutely pay someone for a starting design for sure. Ill get the tools installed and familiar with them and get some measurements to start. I have the old drivers here (Eton, Peerless 10" and some Seas alum tweets to practice on).
REW and XSim is probably best ?
What's a decent audio interface for my laptop to use for measuring ? I have Neve pre's and Apogee converters in the studio LOL but I think I need to take this outside the studio outdoors to measure right ?
Thanks !
REW and XSim is probably best ?
What's a decent audio interface for my laptop to use for measuring ? I have Neve pre's and Apogee converters in the studio LOL but I think I need to take this outside the studio outdoors to measure right ?
Thanks !
These will get you up and running, and many are familiar with them.. so good choice.
Perhaps it has a line in socket, and you have a mic/phantom power you can plug into it. Otherwise you can add a USB interface that can be looped out/in and has phantom power.
Avoid getting a standalone USB based microphone.
No, but it is a good idea as it may improve your measurements, even though there are steps you can take. Therefore it is especially a good idea if you are still learning what to look for.
REW is well-featured, but quite complex to set up due to number of different parameters. For example ARTA is simpler and has some support for automatic turning tables too, but not freeware.
XSim is quite simple, but limited. The biggest restriction in standard XSim is that it does not have acoustical domain with geometry calculation and off-axis responses analyses. No radiator simulations and response file processing either. It's more like basic circuit simulator extended with single (on-axis) frequency response file and impedance response file for each driver.
VituixCAD 2.0 has at least five times more features. Steeper learning curve means that it has five times more features to study. Features which both XSim and VCAD support are as easy or easier in VituixCAD, imo.
Member
Joined 2003
Xsim is obsolete. I would disagree on "steeper learning curve", it's more "limited functionality" and the functionality it does have is more easily captured in VituixCAD IMO. I've mentioned before "For exact same result of Xsim within Vituixcad, load single response for each driver in driver tab, enter delay in driver tab, and leave driver coordinates on crossover at 0,0,0. That’s it."
With Xsim you will need another software for cabinet model, another software for diffraction model, another software for response processing.. How is that easier? VituixCAD includes everything you need apart from physical measurements, covered by ARTA or REW or Clio, and some nice measurement guides that if followed will ensure good data and excellent results.
With Xsim you will need another software for cabinet model, another software for diffraction model, another software for response processing.. How is that easier? VituixCAD includes everything you need apart from physical measurements, covered by ARTA or REW or Clio, and some nice measurement guides that if followed will ensure good data and excellent results.
You will be surprised what you can do with a spec sheet a few measuring tools a freeware crossover design program , you could end up with some excellent sounding speakers , there are a lot of free tools ,look at DIY audio and video there are lots of really informative freeware programs you should give e it a look . I have been building for a long time with a Degree in Electronics and industrial Technology , I really appreciate the information these guys share for speaker builders .
Since I'm new and have to learn one, which measurement software will have the most longevity ARTA or REW or Clio ?
Member
Joined 2003
REW is the cheapest (free) solution, with active development. ARTA is shareware with a small license fee to save files, still supported however the developer is retired so I wouldn't expect much beyond bug fixes moving forward. Great software and easy to use, where REW can be a bit complex in comparison. Clio is a commercial kit with dedicated hardware, software, and mic. Excellent package I'm told but comes at a price. If the price is okay with you, it's a good option as a one stop shop for everything you need, rather than shopping for a mic, audio interface, and building a DIY jig. VituixCAD measure guide for Clio is the most detailed of any of them. Clio pocket is good enough for VituixCAD use.