Hi diyAudio: I have a question about modifying a 3-way crossover, if it becomes necessary.
I’m interested in building a 3-way DIY speaker kit. The components and design come from a reputable supplier and designer. My concern is that the voicing of the loudspeaker might not be to my liking. My current 2-way speakers, which I like a lot, have a “Harman curve” with a slight emphasis in the bass and a de-emphasis in the treble. The off-axis response is also good.
I’ve heard expensive commercial speakers that I didn't particularly like. So this potential issue is one of preference not a criticism of the designer. The problem that we all recognize is that you don’t know how things are going to turn out until you’re done. I did a DIY speaker project 20 years ago and came up short of expectations because of the crossover.
@giralfino has stated that a flat on-axis frequency response doesn’t necessarily make for a good speaker. Fair enough. Is there anything you can do about that situation besides rototilling the whole crossover design and starting over?
I know some will suggest an active crossover, which makes a lot of sense, especially if I had a streaming front end using Audiolense or JRiver. However, I don’t. I’m old school and just listen to CD’s. And I have a lot of them. Plus, I like the idea of set and forget.
I’m on my 2nd act, this time as a furniture maker (unpaid ;-). This project isn’t about saving money but answering the Siren’s song.
I’m interested in building a 3-way DIY speaker kit. The components and design come from a reputable supplier and designer. My concern is that the voicing of the loudspeaker might not be to my liking. My current 2-way speakers, which I like a lot, have a “Harman curve” with a slight emphasis in the bass and a de-emphasis in the treble. The off-axis response is also good.
I’ve heard expensive commercial speakers that I didn't particularly like. So this potential issue is one of preference not a criticism of the designer. The problem that we all recognize is that you don’t know how things are going to turn out until you’re done. I did a DIY speaker project 20 years ago and came up short of expectations because of the crossover.
@giralfino has stated that a flat on-axis frequency response doesn’t necessarily make for a good speaker. Fair enough. Is there anything you can do about that situation besides rototilling the whole crossover design and starting over?
I know some will suggest an active crossover, which makes a lot of sense, especially if I had a streaming front end using Audiolense or JRiver. However, I don’t. I’m old school and just listen to CD’s. And I have a lot of them. Plus, I like the idea of set and forget.
I’m on my 2nd act, this time as a furniture maker (unpaid ;-). This project isn’t about saving money but answering the Siren’s song.
Last edited:
@motokok I have used all kinds, such as conventional component equalisers and software ones. The goal is just a combination of searching for a natural tone while dealing with peaks you hear, and the dips you can discern. There are many things that will not sound right no matter how much you eq and you may need physical fixes for those. How you keep it locked in is your choice. You could measure the output of an equaliser component and used the curve in a crossover simulator, or just leave it in software.
