I did that is why I know that the graphs are useful. The two speakers simulated in those graphs are mine and wesayso's. The speakers were built, tweaked and listened to long before those simulations were made as I had not learned how to make them at the time the speakers were built.
When I built mine I chose the route of a small roundover and a simpler cabinet because I was not convinced that the extra effort was worth it to make the much more complicated shape wesayso used. I always had trouble equalizing the speaker in the 3KHz range and getting a good balance there was hard.
When I made the simulations and saw the directivity issue at the same frequency I understood why I was having trouble. To avoid making the same mistake twice I put more effort into preparation and simulation than I did before.
Unfold it and the length dimension becomes enuff longer than the other dimensions so it becomes a quarter wave resonantor at some frequencies. This can be usded to advantage.
Like these.
dave
Like these.
dave
Keep in mind the context (from post one):
"If you are making a rectangular box and have limited machining/carpentry tools"
By definition, any roundover on a standard, simple rectangular box is small.
Within the stated context, good driver placement and use of felt will work well.
Also within the stated context, scouting around for an appropriate found object (e.g. a bamboo box or tray from IKEA, which already has angled walls) is probably better (if a complex, non-parallel cabinet shape is desired) than buying new tools to cut & join those angles.
Attachments
Now you're cooking with fire. There are many thick wooden bowls available at Walmart now which provide gratuitous roundovers.
This is a false premise. If you record a stereo image at the ears and play it back through a set of speakers of course it will not sound natural. It would have to be played back through head phones to sound right. If it was recorded at the ears then it would need to be played back at the ears. This is why no one records this way. Back in the 80's you might find the occasional bootleg live rock recording recorded this way with a set of bunny ears from the audience but that certainly isn't ideal. The best live recording is done with a multitude of mikes. Most of which are at each instruments source just like the singers mic. There might also be a mix of house mics used to widen the sound stage but no one is placing a set of stereo mics 6" apart and recording unless they are studying psychoacoustics. Any properly recorded live performance isn't going to suffer from this and studio recordings are most commonly recorded one instrument at a time in isolation and can be placed anywhere in the sound stage and hearing various amounts of the performance of any single instrument dominate in one ear over the other is the magic of stereo. So I don't really understand the point of this. It sounds kind of gimmicky to me.
I'm not sure you read that fully - it says if recorded at the ear but played back through a speaker, some of the sound intended for that specific ear will be heard in the other ear.
I'm curious about the use of felt - especially for the taming of sibilance. Has anyone experimented with polyester/acrylic felt or just wool felt?
I think he did, and the recording engineers know this too, they record so that it sounds proper with speakers if that is the goal. If the goal is headphone listening instead, then they might choose to make the ear mic technique if that sounds like it should in the headphones.
The recording technique is not important in a way, only the playback matters, the end result, what the listener hears. Same as with designing crossovers for a loudspeaker, you don't design them just the sake of crossovers but to make the loudspeaker acoustic output good as a whole.
The recording technique is not important in a way, only the playback matters, the end result, what the listener hears. Same as with designing crossovers for a loudspeaker, you don't design them just the sake of crossovers but to make the loudspeaker acoustic output good as a whole.
I understand completely what is said here. Maybe I'm not making myself clear. Nothing you will ever play on your system will have been recorded at the ear and if it where the only way it would sound right is if it where played back through headphones, at the ear. There are a lot of companies out there trying to apply different psychoacoustic effects to their products for various stated purposes and some maybe perfectly valid but most to me just sound weird. All of these effects are at their core distortion.
The one downfall of a great playback system is it makes bad recordings very apparent. Sometimes to a point that you just don't want to play them anymore.
I've agreed with you until this. Some systems can be 'revealing' of issues without necessarily being accurate.
I realise we are digressing from the OP, but I would question what makes a speaker great? ..one that is amazingly accurate or one that makes everything sound enjoyable? ...maybe both 😊
I understand there's a doubt that 'accurate' speakers are enjoyable to listen to. That's why I don't use the word accurate. I've been there and caught myself thinking that, only to find out later when I'd overcome the issues that I had no idea what I was talking about.
In the end it is always subjective. What you like is what you like. Sure any system can be revealing but a truly, as you say "accurate" System (your words not mine) can reveal so much more. If you only listen to a recording on a bluetooth speaker, so common these days, you won't hear any of the stereo separation and likely none of the lower register but played back on a typical PC desk top system with a sub those things can be revealed but the fact that there are two cellists playing and not just one might only be noticed on a real high fidelity system. If you want to be able to hear the lower register and not just a one note base then you need something that can "accurately" or as I would prefer to say faithfully playback what's been recorded. So in my opinion building a system that can faithfully and objectively achieve an "accurate" playback is the right goal. Sometimes I maybe trying to make a set of nearfield monitors for mixing and they would need to be as flat as I could get them as a "reference" but other times I may be building a set of Hi Fidelity speakers that will exhibit a bump in the lower register and a roll off of the highs at around 18khz as that is how most music is structured and in my opinion sounds better because it is more dynamic that way. So the reference speakers are going to be accurate, better for mixing and recording and the Hi Fi speakers are going to be more faithful to the way the music is played and is better for playback. (IMHO)
But these things are only part of a great speaker system though. Just as important is the overall tonality of the speakers and this is a much harder thing to get right and is the subject of debate. It is the one area that is objective but no one has figured out how to measure it yet because it hasn't even been properly defined so until it has it is subjective and is the reason why that set of DIY speakers you have been dicking around with for the last year or so looks good on paper but sounds like crap. They may play back a couple songs all right but everything else sounds horrible. This comes down to driver selection and is compounded with the addition of multiple ways. Picking the right drivers comes down to a matter of experience or luck.
Anyways I don't want to stray to far off topic here so I'll leave it at that.
But these things are only part of a great speaker system though. Just as important is the overall tonality of the speakers and this is a much harder thing to get right and is the subject of debate. It is the one area that is objective but no one has figured out how to measure it yet because it hasn't even been properly defined so until it has it is subjective and is the reason why that set of DIY speakers you have been dicking around with for the last year or so looks good on paper but sounds like crap. They may play back a couple songs all right but everything else sounds horrible. This comes down to driver selection and is compounded with the addition of multiple ways. Picking the right drivers comes down to a matter of experience or luck.
Anyways I don't want to stray to far off topic here so I'll leave it at that.
Tonality, balance and other descriptions of loudspeakers are affected by the room and the acoustic design of the very thing, the physical object you have, the speaker. Drivers are just one part of the system and I think using suitable premium parts over just suitable parts gives diminishing returns. What is suitable driver? That is part of the construct that dictates suitable drivers. And the construct needs to be suitable to meet design criteria, some goal, which is most likely good sound at listening position in a room.
Size(s) of drivers and their relation to each other and to the structure and structure relation to room and listening position has far more effect how the sound is perceived than any individual parts on some random system. You'd have to do lots of work even with ideal drivers with flat 20-20kHz response that didn't have sound of their own, to make good sound at listening spot. As long as the drivers are suitable for what is required from them they should work pretty nicely and I don't see premium parts are necessity, just don't use the unsuitable ones. Fullrange and 2-way system people need to bend the physics to get performance that would need premium parts. But same premium parts on a system that doesn't have to bend physics would sound better.
It is just a hobby and you are right, there are many angles to look at and have fun or grief. My mission has lately been to convert people think more on their system and not their driver brands, as some might have noticed 😀 People often concentrate on technical details when they should concentrate how the whole system plays together, including the room, shape of the speaker as well as the drivers used and what not. The perceived sound depends on so many things and it all has to play nicely together to make good sound.
Size(s) of drivers and their relation to each other and to the structure and structure relation to room and listening position has far more effect how the sound is perceived than any individual parts on some random system. You'd have to do lots of work even with ideal drivers with flat 20-20kHz response that didn't have sound of their own, to make good sound at listening spot. As long as the drivers are suitable for what is required from them they should work pretty nicely and I don't see premium parts are necessity, just don't use the unsuitable ones. Fullrange and 2-way system people need to bend the physics to get performance that would need premium parts. But same premium parts on a system that doesn't have to bend physics would sound better.
It is just a hobby and you are right, there are many angles to look at and have fun or grief. My mission has lately been to convert people think more on their system and not their driver brands, as some might have noticed 😀 People often concentrate on technical details when they should concentrate how the whole system plays together, including the room, shape of the speaker as well as the drivers used and what not. The perceived sound depends on so many things and it all has to play nicely together to make good sound.
Last edited:
So I have my modded B652-Air and I have a single JBL Studio 270. The B652 has virtually no dispersion control, but the JBL uses an M2-type waveguide and in my limited measurements the dispersion is pretty consistent and wide across the spectrum.
In order to get the B652 to sound closer to the Studio 270 I have to lower the treble and boost the bass. By sound closer I mean, I can hear everything I want to hear in balance. The B652 sounds a bit boxy this way, but the details exist from low to high and blend musically with lifelike presence, in spite of the tilted response. And it is interesting to me that the sound can be convincing even though it is audibly not flat. It is easy to get used to the extra response which contributes to envelopment, but if you take away that response to make it flat, the realism goes away.
Before that is possible I have to get rid of peakiness in the mids and treble, like the 3.4KHz woofer peak I addressed with the crossover. Peaks seem to matter disproportionately compared to dips. Too little sound at one frequency is missed but too much is painful, and I prefer not to be in pain. One peak at 3-4KHz will dominate over everything else.
In order to get the B652 to sound closer to the Studio 270 I have to lower the treble and boost the bass. By sound closer I mean, I can hear everything I want to hear in balance. The B652 sounds a bit boxy this way, but the details exist from low to high and blend musically with lifelike presence, in spite of the tilted response. And it is interesting to me that the sound can be convincing even though it is audibly not flat. It is easy to get used to the extra response which contributes to envelopment, but if you take away that response to make it flat, the realism goes away.
Before that is possible I have to get rid of peakiness in the mids and treble, like the 3.4KHz woofer peak I addressed with the crossover. Peaks seem to matter disproportionately compared to dips. Too little sound at one frequency is missed but too much is painful, and I prefer not to be in pain. One peak at 3-4KHz will dominate over everything else.
This makes little sense to me. Most recordings are recorded so as to create the illusion the artist is trying for with a air of stereo loudspeakers. Plsy it over headphones, you get that unnatural sound inside your head. Ambiosonic recordings were made to work on headphones.
Everything is subjective, Even those who say something is “accurate” are leaning on the subjective desicion that the measures show what is heard. Given the poor state of the measure swe typically make a ton of places to go wrong.
In the end, no matter how good your hifi, if it connects you to the music and it brings pleasure then that is what really matters.
dave