Tweet
Boundary issues, acoustic setup, and the inverse square law impact on speaker design
I've been meaning to get this started for a while, but so busy with a variety of topics and projects that I just never get to it by the time the end of the day rolls around!
Well, all I'm going to do is post a few images as teasers for thought right now, with some brief comments, with the expectation that I'll get back to this soon, and of course, welcome other inputs... I think Sven may want to share some recent personal experience, too. And yes, this does relate to things in progress, like my four woofer tower concept the Saint-Säens that evolved into a FR30 inspired system using the same type of midrange. There's good reasons Chris Brunhaver used that partial line array configuration instead of the concepts employed by Paul McGowan and staff on previous efforts to present the B&G midranges and tweeters in the best light. Reasons I was learning in the 70's and which got reinforced with early multi-woofers experiments with projects like the Modula Xtreme, a precursor to the Isiris.
My simple point from decades of experience is that, like in psychology, boundaries matter, and one ignores the inverse square law at one's own risk...
So, a couple of pictures from web sites, one selling speakers, the other, room treatments, but showing a set of speakers that you might say, "piqued" my interest, and then one showing some recent testing. And to a substantial degree, I intend to send you off to read some other articles by pretty competent people, to help put the picture together. You do like reading, yes? Well, we welcome writing in your thoughts and experiences, too- (in this day and age, it seems that if it isn't in a video, 75% of folks can't be bothered...)(is this statistically related to functional literacy, I'm wondering?)
This image is from the landing page of the Dutch & Dutch 8c, a strikingly focused design on achieving consistent horizontal directivity above 100Hz, and judging by 3rd party measured results, quite successfully. Think about how this setup differs from say, the Cardas approach, and how the inverse square law might be relevant. I went through a lot of effort three decades ago to write a MathCAD program (DOS version, not windows!) that could analyze the impact of the three closest boundaries. Did some clever tricks to minimize the size, but they broke rules implemented in later versions of MathCAD- it will still run in the older ones.
See any potential issues here? Especially for the LF.
A directivity plot from a 3rd party review:
They did do their homework on that score. Pretty much on a par with some of Sigfreid Linkwitz's best work.
Here's the other pic I came across, for a place that makes room treatments, it just happens to feature some Fleetwood Sound speakers, and looks like it was taken in a hotel show demo room.
By now, you may be starting to figure out my concerns, based on that old inverse square law bugaboo... that woofer sure isn't very far off the floor. A lot of speakers these days follow a trend that folks sit in those low boy sectionals, and want to have the tweeter at ear level, often around 36", and that of course, the woofer should be below that. Hmmm... and they may need to follow that carefully, because the vertical lobing in the crossover region is significant enough that it really DOES need to be near ear level...
(One of the reasons I favor the not sanctioned concept of LR3 crossover, as well as the Duelund)
And while I do think the room treatments will level out some of the low bass, and likely improve the imaging characteristics in this room, I'm not really a fan of a single woofer system at a relatively low distance off the floor. At least, compared to the Dutch & Dutch pic, this one is a decent distance from the rear boundary. But still....
Working on the CC for Sven was a very real reminder of the problems and challenges that occur, and how this rather typical configuration near a floor boundary and wall boundary messes up the frequency response through the baffle step transition region and below that. In this case, one cannot remotely compare the quasi-anechoic elevated box measurement to what happens in the necessary system play back position for the center channel- some ugly things happen from the low base up through 1kHz due to boundary reflections and comb filtering (from wavelength dependent addition and subtraction).
Well, all I'm going to do is post a few images as teasers for thought right now, with some brief comments, with the expectation that I'll get back to this soon, and of course, welcome other inputs... I think Sven may want to share some recent personal experience, too. And yes, this does relate to things in progress, like my four woofer tower concept the Saint-Säens that evolved into a FR30 inspired system using the same type of midrange. There's good reasons Chris Brunhaver used that partial line array configuration instead of the concepts employed by Paul McGowan and staff on previous efforts to present the B&G midranges and tweeters in the best light. Reasons I was learning in the 70's and which got reinforced with early multi-woofers experiments with projects like the Modula Xtreme, a precursor to the Isiris.
My simple point from decades of experience is that, like in psychology, boundaries matter, and one ignores the inverse square law at one's own risk...
So, a couple of pictures from web sites, one selling speakers, the other, room treatments, but showing a set of speakers that you might say, "piqued" my interest, and then one showing some recent testing. And to a substantial degree, I intend to send you off to read some other articles by pretty competent people, to help put the picture together. You do like reading, yes? Well, we welcome writing in your thoughts and experiences, too- (in this day and age, it seems that if it isn't in a video, 75% of folks can't be bothered...)(is this statistically related to functional literacy, I'm wondering?)
This image is from the landing page of the Dutch & Dutch 8c, a strikingly focused design on achieving consistent horizontal directivity above 100Hz, and judging by 3rd party measured results, quite successfully. Think about how this setup differs from say, the Cardas approach, and how the inverse square law might be relevant. I went through a lot of effort three decades ago to write a MathCAD program (DOS version, not windows!) that could analyze the impact of the three closest boundaries. Did some clever tricks to minimize the size, but they broke rules implemented in later versions of MathCAD- it will still run in the older ones.
See any potential issues here? Especially for the LF.
A directivity plot from a 3rd party review:
They did do their homework on that score. Pretty much on a par with some of Sigfreid Linkwitz's best work.
Here's the other pic I came across, for a place that makes room treatments, it just happens to feature some Fleetwood Sound speakers, and looks like it was taken in a hotel show demo room.
By now, you may be starting to figure out my concerns, based on that old inverse square law bugaboo... that woofer sure isn't very far off the floor. A lot of speakers these days follow a trend that folks sit in those low boy sectionals, and want to have the tweeter at ear level, often around 36", and that of course, the woofer should be below that. Hmmm... and they may need to follow that carefully, because the vertical lobing in the crossover region is significant enough that it really DOES need to be near ear level...
(One of the reasons I favor the not sanctioned concept of LR3 crossover, as well as the Duelund)
And while I do think the room treatments will level out some of the low bass, and likely improve the imaging characteristics in this room, I'm not really a fan of a single woofer system at a relatively low distance off the floor. At least, compared to the Dutch & Dutch pic, this one is a decent distance from the rear boundary. But still....
Working on the CC for Sven was a very real reminder of the problems and challenges that occur, and how this rather typical configuration near a floor boundary and wall boundary messes up the frequency response through the baffle step transition region and below that. In this case, one cannot remotely compare the quasi-anechoic elevated box measurement to what happens in the necessary system play back position for the center channel- some ugly things happen from the low base up through 1kHz due to boundary reflections and comb filtering (from wavelength dependent addition and subtraction).
Comment