ArtsyAllen,
thank you for the compliment on the design, they were exceptional sounding. I still have a second set of enclosures but to fill them with the same components is just so expensive. The base was made with stacked 1" mdf and painted with two component automotive black urethane paint.
ra7,
I am not arguing with you and you actually just said what I was saying, and Wesayso it seems is in agreement, about 10Khz with a 3" cone you run out of clean top end. That was my only point, that no single device I know of can really produce the entire audio bandwidth. Now that being the case what happens about 10Khz really isn't all that much or that important to the overall sound. I am not arguing that your line arrays don't sound wonderful. I just didn't want someone to think they are perfect, there is always a compromise somewhere in all of our designs, we can't beat physics.
Wesayso,
Keep enjoying your speakers, I am all for that. Besides most of us can't hear much above 10Khz anyway anymore if we aren't teenagers and not much information is up there anyway. When I get my speakers completed that I am working on I could send you one of the Be dome tweeters so you can try your experiment. But why would you want to mount the tweeter on the top of your column and not on the side on your listening axis? If you were only adding a single unit, that would be distracting to have that sound off axis I would think?
ps. Last I checked a couple months ago I could clearly hear up to 16khz, that seemed to be my audible acuity limit.
thank you for the compliment on the design, they were exceptional sounding. I still have a second set of enclosures but to fill them with the same components is just so expensive. The base was made with stacked 1" mdf and painted with two component automotive black urethane paint.
ra7,
I am not arguing with you and you actually just said what I was saying, and Wesayso it seems is in agreement, about 10Khz with a 3" cone you run out of clean top end. That was my only point, that no single device I know of can really produce the entire audio bandwidth. Now that being the case what happens about 10Khz really isn't all that much or that important to the overall sound. I am not arguing that your line arrays don't sound wonderful. I just didn't want someone to think they are perfect, there is always a compromise somewhere in all of our designs, we can't beat physics.
Wesayso,
Keep enjoying your speakers, I am all for that. Besides most of us can't hear much above 10Khz anyway anymore if we aren't teenagers and not much information is up there anyway. When I get my speakers completed that I am working on I could send you one of the Be dome tweeters so you can try your experiment. But why would you want to mount the tweeter on the top of your column and not on the side on your listening axis? If you were only adding a single unit, that would be distracting to have that sound off axis I would think?
ps. Last I checked a couple months ago I could clearly hear up to 16khz, that seemed to be my audible acuity limit.
Last edited:
Kindhornman,
I would love to check out those Be dome tweeters. From what you've told us about them so far they seem a pretty good candidate for such a job.
I hadn't really checked out the details but here's the idea that floated in my mind when I said that:
The reason for me to put a tweeter on top would be differences in distance between the array as a line source and the separate point source tweeter. But it would probably only work better getting (much) closer to the array. Here's a quick sketch:
As can be seen, the difference is there in length, but seeing it, I'm not sure it's worth the trouble.
If we figure the line source drops 3 dB per doubling distance and the tweeter would drop 6 dB per doubling distance I'd want to prevent the tweeter getting much louder with slight forward movement and change too much with side to side movement. While there still is an advantage (in distance), it's not a really big advantage.
As we generally have a hard time determining height the idea was to put them on top and aim them at the preferred listening spot (angled down). But seeing the actual change in distance in the sketch I just made might make one wonder if it would be an audible advantage.
With sounds above 10 KHz I would be less worried over small timing changes and suspect we are way more sensitive to SPL level changes. The timing can be adjusted to the listening spot, but that would still change with head movement.
But so would the average SPL level from the point source tweeter vs the line source arrays.
About height cues, when you place tweeters low (on the door card) in a Car, a boost at ~8 KHz can make the sound seem like it's coming from way higher than the tweeter placement would suggest. So the level is more important to our perception than the actual placement/direction.
Last edited:
Wesayso,
Yes the distance difference would be a problem with the tweeter on top, but you are so smart you could have a variable delay. On axis same thing you could have a variable output with distance.
Don't get me started with the stupid position of speakers in cars, they drive me nuts, I just think of it as background sound and leave it at that.
Yes the distance difference would be a problem with the tweeter on top, but you are so smart you could have a variable delay. On axis same thing you could have a variable output with distance.
Don't get me started with the stupid position of speakers in cars, they drive me nuts, I just think of it as background sound and leave it at that.
Defining the unknown greatness I seek...
If I recall, Derek went into great detail promoting his propitiatory BMR that is not available for experimental DIY projects.
As You know, I am leaning towards using cone drivers for my next array build, but I am planning to compare the 3" Tectonic up against those, mainly to see if the wider dispersion characteristic of the highs is something desirable. The Tectonic does "appear" to have a quality disadvantage compared to the other drivers, (esp the bezel design) although it is priced higher then the Vifa/Tymphany TC9, so maybe it offers more then we are led to believe. As I will not be able to hear Derek's BMR any time soon, that may play in my favor, because I may actually like the Tectonic, not having my impression tainted by something I can not have anyway. It is certainly worth a listen, even if just for the learning experience.
Also, to dig in a little here, we have been trying to figure out the mysteries of some of the subjective aspects of sound: i.e. soundstage, width, depth, space, etc. Wesayso has went thru greats lengths in this thread and has backed up his discoveries with objective measurements.
As it stands, many of us who have full-range arrays are preferring the gradual high frequency roll-off as it seems to enhance the ambiance, size and space of a 3D soundstage. With my current arrays, I prefer this, because when I EQ the highs level with the mid-range, I get a very "in your face" direct "spaciously flat" sound. It becomes more apparent that the sound is coming from a left and right speaker. The speakers "appear" instead of "disappear". The question is: Are they doing that, because the highs are "directional" due to the dispersion characteristics of the drivers, or is it that our minds prefer a gradually rolled off high? This is my interest in the BMR (if they do indeed disperse better in the highs). Will better dispersion of the highs have the same effect?
Also, I have not ruled out adding a line of tweeters, now knowing the possibilities of DSP. But, cross-over frequencies must still be aligned with the actual distances between tweeter and mid-range lines in an array. Due to the big bezels on tweeters (including ribbon and planar), it is hard to get a line of tweeters close enough to a line of mid-ranges if one wants to keep the cross-over above 3000 Hz (outside our sensitive range). And try to find a tweeter that works good in a line array (minimal space between the tweeters elements themselves). I could build a 8 foot tall ribbon tweeter, yet due to the powerful magnets, it would still require a big bezel, flanking them a good inch or more away from the bezel of the mid-range line. Is it worth the trouble? All great things are, I have not figured yet if this is a part of the greatness I seek.
A lot of times, less is more!
My hearing is only good to about 14 kHz, so I do not know how important all of this is. The inclusion of the BMR in my driver comparisons is to see how important the highs actually are in my set of given compromises.
The concern I have, is if I do like the dispersion of the BMR, yet find the quality of the highs they produce unsatisfactory, I may need to go down the same path as Derek, and develop a better BMR. An intimidating journey to say the least, but maybe something worth fighting for.
Allen
If I recall, Derek went into great detail promoting his propitiatory BMR that is not available for experimental DIY projects.
As You know, I am leaning towards using cone drivers for my next array build, but I am planning to compare the 3" Tectonic up against those, mainly to see if the wider dispersion characteristic of the highs is something desirable. The Tectonic does "appear" to have a quality disadvantage compared to the other drivers, (esp the bezel design) although it is priced higher then the Vifa/Tymphany TC9, so maybe it offers more then we are led to believe. As I will not be able to hear Derek's BMR any time soon, that may play in my favor, because I may actually like the Tectonic, not having my impression tainted by something I can not have anyway. It is certainly worth a listen, even if just for the learning experience.
Also, to dig in a little here, we have been trying to figure out the mysteries of some of the subjective aspects of sound: i.e. soundstage, width, depth, space, etc. Wesayso has went thru greats lengths in this thread and has backed up his discoveries with objective measurements.
As it stands, many of us who have full-range arrays are preferring the gradual high frequency roll-off as it seems to enhance the ambiance, size and space of a 3D soundstage. With my current arrays, I prefer this, because when I EQ the highs level with the mid-range, I get a very "in your face" direct "spaciously flat" sound. It becomes more apparent that the sound is coming from a left and right speaker. The speakers "appear" instead of "disappear". The question is: Are they doing that, because the highs are "directional" due to the dispersion characteristics of the drivers, or is it that our minds prefer a gradually rolled off high? This is my interest in the BMR (if they do indeed disperse better in the highs). Will better dispersion of the highs have the same effect?
Also, I have not ruled out adding a line of tweeters, now knowing the possibilities of DSP. But, cross-over frequencies must still be aligned with the actual distances between tweeter and mid-range lines in an array. Due to the big bezels on tweeters (including ribbon and planar), it is hard to get a line of tweeters close enough to a line of mid-ranges if one wants to keep the cross-over above 3000 Hz (outside our sensitive range). And try to find a tweeter that works good in a line array (minimal space between the tweeters elements themselves). I could build a 8 foot tall ribbon tweeter, yet due to the powerful magnets, it would still require a big bezel, flanking them a good inch or more away from the bezel of the mid-range line. Is it worth the trouble? All great things are, I have not figured yet if this is a part of the greatness I seek.
A lot of times, less is more!
My hearing is only good to about 14 kHz, so I do not know how important all of this is. The inclusion of the BMR in my driver comparisons is to see how important the highs actually are in my set of given compromises.
The concern I have, is if I do like the dispersion of the BMR, yet find the quality of the highs they produce unsatisfactory, I may need to go down the same path as Derek, and develop a better BMR.
An intimidating journey to say the least, but maybe something worth fighting for. Allen
Allen, I certainly don't regret "keeping it simple", but one might wonder how simple it really is with all of the DSP that I included(*).
The drop off I have did not differ that much (or not at all) from what the research by JBL predicted to do well in a real room. I'm definitely not worried I'm missing something.
If I had to do it all over again I'm not so sure I would change anything, yes, I'm that pleased.
(*) Yet the DSP is different from what people want to believe. It's definitely not "hammering the frequency into shape", it's finding the least amount I could get away with that sounded the best to these ears. And even the graph's agree...
The drop off I have did not differ that much (or not at all) from what the research by JBL predicted to do well in a real room. I'm definitely not worried I'm missing something
.If I had to do it all over again I'm not so sure I would change anything, yes, I'm that pleased.
(*) Yet the DSP is different from what people want to believe. It's definitely not "hammering the frequency into shape", it's finding the least amount I could get away with that sounded the best to these ears. And even the graph's agree...
Last edited:
ArtsyAllen,
Just my knowledge not based on the BMR driver your speaking of but some alternate work I did working with someone on a flat panel speaker is that as the frequency rises the so called flat diaphragm no longer acts as a planar surface. The response will tend to narrow around the drive point, the attachment point on the diaphragm. What I am saying is just as in a cone driver the high frequency dispersion narrows and the diaphragm goes into bending modes, very different than the marketing that says they act as infinitely stiff surfaces. The are definite Eigen modes happening in those diaphragm, if you could look at the diaphragm with a laser it would become apparent how there are many different breakup modes dependent on frequency.
Your comment on the width of most dome tweeters is very true generally but you could look to some of the very small dome tweeters intended for automotive use that use Neodymium magnets in a cup core design, Those can be almost as small in diameter as the dome itself. Now who is making a great sounding unit I can't tell you but they are much smaller than a typical tweeter using a flat ceramic magnet with a front plate and rear T-pole type construction.
Just my knowledge not based on the BMR driver your speaking of but some alternate work I did working with someone on a flat panel speaker is that as the frequency rises the so called flat diaphragm no longer acts as a planar surface. The response will tend to narrow around the drive point, the attachment point on the diaphragm. What I am saying is just as in a cone driver the high frequency dispersion narrows and the diaphragm goes into bending modes, very different than the marketing that says they act as infinitely stiff surfaces. The are definite Eigen modes happening in those diaphragm, if you could look at the diaphragm with a laser it would become apparent how there are many different breakup modes dependent on frequency.
Your comment on the width of most dome tweeters is very true generally but you could look to some of the very small dome tweeters intended for automotive use that use Neodymium magnets in a cup core design, Those can be almost as small in diameter as the dome itself. Now who is making a great sounding unit I can't tell you but they are much smaller than a typical tweeter using a flat ceramic magnet with a front plate and rear T-pole type construction.
Wesayso,
It seems we have been told for so long that the frequency response is supposed to be flat for so long that it has become the norm that so many strive for. On the other hand it seems from actual DBLT that most people prefer a slowly downward sloping curve. Given a perfectly flat output up to the highest frequencies the testing showed that the preference was to cut the highest frequencies rather than have the ruler flat output up to 20Khz. When you think about the normal drop off of spl with rising frequencies with distance this seems more what we hear in nature, not a flat power response across the entire range.
It seems we have been told for so long that the frequency response is supposed to be flat for so long that it has become the norm that so many strive for. On the other hand it seems from actual DBLT that most people prefer a slowly downward sloping curve. Given a perfectly flat output up to the highest frequencies the testing showed that the preference was to cut the highest frequencies rather than have the ruler flat output up to 20Khz. When you think about the normal drop off of spl with rising frequencies with distance this seems more what we hear in nature, not a flat power response across the entire range.
But we have to be careful here. The slowly dropping response is an in-room response, or a power response. Yes, we certainly don't want flat power, too bright. But we don't want a falling on-axis response either. That to me has sounded too dull. What we ideally want is flat on-axis but a falling power response. And maybe it's a coincidence, but traditional cone and dome designs do exactly this -- Combined with typical room absorption, you get a falling room response. The constant directivity designs on the other hand need specific room absorption to achieve this optimum. I was never able to get this balance right with the constant directivity horn designs that I tried. It was either bright or dull, i.e., never flat on-axis but falling power.
ra7,
Yes Earl Geddes uses massive room treatment to make his CD horns work in room. He hates exponential horns but I never have had a problem with them though I usually designed horns as hyperbolic function not a large difference as far as I'm concerned if you don't do it excessively. I used a 0.6 exponent if I recall correctly.
Yes Earl Geddes uses massive room treatment to make his CD horns work in room. He hates exponential horns but I never have had a problem with them though I usually designed horns as hyperbolic function not a large difference as far as I'm concerned if you don't do it excessively. I used a 0.6 exponent if I recall correctly.
That's why I still have an interest in the JMLC horn, as it mimic's a small cone driver in it's dispersion pattern. Couple a pair of those to the Synergy ideas that are booming news lately and I think one might be able to get very nice results.
Vifa tweeters like these come to mind:
Attachments
Last edited:
Interesting, what are the plans... run as dipole for the upper part or are you planning enclosures (maybe we don't see them and you already have corner fitting enclosures)?
Best explanation of line arrays I have come across so far. Scroll down to the section on
Line Array or Column Loudspeakers
Line Array or Column Loudspeakers
Yes they are corner fitting enclosures basically big Toblerone boxes [emoji2]
Sent from my iPhone using Tapatalk
Lol...
That description made it clear in an instant. Probably very similar to what ra7 is using...
Did it help put in perspective why one would use line arrays in a home environment? It misses some of the key points for me about a floor to ceiling array. I've been asked many times if a shorter array could do the same. Not in my opinion. The key here is the "seemingly endless array" we create by using the reflections, that every speaker has to deal with one way or another, to our advantage. Combined with the speed and low distortion of using that many drivers which results in each driver only having do a very small part of that total.
Attachments
Wesayso, been lurking in your thread for a few weeks while I've been planning a build to get back into speaker building, I'm not going to say anything that you haven't already heard, but epic build and even more epic effort! I came across your thread initially when I was researching laminated building techniques.
I am taking on a much simpler project using longitudinal layers, nothing too special but it's been fun to figure it out with limited "hand" tools and experiment with the technique. I'm documenting my build and system in this thread if you get a chance to check it out http://www.diyaudio.com/forums/mult...ayton-reference-adcom-system.html#post4609593
Anyways, thanks again for inspiration and not giving up so easily, great to meet you
I am taking on a much simpler project using longitudinal layers, nothing too special but it's been fun to figure it out with limited "hand" tools and experiment with the technique. I'm documenting my build and system in this thread if you get a chance to check it out http://www.diyaudio.com/forums/mult...ayton-reference-adcom-system.html#post4609593
Anyways, thanks again for inspiration and not giving up so easily, great to meet you
Wesayso
Found this room kit calculator - curious what the suggestion would be for your room.
Universal Acoustics Room Kit Calculator
A you tube demo of room absorption panels was quite striking, with a noticeable cleaning up of sound.
Found this room kit calculator - curious what the suggestion would be for your room.
Universal Acoustics Room Kit Calculator
A you tube demo of room absorption panels was quite striking, with a noticeable cleaning up of sound.
Wesayso, been lurking in your thread for a few weeks while I've been planning a build to get back into speaker building, I'm not going to say anything that you haven't already heard, but epic build and even more epic effort! I came across your thread initially when I was researching laminated building techniques.
I am taking on a much simpler project using longitudinal layers, nothing too special but it's been fun to figure it out with limited "hand" tools and experiment with the technique. I'm documenting my build and system in this thread if you get a chance to check it out http://www.diyaudio.com/forums/mult...ayton-reference-adcom-system.html#post4609593
Anyways, thanks again for inspiration and not giving up so easily, great to meet you
Thanks for the kind words... I will check out your build!
My room doesn't seem to fit into the calculator
. But don't count too much on standard calculators. They show you where to expect troubles but room furniture, speaker placement and listening position all change how we perceive that. Room treatment is important though! DSP is not a real substitute but can help to get better results.I went another way with this. I looked at first reflection points and damped those first. That meant using 2 huge panels, hidden from view and one poster/damping panel in clear sight.
Remember that I made a promise not to add more damping panels? In the bass I swap some energy at low frequencies between both arrays. The left is a little louder at 30 Hz as it does better there due to being in a corner. But that same tower has more problems around 60 Hz. So that's where the right tower is up a little compared to the left one. As the wave lengths are longer than the room dimensions we cannot fully hear the direction where the low notes are coming from. The left still plays those notes at 60 Hz, just a few dB's lower. Together that evens out the most problematic room modes.
Looking at measurements can tell you what is really happening in the room and at what frequency. I tried to get that as even as I could which resulted in this graph:
So it's pretty even and consistent trough most of the spectrum. It still is a lively room though. Due to using tricks like back reflections it "sounds" bigger than it is. Most important is to get the early reflections out of the way:
The first peak is the music signal, the next ~ 20 ms is down by 20 dB or more and at 20 ms you see the "back reflections" I introduce and a gradual decay after that (as far as one can even speak of decay in a small room).
Look at: Room Acoustics for some more food for thought.
I actually spend some time last week to work on the back reflections. As they are created by my ambient speakers I can make them do whatever I want. There's way more "power" in there than I expected. I had a lot of fun playing around with it. Adjusting the timing and the content they play...
Mind tricks
, but they do work! At no point you are aware there are speakers in the back(*). But they do change your perception of what's happening in front of you. Together it makes for a more believable stereo image that's larger than my room when the music asks for it. Enhancing the "you are there" feeling of this setup.(*) those speakers to not fire directly at the listening position. Most of it is reflected energy you hear, diffused and de-correlated from the main signal.
Last edited: