Crazy Austrian/German thread in hififorum.at :
Can Toole's preference criteria for a living room friendly loudspeaker be met
with this simple baffle and driver driver arrangement?
The proposed crossover frequency is below 500Hz:
Proposed driver arrangement for the mid and high frequency range:
https://www.hififorum.at/node/703371/page3#post703675
Simulated frequency response in the far field "on axis":
https://www.hififorum.at/node/703371#post703446
Placement of woofer and simulation of baffle step regarding woofer:
https://www.hififorum.at/node/703809?p=704449#post704449
Here is some discussion of a very similar driver configuration - using
only little smaller dimensions - regarding "off axis response",
"sound power" and "estimated inroom response" of said
midrange/highrange driver configuration.
Since i am the author of the posts linked and the driver configurations
proposed, i am here to help, if e.g. "DeepL" should not translate well ...
Can Toole's preference criteria for a living room friendly loudspeaker be met
with this simple baffle and driver driver arrangement?
The proposed crossover frequency is below 500Hz:
Proposed driver arrangement for the mid and high frequency range:
https://www.hififorum.at/node/703371/page3#post703675
Simulated frequency response in the far field "on axis":
https://www.hififorum.at/node/703371#post703446
Placement of woofer and simulation of baffle step regarding woofer:
https://www.hififorum.at/node/703809?p=704449#post704449
Here is some discussion of a very similar driver configuration - using
only little smaller dimensions - regarding "off axis response",
"sound power" and "estimated inroom response" of said
midrange/highrange driver configuration.
Since i am the author of the posts linked and the driver configurations
proposed, i am here to help, if e.g. "DeepL" should not translate well ...
Sorry, this link was missing (some "rough" off axis and sound power estimations
done in Excel based on "The Edge" baffle diffraction simulator) :
https://www.hififorum.at/node/703371?p=703655#post703655
done in Excel based on "The Edge" baffle diffraction simulator) :
https://www.hififorum.at/node/703371?p=703655#post703655
@TNT i will try to reduce the goals to some "bullet points":
The cabinet is assumed to be rather flat (considerably less than 20cm and/or the width) and "toed in" towards the listener.
Even a slight "tilt" to aim the Midrange/Highrange Section at the listener will be necessary.
So this is - as i think of it - a highly pragmatic attempt to make a "living room friendly loudspeaker".
- mitigating (if not completely neutralizing) baffle diffraction artefacts in the "on axis" frequency response (by that keeping crossover design relatively simple)
- create uncorrelated sound above about 300 ... 600Hz under larger angles horizontally and vertically
- (by doing so) mitigate comb filtering (strong interference needs coherence) with mirror sources in the room (e.g. from side walls)
- aim for directivity index continouusly and smoothly rising with frequency (no "blooming" of directivity at/above crossover frequency)
- aim for inroom response falling smoothly with frequency similar to Bruel&Kjaer "preferred" curve (1974)
- meet Toole's preference criteria for loudspeakers designed for "usual" living rooms, that lack extensive acoustic (esp. absorbing) treatment
- make the speaker work, were most home listeners - at least in Europe AFAIK - will place them anyhow: that is rather close to the front wall of the room and with not very much distance to the side walls (if unlucky).
- no crossover frequency in the most sensitive frequency range of human hearing
The cabinet is assumed to be rather flat (considerably less than 20cm and/or the width) and "toed in" towards the listener.
Even a slight "tilt" to aim the Midrange/Highrange Section at the listener will be necessary.
So this is - as i think of it - a highly pragmatic attempt to make a "living room friendly loudspeaker".
Off axis responses were calculated only for selected angles so far (and summed up):
https://www.hififorum.at/node/703371?p=703655#post703655
Polar plots for certain frequencies will have a main lobe but also side lobes:
They would look like a "cross breed" between a conventional "mostly coherent" radiating
speaker and a DML (Distributed Mode Loudpspeaker) or even a diffuser.
One could even say, that (single frequency) polar plots will "look like ****":
But that "****" is statistically very well behaved, so that power response and DI change
very smoothly with frequency in the mid/high range.
You need only few points in space (say a dozen or more) to simulate or measure to
confirm that claim quickly.
I am sorry: I forgot we are not allowed to write "TIHS" here ...
(Edit: Which in itself is a good thing ...)
(Edit: Which in itself is a good thing ...)
Last edited:
Instead of using multiple, uncorrelated sources all over the place (both on the speaker and the walls as you mentioned) why not just build something a la the Bose 901?
Which of the goals mentioned above
- execpt for the last one of course - could be achieved with the BOSE 901 concept?
- execpt for the last one of course - could be achieved with the BOSE 901 concept?
Last edited:
Nice project but what you are seeking to do seems to be a challenge. I am not aware of any wholly successful implementations only partially successful ones like Sonab.
3 sources on a baffle with a few weaker images seems insufficient for a reasonably decorrelated reflected sound and is likely to cause some interference issues for the direct sound. My guess is that making use of the directivity of the source to weaken the direct sound and strengthen the image sources is likely to be more successful but I don't know the rules for what works reasonably. A while ago I tried to get a handle on them via web publications about the Carlsson "orthoacoutic" approach but most of it seemed to be audiophile marketing rather than engineering. Academic publications are likely to be a more reliable source but also a significantly larger task which I didn't pickup. Good luck in making some progress.
3 sources on a baffle with a few weaker images seems insufficient for a reasonably decorrelated reflected sound and is likely to cause some interference issues for the direct sound. My guess is that making use of the directivity of the source to weaken the direct sound and strengthen the image sources is likely to be more successful but I don't know the rules for what works reasonably. A while ago I tried to get a handle on them via web publications about the Carlsson "orthoacoutic" approach but most of it seemed to be audiophile marketing rather than engineering. Academic publications are likely to be a more reliable source but also a significantly larger task which I didn't pickup. Good luck in making some progress.
Regarding Bose "901" (above):
In fact my proposed concept does quite the opposite:
While my proposal reduces detrimental effects especially from early reflections on the direct
sound , the "901" deliberatelely makes use of reflections from the front and side walls of the
listening room, which are highly correlated to the direct sound. The "901" is known for causing
comb filter effetcs by doing so ...
I am not saying a "901" is (therefore) condemned to sound bad in every room or with every
placement, but the goal ist quite the opposite also in indirect sound being technically
dominant with the "901": The main lobes of radiation are not directed at the listener.
With my concept this is the case instead.
In fact my proposed concept does quite the opposite:
While my proposal reduces detrimental effects especially from early reflections on the direct
sound , the "901" deliberatelely makes use of reflections from the front and side walls of the
listening room, which are highly correlated to the direct sound. The "901" is known for causing
comb filter effetcs by doing so ...
I am not saying a "901" is (therefore) condemned to sound bad in every room or with every
placement, but the goal ist quite the opposite also in indirect sound being technically
dominant with the "901": The main lobes of radiation are not directed at the listener.
With my concept this is the case instead.
Last edited:
The first front wall reflection from e.g. a 901 or similarly a dipole loudspeaker will be correlated to a degree, but not totally. Subsequent secondary reflection will be uncorrelated above 600Hz per your goal. So I think the 901 is not all that much different. But anyway,it was just a suggestion meant to stimulate some thought/conversation. Personally, I have concerns that multiple, uncorrelated sources above 600Hz will destroy any sort of predictable imaging, etc. and will just provide "sound". It's hardly the same sort of goals as hi-fi stereo reproduction, but perhaps your goals are totally different and that is fine. If you do build this system I would love to know your thoughts about things like soundstage, properly locating instruments, etc.
As long as there is a baffle, there will be diffraction. The position of the drivers on the baffle or the shape of the baffle will not change this fact. The only way to mitigate baffle diffraction is to use drivers that tend to "beam" forward, thereby reducing the sound intensity to the sides, where the baffle edges are. Doing this may be practical with tweeters (waveguides!), but not so much with low/mid drivers.
Not sure what "uncorrelated sound" is in this context. Sounds a bit like random noise to me, but I guess that's not what you have in mind.
Again, waveguides will be your friends.
Just wanted to say that waveguides might be worth considering, but I guess I mentioned that already ;-)
Again, waveguides... they help with emphasizing direct sound from the speaker over reflected sound from the room at the listener position. So, yes, waveguides again!
Regarding "issues" with direct - on axis? - sound:
Who has designed and/or built conventional speaker configurations can tell easily "which curve is which" i guess ...
Who has designed and/or built conventional speaker configurations can tell easily "which curve is which" i guess ...
Last edited:
This design is all wrong - three identical wide/fullrangers widely spaced sharing the same frequency band will introduce very bad off-axis behavior, evident from your own simulation (https://www.hififorum.at/node/703371?p=703655#post703655). Flat on-axis frequency response (including baffle edge diffraction) is not enough for good sound.
Power curve is OK-ish, but it is not the whole story.
Last edited:
Imagine e.g. a "central located" phantom source:
So stereo (too) is "all wrong"(?): Two identical (fullrange) Loudspeakers widely spaced sharing the same frequency band will produce very bad off-axis behaviour ...
So stereo (too) is "all wrong"(?): Two identical (fullrange) Loudspeakers widely spaced sharing the same frequency band will produce very bad off-axis behaviour ...
Last edited:
So if you have:
- on axis FR "OK-ish" and
- sound power also "OK-ish"
what is the whole story (missing) then?
Strictly similar looking response under angles (e.g. 15, 30, 45 degrees) like in conventional (strictly coherently radiating?) designs?
Who is sitting there listening stereo(?) and enjoying(?) those "uniform" responses (in the direct sound) under angles?
There is a simple answer to that: No one is sitting there, and if so they are not listening in stereo anyhow,
regardless of the type of speaker they use.
Last edited:
You are misinterpreting stereo "central located phantom source" with your mono three sources located on one of your loudspeakers (left or right). It is not the same! Try to get stereo "central located phantom source" with your two (left and right) loudspeakers - you will get a mess!
I didn't say "on axis FR "OK-ish" - on the contrary, I said: "Flat on-axis frequency response (including baffle edge diffraction) is not enough for good sound.". If that was not clear enough, I will reformulated that like this: "You manage to get exemplary flat on-axis frequency response on-axis, but that is not enough for good sound".
As I said, the main problem with your design is bad off-axis behavior, which is inevitable with three identical drivers sharing the same frequency band.
I didn't say "on axis FR "OK-ish" - on the contrary, I said: "Flat on-axis frequency response (including baffle edge diffraction) is not enough for good sound.". If that was not clear enough, I will reformulated that like this: "You manage to get exemplary flat on-axis frequency response on-axis, but that is not enough for good sound".
As I said, the main problem with your design is bad off-axis behavior, which is inevitable with three identical drivers sharing the same frequency band.
Last edited:
Necessity of good off-axis behavior is well-established science fact. Your denying of that fact is your loss.