If the room is completely dead then it can't make any difference, right? All you will get is direct sound no matter what the DI is. As the room becomes less dead then you get early reflections and degraded imaging. Really, there is no situation where I see low DI speakers working well where image is concerned. Sure, in a more lively room they will cover up a lot of recording problems of large venues where spaciousness is the goal and image doesn't matter, but that's a compromise that you have to make based on what you are looking for. To me, image is paramount and I won't compromise on it.
No room is completely dead though. But lets say a smallish lively room where you are 6ft from the speakers. That's a case where more dispersion might be good if you can avoid early reflections.
How can you possibly avoid early reflections if the room is "lively" and small and the speakers send sound all over the place.
No, "a smallish lively room" is exactly what I have and my speakers have extremely high DI, and the room is often rated as "the best that I have heard." Listening distance is maybe ten feet.
No, "a smallish lively room" is exactly what I have and my speakers have extremely high DI, and the room is often rated as "the best that I have heard." Listening distance is maybe ten feet.
I think someone could get the speakers more towards to the middle of the room on stands. But I still think that too much direct sound is fatiguing, and the ratio of 2 to 1 indirect to direct is about right. I don't think there are a lot of good options for small rooms. A deader room would benefit from dispersion.
Also just to add a point, a sound absorbing baffle would help cut down early reflections.
Also just to add a point, a sound absorbing baffle would help cut down early reflections.
Last edited:
My experience of small rooms (which, lets face, we are stuck with) over some 45 years says that high DI and low absorption is the key. Of course the room needs to be well damped in the modal region, but that's possible to do - just read my books.
Ah, as Earl and 33Polkhigh are pointing out, it really may be that "colour management" really matters.
If you have directional horns like Earl's speakers, you whomp-up the reflections to add warmth, as he says. But for minimally directional typical speakers in sparsely furnished rooms (typical of DIYaudio pictures) the sound may be hideously lively if not absorbed wisely.
Tinkering with diffraction seems a very minor tweak in the larger adjustment of colour.
B.
Diffraction is mostly an imaging problem as it does NOT have much effect on colouration. Room reflections do, for more steady state signals, but its the transients that affect imaging and diffraction interferes with the transients a lot.
Last edited:
One can have both ambience and imaging if the speaker avoids early reflections and narrows dispersion. I think this is partly why the bass unit with smaller mid range and tweeter sections on top is a common hi end design.
I'm afraid that you have that backwards. Smaller speakers have wider directivity. So to achieve a higher DI (for lower early reflections,) one would want to use larger drivers, not smaller.
The lengths some will go to...
Here is an amusing solution to diffraction and similar issues from our cousin web site audio asylum. This is, er, a padded cell so to speak 😀
The ole felt tweak - emailtim - Tweakers' Asylum
Here is an amusing solution to diffraction and similar issues from our cousin web site audio asylum. This is, er, a padded cell so to speak 😀
The ole felt tweak - emailtim - Tweakers' Asylum
Buchardt Audio says that their deep deep waveguide reduces diffractions by 90% and thus they do not use baffle with a rounded edges.
You can check detailed measurements on their web page.
s400 detailed description — Buchardt Audio
You can check detailed measurements on their web page.
s400 detailed description — Buchardt Audio
Last edited:
Which leave us with 10% remaining and bring to Earl Geddes's answer to the issue: to have both waveguide AND rounded edges (of significant radius for the freq range involved ... and if you decide this is the kind of 'loudspeaker type' you like). 😉
Last edited:
If remaining 10% could be translated as -20dB then it might not be hearable, but I'm not sure how they meant it 🙂
Not sure. If Earl is right in his work it'll depend of the overall spl (if it is noticable or not... as the louder the most noticable it should be from his research).
The point about what they mean (and how i understand it) is that once the waveguide effect kick in the directivity control make the high freq diffraction less an issue as the freq radiate in a -6db 90° conical shape though the edge diffraction is de facto less of an issue (in their way to approach things), second being lower in level than the direct wavefront they should be hidden 'behind' the direct signal.
But are we sure that what remains is really hidden behind (level is low enough versus the controlled radiated sound) and it doesn't either gives any clues the delay induced won't harm the direct signal? And what happen to frequency just below the waveguide where the direct radiator is slowly beaming to reach waveguide radiation pattern? If not treated the effect is still present (as this is not a binary effect (on/off): there is a transition area)?
In doubt i think Earl's answer provide better control over both issues and should give a much better end results.
The point about what they mean (and how i understand it) is that once the waveguide effect kick in the directivity control make the high freq diffraction less an issue as the freq radiate in a -6db 90° conical shape though the edge diffraction is de facto less of an issue (in their way to approach things), second being lower in level than the direct wavefront they should be hidden 'behind' the direct signal.
But are we sure that what remains is really hidden behind (level is low enough versus the controlled radiated sound) and it doesn't either gives any clues the delay induced won't harm the direct signal? And what happen to frequency just below the waveguide where the direct radiator is slowly beaming to reach waveguide radiation pattern? If not treated the effect is still present (as this is not a binary effect (on/off): there is a transition area)?
In doubt i think Earl's answer provide better control over both issues and should give a much better end results.
Last edited:
'All over' would surely be as complex an issue as the other way around wouldn't you say?
^No, it is just the stand that has bevel.
Shallow waveguides lay on a wider baffle, which alleviates diffractions. A synergy horn or similar wide and deep horn is more critical, but they can be highpassed above the problematic frequency.
I have measuend Fountek NeoCD 3.5H without a baffle and it works quite well too, up/down to certain F.
An externally hosted image should be here but it was not working when we last tested it.
Shallow waveguides lay on a wider baffle, which alleviates diffractions. A synergy horn or similar wide and deep horn is more critical, but they can be highpassed above the problematic frequency.
I have measuend Fountek NeoCD 3.5H without a baffle and it works quite well too, up/down to certain F.
Attachments
Last edited:
What do you think about Siegfried Linkwitz design approach, which was probably different from yours
The-Magic-in-2-Channel-Sound
http://www.linkwitzlab.com/The_Magic/The Magic in-Edit-SL.pdf
Dipole speakers, wide radiation pattern, both front and rear radiation.