What is the ideal directivity pattern for stereo speakers?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
No takers on building to this concept, eh?

It answers what has been forward as an ideal in this thread: You can have a lively room and speakers of any directivity, including wide directivity, while still achieving a dead period in terms of early reflections.

The author gets lower early energy than you could achieve with diffusors. Probably lower than you could get with especially directional speakers. You could also use the concept with dipoles to good advantage. Since they send little energy to the sides where it would contribute to spaciousness, and much to the rear wall where it can cause problems, this approach would let you redirect the energy in a more beneficial direction.

You also don't need as complex a surface as this. If you have a well defined seating position you can simply put large reflectors at the wall bounce points of the reflections you want to "divert". Plus diffusors at points where you want reflections from (wide sidewall angles).

You could certainly experiment with free standing plywood reflectors, or even foam core.

David S.
I was trying to combine it with some CD racks in a basement. Basically still try to do some absorption of reflected waves after they have traveled to the back of the room while still applying a similar concept to reduce reflections from the back wall reflecting back to the listening area.
 
diyAudio Moderator
Joined 2008
Paid Member
No takers on building to this concept, eh?
I was doing some tracing yesterday and as you can see from the attached image, by doing little more than extending the flat baffle out until it touches the side and rear walls I could be redirecting the rear and both sides.

My ceiling diffusor is basically along these same lines and I can tell you that it works,
You have the plane of your reflector sheets pointed back at the speakers, ie. they are not broadside toward the speakers, is this correct?

I would have assumed the opposite, but then that would produce a Freznel style reflector (with frequency limitations) and I guess yours wouldn't be acting that way?
 

Attachments

  • 1.gif
    1.gif
    5 KB · Views: 227
Administrator
Joined 2004
Paid Member
No takers on building to this concept, eh?
Don't be in such a hurry! I studied the paper while traveling, I need to look again.

While I do like the concept - and my room is of a very similar size - but I ain't sitting anywhere near that close. So I'd have to figure out how and if it would work for my listening position in my room. I do not want to sit that close to the speaker wall. Would you?
 
REW shows ETC and is a freeware. It has all the features one need for small room acoustics.
REW - Room EQ Wizard Home Page

A comment to what was said earlier. Localization error will happen no matter how the recording is if the room has a complete dampened rear wall. It doesn't only sound unatural because it's dead, but it's also incorrect due to how we perceive sound. This has been known for a long time.
 
A comment to what was said earlier. Localization error will happen no matter how the recording is if the room has a complete dampened rear wall. It doesn't only sound unatural because it's dead, but it's also incorrect due to how we perceive sound. This has been known for a long time.

If by "rear wall" you mean the wall behind the speakers, then your comment is contrary to both my experience (I do precisely that) and Floyd Tooles (see his book).
 
As will HomImpulse. My preference is Holm because its full capability is free.
I can't find anywhere in Holm to generate ETC graphs. :confused:

While I agree Holm is a great tool if your budget is "for free", it really can't compete with ARTA in either functionality or usability.

ARTA has a full suite of room acoustic measurement tools that Holm lacks, and for measurements that both programs can do I generally find ARTA does a much better more accurate job of it and is easier to use.

The free version just prevents you from saving impulse response measurements permanently to a file - not something that's really an issue for a quick test proposed in the thread - it will still let you take a measurement and generate any ETC (or other) results from that and save the end results including copy/pasting graphs into jpeg/png. (You just can't go back later and generate more results from the original measurement once the program/impulse response is closed, unless you re-measure)
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.