Q of audible resonances.

[Dr.EM]

I personally find resonance the most common and most offensive artefact of speaker systems I've heard. No matter their origin, even quite a subtle emphasis of a note in a track can immediately spoil the realism of the sound for me.

Of course, a lot of resonance will be room borne, but before dealing with this we surely need the speaker as flat as possible. I can and have heard in speakers a single note come across too strong, this infers a resonance of 1/12th of an octave or smaller if other notes remain unaffected. Most measurements seem to be taken at 1/3rd or 1/6th octave, so my question is, is this really enough resolution? Should we in fact be paying more attention to 1/20th octave measurements and correcting peaks in this where possible?

The reason we don't could be that correcting these high Q peaks (locating their origin then tackling them) is too difficult, or perhaps they are considered inaudible (which is what I am disputing, I certainly think they are). These small peaks may only noticed on certain tracks, but I think they are important to address in making the most realistic sounding speakers.

What are your opinions? I'd like to see discussions on the causes of small resonance, your views on if they are audible and the importance of dealing with them. Some 1/20th octave measurements accompanied by your impression of the sound would be great

[twest820]

I usually use unsmoothed measurements for exactly this sort of reason. There are two basic approaches. Damp the drivers to flatten out their efficiency (typical of hi fi drivers) or leave the driver relatively undamped and EQ to get the desired SPL (typical of pro audio; personally I prefer the latter and would point out digital EQ is easy). Resonance sources are myriad but most of the high Q ones I've run into are cone breakup or similar issues such as hitting modes on the dust cap. You can EQ these out, but it's generally preferable to choose a driver better suited to the task.

[LineArray]

Originally Posted by Dr.EM

I personally find resonance the most common and most offensive artefact of speaker systems I've heard. No matter their origin, even quite a subtle emphasis of a note in a track can immediately spoil the realism of the sound for me.
...

Yes ...

Cabinet

- Cabinet walls vibrational modes excited directly from the drivers motor
- Cabinet walls vibrational modes excited by air pressure

- Cabinet room modes (also in Midrangers/Tweeters)

- Unwanted Helmholtz resonators (e.g. Driver/Basket/Cutout cavity)
(also in Midrangers/Tweeters)

- BR ports resonating at n X lambda/2


Drivers

- Cone resonances
- Basket resonances


Crossover Network

- Some Crossovers together with the drivers may behave resonant
due to unwanted capacitive and/or inductive effects of components
unsufficiently damped or compensated.

[Jonathan Bright]

I am in two minds about getting involved with this thread as I am not sure about the all terms being used and we may find ourselves talking at cross purposes if we are not having some sort of consensus.
I think Dr.EMs introduction of Q is a useful place to gather data. I am drawing on memories going back a few years now but there was a lot of good work done on resonance that is a little counter intuitive and the diy community can use it to have a lot of fun and gain some audible improvements in their gear.
I suspect that many people see sharp high Q peaks as a real problem but research has indicated that it is the broad LOW Q peaks that can be more troublesome. I think it was Barlow (originally Goodmans and Sandwich cones) in the UK mag "Wireless World" or the JAES (or both) who did a paper on this back in the 1970's. It was on the audibility of resonances in speaker enclosures. The point he made was that the broader, well damped resonances were excited far more frequently than the narrow ones and so coloured the sound much more.

Secondly, if you can get a look at the B&W paper that was the theoretical basis for their Matrix enclosures there is a lot to digest there. (I think it was on their website a few years ago, may still be.) Again, from memory and in precis form, the most interesting result of what was serious, decent research was that perceived, subjective, colouration was dependant on the reverberation time of the res'. So I think one of the conclusions were that resonance per se wasn't fatal to fidelity as long as it was well damped. I know, I know, that sounds like a result in conflict with the first point but I am trying to condense well over a dozen pages into as many lines.......

I think this is a fascinating area and one where the diy person has a lot of scope to fiddle meaning fully so I'd be interested in what other throw up.

Also hearing itself is interesting. Dr.EM I'd be curious if you have auditioned speakers with others and seen whether you are unusually sensitive to certain things that others may miss. Susan Parker over on her revolutionary "no feedback amp thread" knows that she has an issue with phase that dictates her choice in speakers.
A few weeks I was with a friend who almost couldn't listen to a cheap pair of towers he'd got for his kids which I just didn't even see any issue at all.

Anyway, thanks for starting this thread and I'll try and dig up a few photo copies of the papers I've referred to so I can engage more constructively.

Cheers, Jonathan.

[planet10]

Originally Posted by Jonathan Bright

a lot of good work done on resonance that is a little counter intuitive...
I suspect that many people see sharp high Q peaks as a real problem but research has indicated that it is the broad LOW Q peaks that can be more troublesome.

I think it pretty well established that high Q resonances are less audiable, even when excited (due to narrow bandwidth, they are harder to get going too). Toole covers off the research in his book.

dave

[planet10]

Originally Posted by LineArray

Yes ...

Cabinet
...
Drivers
...
Crossover Network

Oliver, good list. It is nice to be able to compare to my efforts and be able to tick each one off

dave

[Jonathan Bright]

Dave, name of Toole (O'Toole?) book pls?

Thanks, Jonathan

[PLB]

Hi Dr.EM,

This paper -- http://www.fesb.hr/~mateljan/arta/pa...m-aaaa2007.pdf , details detection methods of audible resonances and also shows the important conclusions of other researchers work as detailed on page 2, items 1 through 6.
Item 3 is in agreement with post #4.

Regards

Peter

[planet10]

Originally Posted by Jonathan Bright

Dave, name of Toole (O'Toole?) book pls?

Thanks, Jonathan

Floyd Toole Sound Reproduction

dave

[DDF]

The definitive series of resonance audibility studies was by Sean Olive. Audibility of resonance and delayed resonance via at least two different AES papers.

Good to see these finally garner some interest a dozen years after the fact!

Dave