D'Appolito on What Matters


For anyone interested in evaluating loudspeakers and measurements thereof I found a very recent (November 10, 2016) article by the good Dr. Joseph D'Appolito. A man who quite literally wrote the book on speaker testing.

I share this knowing it will start endless flame wars, but still hopeful some will find it informative.

http://www.audioxpress.com/article/testing-loudspeakers-which-measurements-matter-part-1
erik_squires
I'll see your D'Appolito and raise you a Floyd Toole...

https://www.amazon.com/Sound-Reproduction-Psychoacoustics-Loudspeakers-Engineering/dp/0240520092
Thanks! for sharing- erik & randy-11
John Dunlavy:
http://www.stereophile.com/interviews/163/index.html#L4C6Gveos1Z7KVMl.97
Dunlavy's article is interesting too, though at fist blush, the idea of matching instrument radiation patterns seems bogus, and not what I read acousticians and other speaker makers attempting to do.

I'm an unqualified pretender though, so I'll have to do more investigation.

Best,


Erik
all hail the line source
I'm not sure all of the above named authors would exactly hail the line source.
no, that was a pro-Maggie comment

how about a book on the history of speakers?  the title could be:

"From Horns to Here"

with a sub-title of:

Edgar Villchur Died For your Sins
Now that was funny!
Something to mull over.
There is a question to be asked with D’Appolito speakers, that have the critical midrange handled by two drivers.
For them to be free of any type of problems, the two drivers that are doing the same job, have to be 100% T/S matched and work in absolute unison to be effective in producing those midrange frequencies as good as one driver can do.
I’ve never had two speakers measure identical T/S parameters even from the same batch.

Cheers George
In case others don't know, D'Appolito has designed many different types of speakers but he is most famous for the configuration that bears his name, the MTM (midwoofer, tweeter, midwoofer). Not all MTM designs are D'Appolito and not every speaker D'Appolito has designed are MTM. :)

As far as the mids matching, the design doesn't suffer from those problem so much. You en up hearing the average of the two drivers, assuming the speaker is properly designed. Some would even argue you'll get more consistent results by using 2 drivers, as the average will tend to vary less than any single speaker.

The D'Appolito also has improved room interactions vs. traditional 2-way speakers and in the midrange act closer to line sources.

I'd give them a listen, but not saying you should buy them.
^Can't comment on the D'Appolito's, but Dunlavy used a similar       W M T M W  driver configuration (with different cross-overs of course). 
It has been claimed that he had all parts tested, logged and matched accordingly, and then the cross-overs for each speaker was individually tweaked to match a reference.
http://www.stereophile.com/content/dunlavy-audio-labs-signature-sc-vi-loudspeaker-specifications#Gkx...
unsound,

If you are making high end, that makes sense! :)  It was probably a crap load of work then too! :)

Not just drivers, but caps and coils are notorious for varying. Depends of course on the type and maker. It's totally worth buying 1% resistors for this reason alone!

With automated testing and characterization software I imagine it's much easier now. Not to mention, if you are buying in bulk, you can request certain specs be tested before delivery.

Best,


Erik
Rumor had it that many suppliers were irked that he sent so many things back.
The last time I spoke with him, he was very excited about going to an all digital self powered multi-amped model. He was holding out for better chips. He said the amount of labor time would drop dramatically,  and though the initial project would be expensive, he thought that the cost of succeeding models would drop at a considerably fast rate. Regretfully, he didn't live long enough to see it to completion.
Sorry to hear that. I was actually just thinking that. It's super convenient to match two speakers via DSP. I wonder how much of the benefit of things like DiracLive or Audessey comes just from matching speaker frequency response hyper-accurately.