Hi Earl,
Yes, I follow. But isn't a MP system dispersive if the amplitude isn't flat, DC to light (theoretically)? For example, a simple electrical HP or LP filter is MP but has non constant GD, thus it is dispersive, no? Anything that doesn't have linear phase vs. frequency is dispersive, no? I.e. the effective propogation velocity is frequency dependent.
Yes, I follow. But isn't a MP system dispersive if the amplitude isn't flat, DC to light (theoretically)? For example, a simple electrical HP or LP filter is MP but has non constant GD, thus it is dispersive, no? Anything that doesn't have linear phase vs. frequency is dispersive, no? I.e. the effective propogation velocity is frequency dependent.
Yes, but isn't "dispersive" more general. One could have a dispersive medium that was perfectly flat.
I think that all filters are dispersive, yes, but they can also be MP, meaning that there is a very tight relationship between the group delay and the frequency response. Dispersive could also be non MP. I guess what you saying is that dispersive is not the right term, and that is probably correct. Non MP group delay, or what I believe is called "excess group delay" is the culprit.
At any rate, I think that its clear that it isn't so easy to sort these things out.
I think that all filters are dispersive, yes, but they can also be MP, meaning that there is a very tight relationship between the group delay and the frequency response. Dispersive could also be non MP. I guess what you saying is that dispersive is not the right term, and that is probably correct. Non MP group delay, or what I believe is called "excess group delay" is the culprit.
At any rate, I think that its clear that it isn't so easy to sort these things out.
Allow me to throw into the discussion Bjørn Kolbrek's excellent summary about horns / wave guides.
Many thanks Bjørn, for writing such papers!
In part 2 some basic distortionn principles are discussed in a way as easy to understand as the subject allowes (also giving the scientifically explanation on "absolute phase" detection some of us are sensitive for):
part 1:
http://www.audioxpress.com/magsdirx/ax/addenda/media/kolbrek2884.pdf
part2:
http://www.audioxpress.com/magsdirx/ax/addenda/media/kolbrek2885.pdf
With respect to the double horn approach it may be interesting that part of the distortion mentioned should be cancelled due to its symmetric front-rear loading / radiation - not sure that impacts this may have on HOM at the moment
Michael
Many thanks Bjørn, for writing such papers!
In part 2 some basic distortionn principles are discussed in a way as easy to understand as the subject allowes (also giving the scientifically explanation on "absolute phase" detection some of us are sensitive for):
part 1:
http://www.audioxpress.com/magsdirx/ax/addenda/media/kolbrek2884.pdf
part2:
http://www.audioxpress.com/magsdirx/ax/addenda/media/kolbrek2885.pdf
With respect to the double horn approach it may be interesting that part of the distortion mentioned should be cancelled due to its symmetric front-rear loading / radiation - not sure that impacts this may have on HOM at the moment
Michael
tiki said:
Hi Timo, what you are referring to is what kept me off trying AMT for a (too) long time - besides the prices (ESS, Mundorf, Beyma) that are a plain rip off IMO.
But once you have listened to AMT this reservations are gone (and with double horn the "goose pimple factor" is even substantially raised - I can tell you) – also distortion measurements of standard AMT implemetations show exceptional results.
If Earl could outline a practicable method to measure HOM - I'd possibly would try to provode own results of the double horn approach to compare whith his OS wave guide - given he would agree and provides his measurements as well.
Michael
gedlee said:
As all these effects from mouth and contour reflections and higher order modes (due to wall reflection like in opical fibers) should show up in the impedance curve in odrer to do non-linear modulation - possibly a standardised processing of the mere electric impedance measurement would possibly make a good starting point?
To me it seems that it is highly a matter of interpretation / definition of - and distinction between - different distortion types - no ?
Michael
gedlee said:
Yes, as I said, anything that does not have a linear phase vs frequency is dispersive. A little OT here, but a good example is a typical 2-way speaker. In the crossover region the response sums flat (at least on axis) but, unless the crossover is of the transient perfect type, the crossover introduces dispersion. That is one reason why I have always been interested in TP crossovers. So then one gets into the argument of how much of this crossover induced dispersion is audible. One school says we can't hear it at all in a typical loudspeaker. Another says it's audible at low frequency, yet another says it's audible at higher frequency. Others argue that it depends on the transient nature of the source material. My argument has always been that if you can eliminate it, at least in the direct sound, without detrimental effects of the off axis and reflected sound, then why not eliminate it?
mige0 said:
I made a device to get a relative measure of the diffraction using the impedance curve.
This is on my site in "Mesure de la diffraction d'un pavillon" which means "Measure of horn diffraction", I know it's in french but with an automatic translator, i hope it's much easier to understand:
http://translate.google.com/transla...illon.html&sl=fr&tl=en&history_state0=&swap=1
With some adjustment you can use a microphone for a no relative measure, but the version in my site is really, realy simple to implement.
mige0 said:
Michael
Please also mention my paper disputing much of Kolbrek's paper.
john k... said:
John, I completely agree and I often do things like this. But the usual case is that there are tradeoffs and the "without detrimental effects" clause takes effect.
When I started my Summa designs, I did not have much data on the audibility of diffraction, but it was easy enough in the early designs to minimize it - so I just did it. Now I see that this was the right thing to do, but that is, quite honestly, just hindsight and the data that I know have was an attempt to explain the good subjective results that I got. I then went on to minimize all of the nearfield diffraction in my room as well as in the speakers themselves. This too was a pronounced effect.
Merci! Very interesting work.
I also like your page on the different tube shapes.
http://thend.chez-alice.fr/Audio/Tube_acoustique.html
I also like your page on the different tube shapes.
http://thend.chez-alice.fr/Audio/Tube_acoustique.html
thend said:
Sorry to disagree, but reflections and diffraction ARE NOT the same thing. They can be related to each other, such as a diffraction can cause a reflection, but don;t have to be. IF there is a reflection you can be pretty sure that there is also diffraction, but the converse is NOT true - if there is no reflection DOES NOT MEAN that there is no diffraction.