i take it your talking about post #897
what category does that fall under? and how does it affect amplitude or should i be calling that linear response?
Wow..about everything really....cone breakup...CD diaphragm breakup....more to the point, even simply increasing excursion....
I find cabinet resonances / driver mount rattles, to be "linear" until a certain SPL is hit...then whoa, lookout!
mark100
see that doesn't help me
cone break up: linear or not? what type of distortion?
i always regarded that as producing new frequencies so non linear.no?
on cabinet resonances to me there are two types. one is a band of frequencies that appears with increasing amplitude. the other is a single frequency that appears only when the exact frequency to excite it is contained in the program material. that kind can go unnoticed for a long time until the correct stimulus comes along.
so resonances are linear distortions?
see that doesn't help me
cone break up: linear or not? what type of distortion?
i always regarded that as producing new frequencies so non linear.no?
on cabinet resonances to me there are two types. one is a band of frequencies that appears with increasing amplitude. the other is a single frequency that appears only when the exact frequency to excite it is contained in the program material. that kind can go unnoticed for a long time until the correct stimulus comes along.
so resonances are linear distortions?
Last edited:
It depends on what exactly happens, if new frequencies are created at some point, then at that point it becomes non-linear.
Lots of things can be considered linear up to some point. An amplifier may be linear until it starts clipping, for example. At that point it is behaving in a non-linear way. If we turn down the level and the clipping stops, then it can be considered linear again.
I don't know how far to go into all this right now. Lots of things are more or less linear, but not perfectly so. If linear systems theory gets us close enough to the right answers for our purposes, then we can use that math, and consider the system linear for our purposes.
Having said that, I don't want to get too caught up in dealing with gray areas and exceptions. That should come after learning the definitions and getting used to thinking in established engineering terms. Later would be a better time to get into nuances and more complexity. Learn the basics first, is probably a good idea here.
on driver mount rattles some can be hidden when the volume is loud and only reveal there presence at lower levels and that assumes the right frequency is present as well.
by my reckoning that is a unique category of distortion i call "spurious noise" (oops now i might take flak from the RF crowd for using that one)
by my reckoning that is a unique category of distortion i call "spurious noise" (oops now i might take flak from the RF crowd for using that one)
I am not in the habit of categorizing these in any way. I think a realistic approach to improving performance is just trying to figure out what non-ideal physical scenario can be improved. Any for of distortion data only suggests that there could be a problem. However, the cause cannot be traced directly from the data.
A single driver can (almost surely will) have different response at different frequencies off-axis versus on-axis because of beaming at higher frequencies. If I follow this thread correctly, this is one of the "linear distortions" being discussed.
LsD
For those that are interested in the subject of Loudspeaker Distortion, see the work of Wolfgang Klippel. I have attached a copy of his article that appeared in Voice Coil Magazine. On his website, Literature, additional information may be found. He has also authored numerous AES papers as well.
In the attached article he groups distortion products into 3 categories, the first two of which, he addresses here in detail:
1) Regular Distortion - Linear
2) Regular Distortion - Non-Linear
3) Irregular Distortion
Regards,
WHG
For those that are interested in the subject of Loudspeaker Distortion, see the work of Wolfgang Klippel. I have attached a copy of his article that appeared in Voice Coil Magazine. On his website, Literature, additional information may be found. He has also authored numerous AES papers as well.
In the attached article he groups distortion products into 3 categories, the first two of which, he addresses here in detail:
1) Regular Distortion - Linear
2) Regular Distortion - Non-Linear
3) Irregular Distortion
Regards,
WHG
Last edited:
Regarding Klippel's categoration system and control theory, what Klippel refers to Irregular Distortion would fall under the control theory categories of linear or non-linear distortions depending on things like whether the Irregular Distortions create new frequencies or not.
Apparently, he sees no need to divide up his Irregular Distortion class into linear or non-linear types, but technically speaking he could.
Apparently, he sees no need to divide up his Irregular Distortion class into linear or non-linear types, but technically speaking he could.
i lost sleep again cause now i need to understand a new system of classification with respect to distortion.
i have a question that may help me understand something about loudspeaker distortion.
removing the room and box i'm left with a driver/speaker.
now on it's own the loudspeaker is still subject to problems such as shape geometry, mechanical properties, impedance etc (if there are factors i'm missing here please tell me).
so now i have a device that is the sum total of all linear and non linear properties which comprise it.
so is it wrong to conclude that the device's linear properties are affected by it's non linear properties?
i have a question that may help me understand something about loudspeaker distortion.
removing the room and box i'm left with a driver/speaker.
now on it's own the loudspeaker is still subject to problems such as shape geometry, mechanical properties, impedance etc (if there are factors i'm missing here please tell me).
so now i have a device that is the sum total of all linear and non linear properties which comprise it.
so is it wrong to conclude that the device's linear properties are affected by it's non linear properties?
Probably that's wrong. Linear and non-linear classifications exist because they correspond to different kinds of math used with them. They aren't classified that way because of a cause and effect relationship between them.
Last edited:
It may or may not be, depending on the specific case.
It's kind of like you are asking if a persons age is affected by their gender. It depends. Women live longer than men on average, so in some age ranges there might be some such effect. But not in all cases, or at every age.
For speakers it depends too, but not like with people and age. It just depends on the particular details in a particular case.
EDIT: I think Earl Geddes was saying that speakers may be assumed to be minimum phase linear systems, so knowing frequency response and phase as a function of frequency is sufficient to imply a particular transfer function for the speaker. Other people may argue about the assumptions.
Last edited:
so then there are specific case/conditions where non linearity does effect response.
i fail to see how this has any correlation to loudspeakers (unless someone can show me evidence that loudspeakers have gender)
interesting comments here.
