Well, this is certainly strange.
For the rest of you who want to use the vocabulary correctly, I'll take a stab at the basics.
If I put a signal into a device and it is composed of 8 components, I will measure 8 components at the output. However, the amplitude and phases of these components may have been altered by the device. At this point, there is no non-linearity and consequently there is no distortion. If the amplitudes and the phases have been altered by the device, that is okay. Both those are perfectly fine linear operations.
However, if the output contains more than the original 8 componets, then there is a non-linearity. Those additional components (not in the original input) are distortion.
This is an essential aspect of linear systems analysis. Why these guys are trying to confuse you is beyond me. They know better.
The idea that a device is linear if it puts out a flat spectrum is silly and the notion of "linear distortion" is an oxymoron.
I guess I am not "new age". When you have a technical conversation, you need to use technical terms correctly. Believe me, you would be eaten alive if you were ever at the blackboard in front of an expert group and were sloppy in your communication. My collegaues would have a good chuckle if they saw the statements being made in this thread. Some would simply be depressed that the educational system was appearently failing.
I'll stop here. These guys know better ....
For the rest of you who want to use the vocabulary correctly, I'll take a stab at the basics.
If I put a signal into a device and it is composed of 8 components, I will measure 8 components at the output. However, the amplitude and phases of these components may have been altered by the device. At this point, there is no non-linearity and consequently there is no distortion. If the amplitudes and the phases have been altered by the device, that is okay. Both those are perfectly fine linear operations.
However, if the output contains more than the original 8 componets, then there is a non-linearity. Those additional components (not in the original input) are distortion.
This is an essential aspect of linear systems analysis. Why these guys are trying to confuse you is beyond me. They know better.
The idea that a device is linear if it puts out a flat spectrum is silly and the notion of "linear distortion" is an oxymoron.
I guess I am not "new age". When you have a technical conversation, you need to use technical terms correctly. Believe me, you would be eaten alive if you were ever at the blackboard in front of an expert group and were sloppy in your communication. My collegaues would have a good chuckle if they saw the statements being made in this thread. Some would simply be depressed that the educational system was appearently failing.
I'll stop here. These guys know better ....
Last edited:
I would like to make a minor point about something that I was assigned to do 43 years ago, when I was first hired at Ampex. They gave me the job of calculating the linear distortion correction circuit components for a linear 120 ips instrument tape recorder that we sold to NASA.
I am embarrassed that I can't specifically remember how to do it, I guess I can't remember everything, but I do remember complex math and I used a mechanical calculator. Trust me it was drudgery for anyone else, but me.
I am embarrassed that I can't specifically remember how to do it, I guess I can't remember everything, but I do remember complex math and I used a mechanical calculator. Trust me it was drudgery for anyone else, but me.
Last edited:
The irony is that the people with the technical background here have no problem with both linear and non-linear distortions. Perhaps consideration of error functions (Malcolm Hawksford is but one exponent of these) might help you? Both kinds of distortion show up when error functions are considered.
Of course, nobody else wanted to do the calculations, but I liked running the mechanical calculator and used to drive my office mate crazy, calculating with the maximum number of places. You see, I had just spent a year a Friden, and this was a Friden calculator, so I was already broken in on it. I used to do M derived filters as well. I had only been out of college for about 1 year, so I was easily able to do the math. I was glad to be useful.
It's just a matter of definition. Some guys got together and decided they needed a term to describe deviations in freq response without a non-linearity in the transfer curve. Someone said 'linear distortion'? and they said, OK, that's how we will call it.
Is it a good choice? Who knows. Problem with definitions is that they are very persistent
jd
I've done a stint at Philips's labs working on color TV deflection coils. We would send out newbees to the warehouse to get an horizontal deflection angle....
One guy got smart and didn't return until next morning asking innocently if we wanted him to continue his search.
jd
An official definition of the phrase ´distortion` is this one:
Definition: distortion
We had similar discussions with syn08 about linear vs. nonlinear and harmonic vs. non-harmonic distortion last year in the marvelous cable thread and it occured that even these fairly common terms were not as definite in use as one would have thought.
Definition: distortion
We had similar discussions with syn08 about linear vs. nonlinear and harmonic vs. non-harmonic distortion last year in the marvelous cable thread and it occured that even these fairly common terms were not as definite in use as one would have thought.
Well, It depends ...
Depends what you mean by "tele"
Depends what you mean by "...port ..."
Depends what you mean by "quantum"
Do I believe in results of experiments?
More than in models that explain them. I can't repeat such experiments, so the only what I can do, either to believe, or not. Speaking of models, I don't need to believe in them in order to use them as working tools, while they work. However, when I discover better tools, I would rather use them.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.