Pan said:
The averaging is synchronous to the signal. Thus the noise is reduced and the signal raised. Spectrum analysers that can do this are very rare and even when they can do it I don;t know of anyone who has tried. But in my method you have to sychronize the sine wave to the window length - I don;t think that any spectrum analyzer will do that.
soongsc said:
I don't mean to get too far off topic, but I understand the distinction that is being made here and would like to clarify it.
The definition of proof is conclusive demonstration of truthfulness, i.e. to prove a statement is to show conclusively that it is true. In math and logic this is possible; one deduces a statement starting from a collection of others and using only basic logical principles. This is conclusive, for if the proved statement were false, then either you have to throw out logical rules (eg allow things to be true and false at the same time) or one of the statements you started with was false. What you're actually proving is the implication (hypotheses --> conclusion).
The reason why this cannot happen in empirical sciences should be clear. Statements about the nature of reality, ie "laws," are of the form "xxx happens all of the time," and the only real assumption (and this is somewhat vague) that one starts with is that what is observed is reality. You cannot deduce the law from any finite set of measurements unless you also assume that verifying it a finite number of times implies it holds always. This would be a foolish assumption- certainly not one any self-respecting scientist would hold. The discrepency between the way a scientist uses this word and the popular usage lies in this assumption. Most people have no notion of rigor and apply this sort of half-assed induction, which is often sufficient for practical usage, to life in general. In any case, this is the reason why one cannot prove these statements. Notice that the statement can be shown to be false by providing a counterexample. That would constitute proof. However, the best we can do in the direction of truthfulness is know that it may be true because it is consistent with all observations.
Rybaudio said:
I thought about this again and you are correct - No one can ever prove that something is correct, but they can prove that something is false. This a well know problem in logic.
So I did prove that THD and IMD were unreliable because I showed at least one case where they were not reliable. I could never prove what was more reliable, I can only show evidence that it doesn't fail where its predicesors have.
You can certainly prove one mathematical equation to be true by using other mathematical methods, but the equations only proximate real world phenomona. Therefore is used only as a method in the design process to aid in design work, not to prove a phenomena to be true or false. We can only say whether the math is applicable or not for prediction of results, but never prove a physical phemonena to be true or phase. Whos knows if the right math is applied or not?Rybaudio said:
This is in line with what we are seeing in regards with distortion figures and audio reproduction. You can never tell how a system will sound just by the figures that were derived from math.
soongsc said:
That is completely irrelevant in the discussion at hand.
soongsc [/i] [B]Therefore is used only as a method in the design process to aid in design work said:
The point is not whether or not you have the right math- you don't even need math. It is that in principle it is not possible to prove physical laws. The nature of logic, of the statements of physical laws, and of our capacity to measure do not allow this.
You need to step up a level of abstraction to understand the statement that is being made here. There are more fundamental roadblocks than the ones you stated.
Using a large woofer and a compression driver makes a system in which nonlinear distortion is not going to be a consideration.
My idea exactly (and many many others), However the drivers that are presently on the market cannot get down to approx. 250 Hz and require an XO , at the minimum , of around 500 Hz. This IMHO is in the audible range. A good smaller WB driver that has the proper Qts and is positioned close enough to the LF driver according to the wavelength at XO can be a simple good performer either OB,waveguide,horn loaded with either a true dipole or a monopole with enough consideration to reflected energies at the rear of the cone and volume of the recieving wave/energy to accomodate an IB action at the XO displacement.Simply stated, if the XO is high enough it requires less volume to perform an IB action.
I object to a very high frequency XO to a ribbon or horn or such due to the spacing at the centerline of energy production at a given wavelength. A coiax solves many problems, but the options are limited.
I would be interested in seeing your designs Earl. Are there links?
ron
(the hardest thing you can do is to make something simple)
My idea exactly (and many many others), However the drivers that are presently on the market cannot get down to approx. 250 Hz and require an XO , at the minimum , of around 500 Hz. This IMHO is in the audible range. A good smaller WB driver that has the proper Qts and is positioned close enough to the LF driver according to the wavelength at XO can be a simple good performer either OB,waveguide,horn loaded with either a true dipole or a monopole with enough consideration to reflected energies at the rear of the cone and volume of the recieving wave/energy to accomodate an IB action at the XO displacement.Simply stated, if the XO is high enough it requires less volume to perform an IB action.
I object to a very high frequency XO to a ribbon or horn or such due to the spacing at the centerline of energy production at a given wavelength. A coiax solves many problems, but the options are limited.
I would be interested in seeing your designs Earl. Are there links?
ron
(the hardest thing you can do is to make something simple)
Getting WAY back to the topic at hand...
This thread (started out) extremely interesting to me because I'm still just a sophomore designer, doing my own thing. However, I can't help but notice that drivers measured with really low distortion are held as some sort of panacea amongst DIYers. (eg: "The Vifa ______ is a real steal at its price, having far lower distortion measurements than anything in its price range" (no offense to Vifa)) But I've heard designs using these heralded drivers and I can't say that I felt any inclination to listen further. Not that there wasn't anything WRONG with these designs using low-distortion drivers... just that I couldn't actually hear any reason why I should choose it over another driver which has a much higher-measured distortion.
What has made the biggest difference to my ears with most designs is not the drivers, but rather their use as part of the system as a whole (how they balance FR, power response, phase integration, etc). Is this the gist of the study? Or are my ears just shot?
If it's NOT distortion artifacts people think they're hearing, what ARE they hearing that they believe to be said distortion?
-Paul Carmody
gedlee said:
This thread (started out) extremely interesting to me because I'm still just a sophomore designer, doing my own thing. However, I can't help but notice that drivers measured with really low distortion are held as some sort of panacea amongst DIYers. (eg: "The Vifa ______ is a real steal at its price, having far lower distortion measurements than anything in its price range" (no offense to Vifa)) But I've heard designs using these heralded drivers and I can't say that I felt any inclination to listen further. Not that there wasn't anything WRONG with these designs using low-distortion drivers... just that I couldn't actually hear any reason why I should choose it over another driver which has a much higher-measured distortion.
What has made the biggest difference to my ears with most designs is not the drivers, but rather their use as part of the system as a whole (how they balance FR, power response, phase integration, etc). Is this the gist of the study? Or are my ears just shot?
If it's NOT distortion artifacts people think they're hearing, what ARE they hearing that they believe to be said distortion?
-Paul Carmody
Well, when I listen to speakers, I usually think (in order of frequency of happening) :
99%- This thing sounds like a box ! (uncontrolled resonances, etc)
98%- Frequency response isn't flat at all, lol, who designed that thing ?
97%- This room sucks. I know you have a 15 square meter living room and an obsession for 2m tall Klipsch, but, lol, come on.
10%- Where is that damn crossover ? I can hear the breakup modes screaming (very common on high-eff speakers which cross a 12" or 15" at 2 kHz, most 15" scream at 1K, give it up already, make a 3 way, dammit)
20% - I think this driver has distortion (for speakers where the "bass" driver is too small and it's making more harmonic than fundamental)
5%- Where is that damn crossover ? I can hear the tweeter dying (when pumping the volume on a low order crossover)
4%- Horn special : what is that screeching sound ? (puts cottonballs in ears)
1%- I think this driver has distortion (for all other cases)
0.1%- Let's try another set of cables (I'd really have to be drunk to pull of this one)
99%- This thing sounds like a box ! (uncontrolled resonances, etc)
98%- Frequency response isn't flat at all, lol, who designed that thing ?
97%- This room sucks. I know you have a 15 square meter living room and an obsession for 2m tall Klipsch, but, lol, come on.
10%- Where is that damn crossover ? I can hear the breakup modes screaming (very common on high-eff speakers which cross a 12" or 15" at 2 kHz, most 15" scream at 1K, give it up already, make a 3 way, dammit)
20% - I think this driver has distortion (for speakers where the "bass" driver is too small and it's making more harmonic than fundamental)
5%- Where is that damn crossover ? I can hear the tweeter dying (when pumping the volume on a low order crossover)
4%- Horn special : what is that screeching sound ? (puts cottonballs in ears)
1%- I think this driver has distortion (for all other cases)
0.1%- Let's try another set of cables (I'd really have to be drunk to pull of this one)
Re: Getting WAY back to the topic at hand...
To me the system design virtually dominates the problem. I truely believe that I could make a great system out of almost any drivers - within the limitations of the driver style, such as direct radiator vs. compression driver.
There is no accounting for what people "think" that they hear, but for my money, most of the time its not nonlinear distortion, no matter what they say. People do tend to loosely use the word "distortion" whenever something does not sound right and, of course, any deviation from ideal is distortion, but we need to be better quantified than that - are we talking about linear or nonlinear distortion?
Undefinition said:
To me the system design virtually dominates the problem. I truely believe that I could make a great system out of almost any drivers - within the limitations of the driver style, such as direct radiator vs. compression driver.
There is no accounting for what people "think" that they hear, but for my money, most of the time its not nonlinear distortion, no matter what they say. People do tend to loosely use the word "distortion" whenever something does not sound right and, of course, any deviation from ideal is distortion, but we need to be better quantified than that - are we talking about linear or nonlinear distortion?
Re: Re: Getting WAY back to the topic at hand...
Drivers with a ragged response curve (eg: linear distortion), I'm not so interested in. Spaghetti-like crossovers are not my thing.
Hm. Now THAT sounds like a good challenge! Makes me want to go through Krutke's site and see if I can make a system out of drivers he claimed were "garbage" due to high amounts of non-linear distortion.gedlee said:
Drivers with a ragged response curve (eg: linear distortion), I'm not so interested in. Spaghetti-like crossovers are not my thing.
gedlee said:
A driver does not have low/non audible distortion just because it's a big one or a compression driver. That was my point. Dogmatic statements like the one I replied to makes little sense.
I can only conclude that yours and mine definition of 'great system' is lightyears apart.
Don't think there's a coincident that the cleanest most resolving speaker systems I have heard all have been in the very low (for being speakers) distortion camp.
Would you mind posting (or telling) what distortion levels the Summa has?
/Peter
Translated from Jerkanese into English, I think what you are saying is: "The best systems I have heard used drivers with very low harmonic distortion. Because of that, I question what you say when you claim you could make a great system out of pretty much any drivers, regardless of THD. Could you elaborate further on what you meant by your statement?"Pan said:
By the way, Pan, "Cleaneast" and "Most resolving" are highly subjective terms, and come across like something out of the ad copy for speaker cables or a power conditioner. I honestly don't know what you mean when you say them.