Off-topic split off from http://www.diyaudio.com/forums/full-range/268626-vifa-tc9fd18-08-best-bang-buck-11.html 
At this point it is probably important to remind people that the only scientific study done on measured distortion of speakers (Geddes) is that the figures are meaningless.
dave

What is the point of increasing xmas when all these drivers are useless below 500 Hz because of distorsion..
At this point it is probably important to remind people that the only scientific study done on measured distortion of speakers (Geddes) is that the figures are meaningless.
dave
Something you did not there well read or understood..At this point it is probably important to remind people that the only scientific study done on measured distortion of speakers (Geddes) is that the figures are meaningless.
I've compared the CSD for more drivers, which clearly shows response speed at certain frequencies.. Geddes does not dispute this..
As far as I know, he argues that THD measurements do not show good overview of driver.. but for now he do not offers better method for measurement of distorsion..
My untrained ears can hear 5% distortion vs 0.5% distortion at 50Hz or 100Hz. Guess which one sounds better? Many people may not know what a speaker that can do 100dB at 50Hz at sub 0.5% distortion sounds like because they are so used to speakers making 3% to 7% distortion, and they think they know what bass sounds like. Until you hear a speaker that is clean, you won't know. Once you know, it's like watching HD TV at 1080p and going back to CRT at 360lines. Not very satisfying.
That's why I claim that small fullrange drivers should not be used as a woofer, no matter how much xmax is..My untrained ears can hear 5% distortion vs 0.5% distortion at 50Hz or 100Hz..
3-4 inch drivers can be great to cover the midrange, and tweeter if they quick enough as B80 or 10F.. but they should not be forced to do something that they do not like..
That's why I claim that small fullrange drivers should not be used as a woofer, no matter how much xmax is..
3-4 inch drivers can be great to cover the midrange, and tweeter if they quick enough as B80 or 10F.. but they should not be forced to do something that they do not like..
Well it depends on your intended SPL listening levels, if you are using it like background music speakers at 75dB, the distortion can be quite low for the bass. See my DCR made with the TC9FD, the THD is -35dB at 60Hz, a very good figure. At 85dB the value goes up accordingly to -25dB, still quite tolerable.

If you want 100dB and low THD, sure you need a woofer or sub woofers and this is the reason I really like a FAST system with the TC9FD.
At this point it is probably important to remind people that the only scientific study done on measured distortion of speakers (Geddes) is that the figures are meaningless.
dave
Meaningless or just not well correlated to what we hear?
In any case less distortion is better. Don't you agree?
Meaningless or just not well correlated to what we hear?
Meaningless in the sense that until the numbers hit 25% or so there is no correlation with what people hear.
In any case less distortion is better. Don't you agree?
Low distortion -- in the all-encompassing sense -- should be as low as possible. But we have no way to quantify that, according to Geddes' experiment, what we currently measure is meaningless. So you can make no judgements on quality based on the graphs people post.
In Geddes' book there is a whole chapter on minimizing distortion. After he did the experiment he told people to just ignore that chapter.
Some people think they can look at the measures and predict which DUT is better -- there is far too much still unknown. Treat those people with some trepedation.
dave
Chalk up another one for the palpitations Cal. I don't even want to think about trying to get the drivers into that enclosure.
You thinking of this phrase Dave?
'The end result of this is that traditional measures of harmonic or intermodulation distortion are almost meaningless. They do not quantify distortion in a way that can, with any reliability, predict a human response to it while listening to music or movies. They do not correlate because they ignore any characteristics of the human receptor, an outrageously nonlinear device in its own right. The excessive simplicity of the signals also remains a problem. Music and movies offer an infi nite variety of input signals and therefore an infi nite variety of distorted outputs...
In the general population of consumer loudspeakers, it has been very rare for distortion to be identified as a factor in the overall subjective ratings. This is not because distortion is not there or is not measurable, but it is low enough that it is not an obvious factor in judgments of sound quality at normal foreground listening levels.'
Toole, F. Sound Reproduction: Loudspeakers and Rooms (London: Focal Press, 2008) pp.452-453
YMMV.
You thinking of this phrase Dave?
'The end result of this is that traditional measures of harmonic or intermodulation distortion are almost meaningless. They do not quantify distortion in a way that can, with any reliability, predict a human response to it while listening to music or movies. They do not correlate because they ignore any characteristics of the human receptor, an outrageously nonlinear device in its own right. The excessive simplicity of the signals also remains a problem. Music and movies offer an infi nite variety of input signals and therefore an infi nite variety of distorted outputs...
In the general population of consumer loudspeakers, it has been very rare for distortion to be identified as a factor in the overall subjective ratings. This is not because distortion is not there or is not measurable, but it is low enough that it is not an obvious factor in judgments of sound quality at normal foreground listening levels.'
Toole, F. Sound Reproduction: Loudspeakers and Rooms (London: Focal Press, 2008) pp.452-453
YMMV.
Last edited:
A single THD number may not be useful except in the rare case where it is for example, where itbis exceptionally low - say less than -55dB from 200Hz to 15kHz at some reference level like 90dB. When it is that clean - there IS an audible difference with a speaker that has several percent THD over same range.
Third harmonic or odd orders are more objectionable which is why I like to post graphs of 2nd thru 5th order.
Third harmonic or odd orders are more objectionable which is why I like to post graphs of 2nd thru 5th order.
Some people think they can look at the measures and predict which DUT is better -- there is far too much still unknown. Treat those people with some trepedation.
If you read Geddes' book, surely you must have read Toole's too. In the end, he shows a clear relationship between preference and loudspeaker measurements. Turns out that there are a set of measurements that can predict listener preference.
This little TC9 is one heck of a driver. Vifa hit the sweet spot with this one: smooth response, copper cap on the pole piece, and ultra-low price. I'm building a corner line array, similar to Wesayo's, with this driver.
...you must have read Toole's too. In the end, he shows a clear relationship between preference and loudspeaker measurements. Turns out that there are a set of measurements that can predict listener preference.
Within the limits of their testing & testing environment.
dave
A single THD number may not be useful except in the rare case where it is for example, where itbis exceptionally low - say less than -55dB from 200Hz to 15kHz at some reference level like 90dB. When it is that clean - there IS an audible difference with a speaker that has several percent THD over same range.
Third harmonic or odd orders are more objectionable which is why I like to post graphs of 2nd thru 5th order.
Lumped THD is worthless. I prefer 2nd - 7th order which may indicate something useful.
Second and third are enough, for me CSD is more important..Lumped THD is worthless. I prefer 2nd - 7th order which may indicate something useful.
A bunch of conclusions which is in this book was made on the basis of tests which involved people who would have previously passed the training.. we do not have any information what kind of training, we only know that it is essentially influenced their later responses.. practically, Toole has no evidence for what he claims.. also, keep in mind that all of his research was sponsored by his former employer.. who knows what they actually research findings have..Toole, F. Sound Reproduction: Loudspeakers and Rooms
Second and third are enough, for me CSD is more important..
And for me, 2nd & 3rd are not enough and CSD has almost no importance. 😉 Nor am I alone in the latter, (Zaph for example doesn't regard it as being of any real value either).
As for Floyd Toole, nothing is perfect nor pretends to be. But I would attach more credence to his work than that of the vast majority.
Meaningless in the sense that until the numbers hit 25% or so there is no correlation with what people hear.
Low distortion -- in the all-encompassing sense -- should be as low as possible. But we have no way to quantify that, according to Geddes' experiment, what we currently measure is meaningless. So you can make no judgements on quality based on the graphs people post.
In Geddes' book there is a whole chapter on minimizing distortion. After he did the experiment he told people to just ignore that chapter.
Some people think they can look at the measures and predict which DUT is better -- there is far too much still unknown. Treat those people with some trepedation.
dave
With wide-band drivers I'd think there's a good chance to hit those numbers easily. If distortion would be meaningless Geddes would advocate full range speakers. He doesn't.
Devices that measure better sound better. Isn't that what Klippel is all about?
Geddes also proposed a new metric for distortion measurements. Free download at his site http://gedlee.azurewebsites.net/Papers/papers.aspx
Last edited:
I don't fell that measured distortion is that important, even thou distortion in the global sense is. Often a multiway will not have as good a dispersion characheristic as a good full-range.
I used to believe that as well until I looked at the dispersion characteristics of wide-band drivers. They don't look good at all. An example of not very well behaved dispersion would be the Markaudio 7.3. A positive example is the Tymphany TC9.
More data can be found at Loudspeaker Raw Driver Measurements
Furthermore the directivity of full range drivers dramatically rises with frequency. Geddes presents some good data on his website at GedLee LLC
Multiways can go lower & higher & usually play louder than a full-range. They appeal to many humans that like to say "mine is bigger than yours", or more is better.
SPL requirements for movie playback are real. Skyrocketing distortion numbers of full range drivers are real. Those are simply system requirements based on desired SPL and not a weakness of character as you seem to imply.
...the dispersion characteristics of wide-band drivers....
The thing is that they are generally smoothly changing. That the dispersion tends to narrow is not the point.
Geddes likes to see a flat DI curve, Toole/Olive like to see a falling one.
I am also always suspicious of measures i see. I have seen measures of the same driver with quite different measured responses.Which one doyou believe?
dave
The thing is that they are generally smoothly changing.
I found this generally not to be the case:
Loudspeaker Raw Driver Measurements
Any cheap Behringer box is way smoother.
- Status
- Not open for further replies.
- Home
- Loudspeakers
- Full Range
- Limitations (or not) of current speaker measurement technology