Jeff Wheeler wrote on Tue, 04 January 2011 01:57 |
[I'd like to see the specified useful frequency response range actually used for specified peak power testing, without driver damage or outrageous distortion from cone flap. .
|
I hate to highjack this thread-as this is a totally different subject and has nothing in common with the origional post.
The problem with that "test" is what to use for the source waveform. Some say pink noise-OK what crest factor? 6dB-12dB. That makes a lot of difference.
Or should you use a specific "curve"? Which one? There are all sorts of standards that the developers each think theirs best describes a music source. What kind of music. Metal and classical have very different freq content. Most of the standard curves roll off the low freq, yet it is the low freq that is actually putting the strain on the cabinets.
Have you ever looked at the typical time vs freq content of a song? I am thinking of starting a thread on that one. but since it is time that produces heating, I would think that it should be considered-but I have not seen any of the "standard" curves that have what looks like it agrees. Then again-musical styles vary.
If one manufacturer uses one curve and somebody else uses a different one-how do the two compare?
How much distortion is "to much".
Speaking from a manufacturer perspective, we have been asked to provide certain specific data. Yet no other manufacturer provides this data. So somehow the cutomer who requested it is going to use it to "evaluate" the product and then attempt to compare it to a product that does not have that data.
How is that possible? Let's say they see something they don't like. What about the other product under consideration? They don't provide the data, so they get a "free pass"? How does the customer know that their data is better or worse-if they don't have it?
There are some things that can be kinda compared. And then other things are very dependant on the measurement conditions.
I do agree, I would LOVE to have a standard that all manufacturers would have to adhear to. But I don't think that is gonna happen in our life time. It is hard enough to get manufacturers to get the spec numbers and their own measured data/graphs to agree.
Even something as simple as the -3dB point in freq response. It SHOULD be the freq at which the response is 3dB down from the rated sensitivity. The two HAVE to be tied together-but very rarely are. Take a look at the graphs yourself.
Let's say a loudspeaker has a sensitivity of 100dB. Then the -3dB point would be the freq that is at 97dB on the curve. SImple as that. Now if the manufacturer wants a lower cutoff number, then they need to state a lower sensitivity. Or if they want a higher sensitivity, then they need to be willing to settle for a higher cutoff freq.
You can't have both.
I better stop now before I get in trouble. I can get very "emotional" when it comes to specs and honesty