Okay, here is what's going on with the Fountek FR88EX.
Here is the frequency response measured in Jeff's box.
It's not bad actually. There is an expected baffle step loss below 1kHz. That is typical for any driver used in a small box.
Here is the effect of Jeff's grill.
Great job on the grill. This thin design has a very minimal effect on the response.
The off axis response looks pretty typical of a small driver too.
The impedance looks good and is free of any box resonances. Tuning frequency is just over 65Hz.
And here are the two speakers. There is a little variance between the two in the lower range, but it's not too bad.
This is where the rubber meets the road though. If there was just one measurement that you can do that will tell you more about how a speaker will sound it's the cumulative spectral decay. It will show you stored energy or ringing issues over time. Stored energy or cone break up is very easy to hear. It will add a little residual ring to everything. And this little driver clearly has some issues in that department. Here is a csd response.
There is actually a peak in the response above 20kHz that shows up in the csd. The top of this peak is where the computer sets the 0 on the scale. So everything is artificially pushed down on the scale. Let me bring the whole scale up 5db and you will see that there is significant break up in the 1kHz to 4kHz range. It is taking a long time to go away.
This is actually typical metal cone issues. For some reason a lot of people have been lead to believe that these stiffer cones will have lower distortion. As that lower distortion will equate to cleaner sound. The typical distortion levels are 40 to 50db down from the average output and only +/- a couple of db between various drivers. It's really hard to hear those differences because they are at such a low level. Cone break up adds a ringing as stored energy that continues to ring long after the original note has ended. This is easy to hear.
So I added a small inductor and Mills resistor to this as a baffle step loss compensation filter. This will help the sound a lot. See the before and after curves.
This keeps 1kHz range and above from being pushed forward in the sound stage. So imaging is improved And check out the csd.
If you flip back and forth from the first csd posted then you will see that the range from 1kHz and down is up now. But if you really consider what is going on then you can understand that with the output of the 1kHz range and down being equal, then the break up from 1khz and up has been reduced by 5db. A 5db reduction in output means it didn't play near as loud and wasn't as excited. So it induces less ringing. It is like looking at the second csd response that I posted that looked really scary and having that graph from 1kHz and down, but having the results of the first one from 1kHz and up. In short, at the same output level the ringing will be less sever.
Compare that with the csd of the LGK's that I posted in this thread:
http://www.audiocircle.com/index.php?topic=120804.0
Now can you see how the spectral decay can tell you more about how a driver will really sound than any other measurement? Pretty cool huh.
