Fe83en with OB and measurements

[tuxedocivic]

In MY room and for MY ears, I have to agree that the OB is the better sounding, cradeldorf.

I wanted to see if I could objectively show a difference between OB, sealed and ported enclosures, for this driver.

You can't really say that though, cause you haven't compared them. You've only listened to the OB, and a sealed enclosure that would be very similar to the OB due to it's relatively massive size. On top of all that, you're using a driver that can't do sub-200hz well regardless of the baffle/enclosure type.

 

[chrisb]

So, you've already decided that the OB solution works for you, with this driver, current room / system, etc. - that's great.

Many of us have tried OBs, some more than once, and don't find the magic - (I 'm definitely in that camp), and trust you and cradeldorf agree that's just as valid

Neither have I yet caught the measurement or modeling bug - as I said earlier, I rely on the skills of others for that, but I should think you'd expect to find different results between the 3 named types of "enclosures" for this driver.

If you're truly interested in compiling a meaningful data set, you should probably try at least one pair of more "appropriate" enclosures such as those I mentioned a couple of posts ago - i.e. somewhere between 3 - 6 liters, and probably not sealed at all?

 

[picowallspeaker]

Nice to see that little speakers in OB produce those graphs 😀

 

[Bando]

A few helpful hints:
1. Looking at the last graph you posted, this speaker is going to sound shouty. A positive spin on this sound can be called detailed, revealing, ruthless. However, a quick comparison with a neutral (read flat response) speaker will show that it is quite harsh on most recordings.

I wouldn't worry too much about the 2.5 kHz dip. There's not much you can do about it unless you have DSP.

2. You want to see something like the attached.

Even then you are lacking bass. But you mentioned you plan to add helper woofers.

3. Try an inductor with a shunt resistor. That should help flatten the response above 100-200 Hz or so. You can look at Jeff Bagby's excellent PCD software to model the inductor/resistor filter and simulate the response. There's some newer simpler versions of the original excel spreadsheet on these forums. Check software pages.

If you hit the dark black line I drew, you should have some reasonable balance, but you will run out of steam pretty quickly. This can be perceived as sounding dead, flat, boring, lifeless, compared to the original sound. Don't blame the filter, it's because you can't turn it up. Your problem is the driver.

Finally, keep up the good work. You are already far ahead of the rest because you can measure. Next step is a good guide on measurements. D'Appollitto's book is excellent and a must have. It will save a lot of time and frustration. Before you do anything else, order the book.

Good luck!

I was able to find testing loudspeakers on audioxpress via google. I have read part 1 and he recommends using a CSD plot for finding decay for getting a quasi-anechoic freq. response. Do you agree? If so, what limits should I have 40 - 90 dB (y axis), 20 - 20khz (x axis) and what 300 ms (z axis)?

My understanding is that waterfalls also help identify room modes and early reflections.